The Racing Partnership Ltd & Ors v Done Brothers (Cash Betting) Ltd & Ors

Introduction

These two combined actions concern the alleged infringement of the claimants’ rights in respect of data relating to horseracing at, principally, six racecourses (the “Arena Racecourses”) owned by the second claimant, Arena Leisure Limited (“Arena”).

The first claimant, The Racing Partnership Limited (“TRP”) produces live betting and horseracing data collated at racecourses pursuant to agreements with the course owners and sells it to off-course (including online) bookmakers. In relation to the Arena Racecourses, TRP’s right to conduct this business, from 1 January 2017, derives from an agreement with Arena dated 13 May 2016 (the “TRP Media Rights Agreement”).

TRP also entered into an agreement with Northern Racing Limited, an associated company of the third claimant, permitting it to carry out similar activities as from 1 January 2018 from nine further racecourses. TRP entered into further agreements with the owners of seven independently owned racecourses (the “Independent Racecourses”), entitling it to carry on similar activities from those racecourses, in some cases from 30 April 2017 and in the remainder from 31 December 2017. I will refer to all the racecourses where TRP operated as the “TRP Racecourses”.

There were, across the two actions, originally six defendants, but confidential settlements were reached before trial with five of them. The sole remaining defendant is Sports Information Services Ltd (“SIS”). SIS previously had the right to collect and distribute to off-course bookmakers data from the TRP Racecourses, pursuant to agreements between SIS and each of Arena, Northern Racing Limited and the owners of the Independent Racecourses. In each case, the term of the relevant agreement ended immediately prior to the commencement of the replacement agreement between the respective owners of the relevant racecourses and TRP.

The claim in essence is that SIS, notwithstanding that it lost the right to collect and distribute the data from the Arena Racecourses on 1 January 2017, continued to do so in ways which involved infringing the claimants’ copyright, confidence and contract rights. The overarching claim is that SIS conspired with others to injure the claimants by unlawful means. The trial was concerned with (and this judgment deals with) issues of liability only.

Before dealing with the claims in more detail, I will explain the nature of the data and the contractual relationships relating to the exploitation of the data.

Horseracing Data

Two types of horseracing data are involved: (i) betting prices; and (ii) factual information relating to the courses and the relevant races.

Betting Prices

The betting prices relevant in this case are the fixed odds either offered by bookmakers (both on-course and off-course), or offered and accepted between market participants on online betting exchanges.

A vital piece of information collated from racecourses in relation to betting prices is a single representative price for each horse in a race, sometimes referred to as the consolidated pre-race price, but which is known, when transmitted to off-course bookmakers, as a “Betting Show”. Each Betting Show is the product of an algorithm (which essentially calculates a form of average price) into which is fed a selection of the fixed odds being offered by a sample of on-course bookmakers. It is common ground that the task of identifying bookmakers to include within the sample involves an exercise of skill and judgment. Betting Shows are produced from time to time in the period of eight to ten minutes before the start of the relevant race.

The Betting Shows are vital information for off-course bookmakers, because they enable them to offer odds to their customers that reflect the prices being offered by the on-course bookmakers.

Moreover, as the start-time of a race approaches, the Betting Shows offer the best indication as to the likely “Starting Price” for each horse in the race. The “Starting Price” is the official single price pertaining to a particular horse at the start of the race. It is settled through an industry-standard process, in accordance with standards laid down by the Starting Price Regulatory Commission Limited (“SPRC”), once the race has begun. It is thus unknown until after the start of the race. In the case of the Arena Racecourses, it is produced in precisely the same way as the Betting Shows, based on prices offered by the same sample of bookmakers.

The Starting Price is of considerable significance because a high proportion of bets are placed on terms governed by the Starting Price. It is important to off-course bookmakers that the Betting Shows (and certainly the final Betting Shows before the start of the race) are as close an approximation to the Starting Prices as possible, so as to avoid customers who place bets on the basis of the Starting Price being disappointed as a result of large discrepancies between the prices at which bookmakers offer or take bets and the Starting Price.

Raceday Data

The second type of data relevant in this case comprises information specific to the racecourse on the day of the race, such as the weather conditions, the state of the course (the “going”), the withdrawal of any horses, changes in jockeys, the start-time, the finish-time, any stewards’ inquiry and the result. This is referred to as “Raceday Data”. It is of particular value to off-course bookmakers as it enables them to take bets right up to the start-time of the race, but not beyond, and pay out winnings as promptly and accurately as possible after the conclusion of the race.

The commercial value of Betting Shows and Raceday Data

For the reasons I have outlined above, in order for off-course bookmakers to offer bets on horseraces to their customers, it is essential for them to have access to Betting Shows and Raceday Data as quickly as possible. It is also essential that the information is accurate and reliable.

There is accordingly a commercial value in being able to collate Betting Shows and Raceday Data and distribute them in real time to off-course bookmakers. It is for that reason that TRP (and SIS before it) are (and were) willing to pay substantial sums to the owners of racecourses to be allowed the exclusive right to collate and distribute such information to off-course bookmakers. The value in the information is, however, short-lived. Its value lasts for a matter of minutes only, and much of it does not extend beyond the start of the race.

In order to protect its interest in such information generated at the racecourse, including its interest in granting, for reward, the exclusive right to collate and distribute it, Arena imposes restrictions on the use to which those who attend races can make of the information they acquire. Those restrictions are contained in terms and conditions of entry, published on Arena’s website, signposted at the entrances to the racecourses and incorporated into contracts with those that enter (for example upon purchase of a ticket) (the “Arena Terms”). The Arena Terms are central to the claimants’ case that the distribution by SIS of Raceday Data to off-course bookmakers constituted a breach of contract, and I will return to consider them in more detail below [see section F3 below].

The commercial arrangements entered into by TRP for collation and distribution of Betting Shows and Raceday Data

TRP Media Rights Agreement

Pursuant to the TRP Media Rights Agreement entered into between TRP and Arena dated 13 May 2016, TRP was granted (among other things) “Access Rights” and “LBO Rights” in respect of the Arena Racecourses. The Access Rights included “…the right to create, collect, collate, arrange, present and/or store any Raceday Data at the Course(s) and freely to use and exploit that Raceday Data…” Raceday Data comprised a list of approximately 30 items, including the Going (i.e. the state of the course), Betting Shows, Starting Prices, the Off, non-runners and withdrawals. The LBO Rights included the right to broadcast and distribute live to licensed betting offices audio-visual coverage of races (including commentary) and Raceday Data.

The rights granted to TRP were described as “exclusive, meaning that, subject always to [certain reserved rights], neither Course Owner nor any of its Affiliates shall, nor shall they permit or authorise any third party to … (c) create, collect, collate, arrange, present and/or store any Raceday Data at the Course(s) and/or transmit, distribute or otherwise make available any such Raceday Data.”

TRP/PA Agreement

TRP delegated to the Press Association Limited (“PA”) the collection and distribution of Betting Shows and Raceday Data. This was done pursuant to an agreement dated 23 March 2016 under which TRP granted PA the “exclusive right” to collect “RaceDay Data” at the TRP Racecourses. Race-Day Data was defined to include essentially the same data that was defined as Raceday Data in the TRP Media Rights Agreement. Clause 8.1 contained an assignment from PA to TRP of intellectual property rights. PA was also appointed by TRP to validate the Starting Price (in accordance with the SPRC regulations) for each race at the Arena Racecourses. I will refer to the Betting Shows created for TRP by PA as the “PA Betting Shows”.

TRP/RDT Agreement

TRP also entered into an agreement (dated 24 August 2016) with Racecourse Data Technologies Limited (“RDT”) under which TRP granted RDT the right to access the racecourses for the purposes of providing access to RDT’s automated betting information collation system to on-course bookmakers, and providing “Betting Data” and the “Database” to the “Official Data Provider”, i.e. PA.

The “Betting Data” meant, in essence, the Betting Shows and Starting Prices for races at the racecourse.

The Database (the “RDT Database”) meant such part of any database in which the Betting Data was collated, arranged, presented and/or stored.

By clause 5.2, RDT assigned to TRP all intellectual property rights in the Betting Data and/or the RDT Database.

TRP/Bookmakers Agreements

TRP entered into agreements with over 100 on-course bookmakers for the exclusive use of their betting prices. Where the on-course bookmaker connects to the RDT system, their betting prices go into the RDT Database.

The commercial arrangements entered into by SIS for collection and distribution of (i) Betting Shows and (ii) Raceday Data

Until 31 December 2016, SIS was permitted to collect and distribute Betting Shows and Raceday Data from the Arena Racecourses pursuant to an agreement with Arena dated 22 July 2009 (the “SIS-Arena Agreement”). Under the terms of the SIS-Arena Agreement, SIS was granted the “sole and exclusive right and licence (without the right to grant sub-licences) to provide any and all Programming and Race-Day Data … to Subscribers”. “Programming” included live feeds taken from and relating to events at Arena’s racecourses, and “Subscribers” included off-course bookmakers.

As from 1 January 2017, notwithstanding it no longer had any contractual arrangement with Arena, SIS continued to provide a “feed” comprising both Betting Shows (the “SIS Betting Shows”) and Raceday Data relating to races at the Arena Racecourses to two of the major off-course bookmakers, known as Ladbrokes Coral (“LC”) and Betfred (“Betfred”) as well as to certain independent bookmakers.

LC comprises various corporate entities including Ladbrokes Betting & Gaming Limited, Coral Racing Limited and Ladbrokes Coral Group PLC (the first to third defendants in the action numbered 2017-000094). Betfred is operated by Done Brothers (Cash Betting) Limited (the first defendant in the action numbered 2017000093).

SIS continued to provide Betting Shows and Raceday Data to LC and Betfred until

There are certain further actions of SIS which it admits constituted breaches of the claimants’ rights. In light of SIS’s admissions, and given that I am not concerned in this judgment with issues of damages or other relief, I merely note these additional causes of action: (1) SIS entered into a data licence with TRP in May 2018 and thereafter transmitted to certain bookmakers data relating to racing at the TRP Racecourses in breach of that licence; and (2) between about July 2017 and October 2018 SIS collected Raceday Data relating to races on the TRP Racecourses from third party sources (the Irish Racing and Sporting Life websites) and inserted it into the feed it supplied to independent bookmakers, in breach of TRP’s database rights in a database referred to as the “Racing Store”.

I describe in the following paragraphs the arrangements made by SIS to provide the SIS Betting Shows and Raceday Data to LC and Betfred.

Betting Shows

From 1 January 2017 SIS had no contractual relationship with any third party relating to the creation of the Betting Shows. Instead it created the Betting Shows from information gleaned from third party sources. The precise nature of those sources is a matter of dispute between the parties.

SIS maintains (and the claimants accept) that its Betting Shows were created in the first place by its traders referring to (i) the “Early Prices” that were published in various places including in particular a publicly available comparison website known as Oddschecker, which displayed the prices being offered by all major bookmakers and (ii) the prices appearing on the two major online betting exchanges known as “Betdaq” and “Betfair” (together, the “Exchanges”). The Early Prices are those published prior to the period in which Betting Shows are produced. It is not the claimants’ case that SIS’s reference to Early Prices involved any infringement of their rights.

The prices on the Exchanges are presented as decimal points. SIS traders exercised judgment to convert those prices into fractional odds suitable for off-course bookmakers. Bookmakers earn their cut from betting by building a commission into the prices they offer (referred to as an “overround”). The Exchanges, on the other hand, charge a separate commission to parties who execute trades on the Exchange. It is accordingly not possible to translate the Exchange prices directly into fixed odds suitable for bookmakers. The decimal points on the Exchanges need to be adjusted to cater for the bookmakers’ overround.

The claimants contend, however, that SIS’s traders also referred to the prices appearing on Oddschecker during the eight to ten minutes prior to the start of each race. In that period, it is common ground that the prices appearing on Oddschecker reflected the PA Betting Shows which were being fed to the bookmakers whose prices appear on Oddschecker. The precise way in which SIS’s traders are alleged to have made reference to the PA Betting Shows has changed over the course of these proceedings but, in closing submissions, its case was limited to the traders having made reference to the prices on Oddschecker “in order to minimise differences between” the PA Betting Shows and the SIS Betting Shows. This is the subject matter of the claimants’ copyright and Database infringement claims which I address in detail below at section F1 below.

SIS accepts that its traders did make reference to the PA Betting Shows (via Oddschecker) on limited occasions in the early days of January 2017, but insists that this practice stopped at some point in January 2017.

Raceday Data

SIS collected Raceday Data through an arrangement with Tote (Successor Company) Limited (“the Tote”).

The Tote was originally established by the Racecourse Betting Act 1928. This made it lawful for the Racecourse Betting Control Board to operate a “totalisator” on approved racetracks. The Tote remained a public body until it was privatised and sold to Betfred in 2011. It has had a presence on British racecourses, including the Arena Racecourses, for many years.

Apart from a few occasions when the Tote operated “Totesport” fixed odds betting outlets on certain TRP Racecourses (which it did under separate agreements), the Tote has only ever undertaken a pool betting service. That service includes collecting certain elements of Raceday Data from racecourses and making it available to offcourse bookmakers via a designated data feed called the Tote Information Gateway (“TIG”). The Tote had carried out this service at the Arena Racecourses during the period covered by the SIS-Arena Agreement and continued to do so after 1 January 2017.

The arrangement between SIS and the Tote was formalised in heads of terms dated 30

The “Events” were set out in Appendix 1, and included all race meetings taking place at a number of racecourses including, but not limited to, the TRP Racecourses, depending on the time period, the time periods reflecting the different dates upon which SIS’s rights pursuant to its existing agreements with the TRP Racecourse owners terminated. Thus, from 1 January 2017 until 30 April 2017, the Events covered race meetings at the Arena Racecourses and a number of racecourses (referred to as the RMG Racecourses) that have nothing to do with the claimants. From 1 May 2017 to 31 December 2017 the Events included, in addition, three of the Independent Racecourses. From 1 January 2018 certain other racecourses owned by Northern Racing Limited were added and from 1 April 2018 the remaining Independent Racecourses were added. The Tote HoT also contained the following terms:

The claimants’ claims in outline

At the forefront of the claimants’ case (certainly by the end of the trial) is a claim in conspiracy. It is alleged that SIS conspired with LC, Betfred and the Tote to use unlawful means to injure TRP.

The alleged conspirators are related to each other as follows: SIS is wholly owned by Sports Information Services (Holdings) Ltd (“Holdings”); LC owns 23% of the shares in Holdings; the Tote owns 6% of the shares in Holdings; in turn, the Tote is wholly owned (indirectly) by Betfred; and Fred Done, the founder of Betfred, owns 8% of the shares in Holdings.

The allegation that the conspirators intended to injure TRP contains two elements.

The alleged unlawful means consist of the following:

In addition, in respect of three of the alleged unlawful means (infringement of copyright, infringement of a database right and breach of confidence) the claimants assert independent claims against SIS.

I will deal first with each of the substantive allegations of unlawful conduct in turn, before turning to consider the elements of the cause of action in conspiracy (and without prejudice, therefore, to the question whether the relevant unlawfulness, even if established, suffices for the purposes of the tort).

The alleged unlawful means

The claimants’ original case was that it is the owner of copyright in the database created by PA consisting of the compilation of the Betting Shows for each horse in a particular race (the “Betting Shows Database”).

By an amendment made on the first morning of the trial, the claimants added an alternative claim to copyright in the PA Betting Shows themselves (originally vested in PA but alleged to have been assigned to TRP pursuant to clause 8.3 of the TRP/PA Agreement). After the close of the factual evidence, the claimants abandoned entirely the claim to copyright in the Betting Shows Database, leaving only its belatedly added claim to copyright in the Betting Shows.

Does copyright subsist in the Betting Shows?

The claimants contend that the Betting Shows fall within the definition of “literary work” in the Copyright, Designs and Patents Act 1988 (“CDPA”). By s.1(1)(a) of the

Neither counsel was able to cite any case where copyright has been held to subsist in a single betting price (for example “Utopian Dream: 5:2”) or anything of a materially similar nature. The claimants relied on PCR Ltd v Dow Jones Telerate Ltd [1998] FSR 170, per Aldous J at 734-735, for the proposition that copyright can exist in statistical information, but the facts were far removed from the present circumstances.

The only case cited in which the issue was considered was Bookmakers’ Afternoon Greyhound Services Ltd v Wilf Gilbert (Staffordshire) Ltd [1994] FSR 723. In that case, the claimant (“BAGS”) claimed copyright in forecast dividends which it produced for greyhound meetings (as well as in racecards for those meetings). The forecasts were produced by inputting a number of factors, including the starting odds of all the dogs running in a race, into a computer into which had been programmed a formula. The output from the operation of the formula was given in money terms, expressed on the basis of a £1 stake. That was disseminated as, for example, “Bags Forecast £5”.

The applicable statutes were both the Copyright Act 1956 (because the infringement was alleged to have occurred while that Act was in force) and the CDPA (because an injunction was sought in light of threats of further infringement after the coming into force of the later Act). It was common ground that (apart from one matter which is irrelevant for the purposes of this case) any conclusion under the 1956 Act would be the same under the 1988 Act.

BAGS claimed that copyright subsisted (pursuant to s.2 of the 1956 Act) because the forecast dividends were an original literary work. Aldous J held (at p.731) that it was settled law that “…the word ‘original’ does not require original or inventive thought, but only that the work should not be copied and should originate from the author”, citing Ladbroke (Football) Ltd v. William Hill (Football) Ltd [1964] 1 WLR 273 at 291. He also held that it was settled law that:

“…to acquire copyright the author must have expended a sufficient amount of labour, judgment, literary skill or other such quality. What will be sufficient in any particular case is a question of degree and will depend upon the work produced and all the circumstances of the case. However the courts are not astute to enable one man to appropriate, without payment, the fruits of another's labours (see Ladbroke supra at p.291). I must therefore consider the production of … the dividend forecasts and decide whether sufficient labour, judgment and skill were used so as to render them original literary works.”

Aldous J held, applying this test, that although he saw no reason why the formula itself could not constitute a literary work, and that “thought, skill and labour” went

He went on (at p.736):

“I have already indicated that I do not believe that once the formula was derived, sufficient skill, labour and judgment is used when calculating the dividends from the starting prices supplied to acquire copyright. It is a repetitive job requiring a certain amount of education and thereafter a meagre amount of labour. Further I cannot believe that the person who, day after day, year after year, does the calculation is producing what the statute calls “an original literary work”. The words “BAGS Forecast” appear every time and the only figure that may or may not vary is the number. I do not believe that calculation using the formula will produce an original literary work every time the calculation is done. I therefore conclude that copyright does not subsist in the individual BAGS forecast dividends.”

SIS contends, on the basis of this decision, that TRP cannot establish that copyright subsisted in the Betting Shows. It points out that PA creates Betting Shows pursuant to a process devised by the SPRC for creating Starting Prices, and that in any event the Betting Shows are the product of the algorithm into which are fed prices from oncourse bookmakers. On the basis of the BAGS case, SIS contends that this is insufficient to create copyright in the output of the algorithm, namely the Betting Shows.

The evidence as to the process by which PA Betting Shows are created was contained in the witness statements of Simon Lamb and Nigel Burns, both employees of PA. Their evidence was accepted as true by SIS, and neither was required to attend to be cross-examined on their statements. That evidence can be summarised as follows (ignoring such technical detail as is unimportant to the points in issue). The prices offered by all of the on-course bookmakers are continually fed into, and updated on, an electronic database. One of the PA employees, known as the “validator”, who has access to the full list of bookmakers on course who publish prices, selects a sample of those bookmakers. That sample is used for the purposes of establishing Betting Shows and the Starting Price. The prices being offered by the bookmakers included within the sample are automatically fed into an algorithm (which has been created in line with guidelines issued by the SPRC). The algorithm calculates a representative price for each horse in the race. The prices for all of the horses in the race are transmitted

Because the same sample is used for creating Betting Shows and the Starting Price, its selection is subject to the rules of the SPRC. According to Mr Burns, within those rules considerable discretion is vested in the validator in selecting which bookmakers to include within the sample. The validator is required to exercise judgment, for example, as to: the size of the sample (within parameters laid down by the rules); the split as between “Rails” bookmakers (those operating between the fence and the Members and Tattersalls’ enclosures) and bookmakers within the Tattersalls area; whether bookmakers should be excluded because, based on past experience, they might cease displaying prices before the off, or not stay for the duration of the day, or have been subject to past suspension. The claimants contend that the product of the skill and judgment exercised by PA’s on-course operatives attracts copyright protection.

While I accept that there is some level of skill and judgment involved in the selection of the bookmakers to be included in the sample, so that copyright might subsist in a list of the bookmakers included within the sample, it does not follow that copyright subsists in the Betting Shows. For reasons which I develop below, I have concluded that even if copyright subsisted in the PA Betting Shows, the claimants have failed to establish there was any infringement by SIS. Accordingly, it is not necessary to reach a firm conclusion on the question whether copyright subsisted. Had it been necessary to do so, however, I would have found that there was no copyright in each Betting Show. In my judgment, the skill and judgment which goes into creating the sample is akin to the skill and judgment which went into the creation of the formula in the BAGS case. Accordingly, while copyright might subsist in a list of the names selected for inclusion in the sample, the process of arriving at each price was (in the words of Aldous J in the BAGS case) “pure routine work”, involving no sufficient skill, labour and judgment. I will deal with the remainder of the issues relating to infringement of copyright on the assumption, however, that copyright does subsist in the Betting Shows.

Was copyright in the Betting Shows assigned to TRP?

SIS also contend that, even if copyright subsisted in the PA Betting Shows, there has been no effective assignment of the right from PA to TRP.

The assignment of PA’s intellectual property rights pursuant to clause 8.1 of the TRP/PA Agreement is of “any and all Intellectual Property Rights in and to the data collected by PA and/or any of its employees, agents or contractors at the [Arena Racecourses] (except for the [Starting Price]), and any database comprising such data…”

SIS contends that the assignment was only of data “collected” by PA and that the Betting Shows were produced by, but not collected by, PA.

I reject that contention for two reasons. First, the assignment expressly excluded the SP. That indicates that, but for the exclusion, the SP otherwise would have fallen within the terms of the assignment. Yet the SP was created by PA in precisely the same way as it created Betting Shows. This suggests that data “collected by” is intended to extend to information (like the SP) that PA creates from the information it collates.

Second, by clause 3.1.1 of the TRP/PA Agreement, PA is licensed to “collect” Raceday Data. Raceday Data is defined (by Schedule 4) to include the Betting Shows. This is a further indication that the phrase “data collected” in clause 8.1 of the TRP/PA Agreement is intended to be read as including the Betting Shows.

Accordingly, on the assumption that there was copyright in the PA Betting Shows, I find that it was assigned to TRP.

Did SIS infringe TRP’s copyright in the Betting Shows

A copyright owner’s restricted rights are defined by s.16 of the CDPA. The only relevant rights for present purposes are the right to copy the work and the right to issue copies of the work to the public. Copying the work and issuing copies of the work to the public are therefore infringements, pursuant to, respectively, s.17 and s.18 of the CDPA.

The essential ingredient in the claimants’ infringement claim is, therefore, copying. Mr Mill QC accepted that infringement requires at least some element of copying, but that this was satisfied where there was a “substantial taking manifested and established by looking at the objective similarity” between the works.

By s.16(3)(a), infringement can consist in copying any substantial part of the work, and by copying either directly or indirectly: see also Sawkins v Hyperion Records Ltd [2004] EWHC 1530 (Ch) (affirmed on appeal at [2005] EWCA Civ 565; [2005] 1

The claimants accept that the mere consultation of a copyright work in order to crosscheck the result of one’s own equivalent exercise may not constitute an infringement, citing Kelly v Morris (1866) LR 1 Eq 697. That case concerned a directory of names and addresses. The defendant had copied the claimant’s directory, then sent out canvassers to see if the information was correct. Sir Page-Wood V.C, at p.702, gave the following example:

“In the case of a map of a newly-discovered island … he must go through the whole process of triangulation just as if he had never seen any former map, and, generally, he is not entitled to take one word of the information previously published without independently working out the matter for himself, so as to arrive at the same result from the same common sources of information, and the only use that he can legitimately make of a previous publication is to verify his own calculations and results when obtained.”

There are four logical possibilities as to the use which could be made of the PA Betting Shows if they were viewed in the course of creating the SIS Betting Shows. In the first three possibilities, SIS looked at the PA price, saw that it was different to its own proposed price, and then either: (1) did nothing; (2) changed its price to match

The claimants’ case in this respect has evolved over the course of the proceedings. The Re-Re-Amended Particulars of Claim alleged that SIS “copied and/or communicated to the public …. certain of the [PA Betting Shows]”. It was further alleged that the SIS Betting Shows were the same as or substantially similar to the PA

While the claim that SIS “copied” the PA Betting Shows had not been formally withdrawn, Professor McHale, the expert witness called by the claimants, commented in the joint report of the experts that: “I do not understand the Claimants’ case to be that SIS is “copying” or “heavily relying” on TRP prices. I am instructed that the Claimants contend that SIS needed to refer routinely to the TRP Show Prices to ensure that the prices SIS supplied to bookmakers were the same as, or very similar to, the TRP Show Prices.”

In supplemental submissions filed after the end of the hearing, the claim was limited further to the following: “SIS’s traders referred routinely to the PA Betting Shows on Oddschecker when creating SIS Betting Shows for TRP Races in order to minimise differences between the SIS Prices and the TRP Prices” (emphasis added). The claimants’ case as to copying, as ultimately refined, appears therefore to be limited to SIS moving its prices closer to, but not precisely copying, the PA Betting Shows and SIS relying on the PA Betting Shows to make no change to its own prices, where they were the same. In other words, the claimants assert that SIS’s behaviour fell within the third and fourth possibilities identified in paragraph 70 above.

In my judgment, such behaviour does not, as a matter of law, constitute infringement of copyright.

As to the third possibility, the claimants’ case depends upon the proposition that publishing a price that is close to, but not the same as, the price appearing in the PA Betting Show, constitutes copying a substantial part of the original work. Mr Mill QC drew an analogy with a musical work. In that context, taking someone else’s skill and effort in composing a work and using it in order to create a composition that was similar, but not the same, is sufficient to constitute infringement. He submitted that if SIS moved its own price closer to the PA Betting Show, having looked at the PA Betting Show, then that amounted to copying a substantial part of the PA Betting Show.

I reject this submission. In the context of a Betting Show, assuming that copyright subsists in something as simple as, for example, odds of “5:2” as applied to a particular horse, then another price is either the same as the original work or different from it. There is no half-way house where another price might be a copy of a substantial part of the original work. I do not see, for example, how odds of “5:1” or “10:3” could in any relevant sense be said to be a copy of a substantial part of “5:2”.

As to the fourth possibility, this would not constitute infringement because it is no more than that which was regarded as legitimate in Kelly v Morris. It would involve SIS having reached its own conclusion by its independent work, and then checking that conclusion for accuracy against the PA Betting Show. It is no different, in my judgment, if – as the claimants suggest might have happened in some cases – SIS reposted its first price after having seen the PA Betting Show. By definition, in the example given in Kelly v Morris, the defendant would have only republished its work after having checked its conclusions against the original work. Page-Wood V.C did not, however, regard that as an infringing act.

Accordingly, I find that SIS’s conduct, as alleged by the claimants, did not constitute an infringement of TRP’s copyright in the PA Betting Shows. That is sufficient to dispose both of the claimants’ direct claim for infringement of copyright and its claim that infringement of copyright was one of the unlawful means as part of the conspiracy claim.

Notwithstanding that finding, in case the matter goes further, I will address the matters that the claimants specifically rely upon as evidencing infringement. This, too, was the subject of development and refinement as the case progressed.

When the claimants amended, on the first day of the trial, to plead infringement of copyright in the Betting Shows, it was on the basis that they would rely on precisely the same particulars and evidence of infringement as they had relied on in respect of the claim for infringement of copyright in the Betting Shows Database. An obvious difficulty for the claimants in this respect was that they had not identified the particular Betting Shows in which they alleged that copyright had been infringed: the pleading referred to infringement of copyright in “certain of” the Betting Shows.

The expert evidence adduced by the claimants (which I explain in further detail below) identified a number of horses in particular races, where it was contended that the similarities between the SIS Betting Show and the TRP Betting Show could only be explained on the basis of actions which constituted an infringement of TRP’s copyright. I need not go into all of these examples, many of which were shown to be flawed in the course of preparation of supplemental expert evidence adduced on the part of SIS, because in closing submissions the claimants limited their claim (so far as the direct claim for infringement of copyright was concerned) to the Betting Show relating to just four horses (the “Four Horses”):

In relation to the conspiracy claim, however, where the claimants rely on infringement of copyright as one of the unlawful means, the claimants rely in addition to the Four Horses on two further matters: (1) admissions made by SIS and its witnesses that its traders made reference to the PA Betting Shows in connection with the creation of the SIS Betting Shows for at least part of January 2017; and (2) the fact that in relation to the dataset of 654 horses supplied with the supplemental report of the claimants’ expert witness, Professor McHale, in 36 instances the final SIS Betting Show was the same as the last price before the off showing on the Betdaq exchange, and that SIS’s expert witness, Mr Hunt, did not have an explanation for this.

It is the claimants’ case that SIS’s traders continued to make reference to the PA Betting Shows after the end of January, but Mr Mill QC submitted in closing that it was sufficient for his purposes (i.e. to establish unlawful means as part of the conspiracy claim) that on the basis of SIS’s admissions there was such unlawful conduct in January.

So far as the Four Horses are concerned, it is an oddity (given the claimants’ refinement of their case as to what constitutes copying) that in each case the final SIS price was the same as the final TRP price. It was not put, however, to any of SIS’s witnesses – in particular Mr Tompkins, a trader who created SIS Betting Shows during the relevant period, or Mr Conway, the head of trading at SIS – that SIS had copied the PA Betting Show either in producing the final price in respect of the Four Horses or any of the movements in price that took place within the period of the Betting Shows.

Mr Tompkins admitted in his witness statement that he and other traders on occasions in January 2017 made reference to the prices showing on Oddschecker during the period they were creating the SIS Betting Shows, but that this was only to ensure that the SIS prices were “reasonably close to” those appearing on Oddschecker and to check that “we weren’t too far from the PA Betting Shows”. While he was challenged on his evidence that this practice ceased at some point in January 2017, he was not challenged on the practice itself. In other words (consistent with the way in which the claimants’ case was refined after the end of the trial) there was no challenge to Mr Tompkins’ evidence that the extent of the reference made by SIS traders to Oddschecker was to ensure that the SIS Betting Shows were close to the PA Betting Shows.

Moreover, it was not put to any SIS witness that they had made any reference to the PA Betting Show in respect of any of the Four Horses.

I was provided with graphs, for each of the Four Horses and all the other horses running in the races in which the Four Horses ran, showing the movement in prices in the minutes leading up to the start of the race of (1) the PA Betting Shows; (2) the SIS Betting Shows; and (3) the Exchanges (for the first race) or just Betdaq for the remaining races. In viewing these graphs, it is important to keep in mind that the SIS Betting Shows would ordinarily be lower than the Exchange prices, by reason of the overround added by SIS traders in converting the Exchange price to fixed odds that a bookmaker could offer to its customers.

On the basis of the graphs alone (this being the only evidence available relating directly to the Four Horses) I am unable to conclude that any reference was made by

Overall, there is nothing in the graphs for the Four Horses that justifies the conclusion that the SIS Betting Shows were created by reference in any way to the PA Betting Shows.

Comparison with the graphs of all the other horses in the four races (none of which is relied on as an example of infringement of copyright) provides no further illumination.

In the absence of any direct evidence, the claimants rely on expert statistical evidence to demonstrate that the correlation between the SIS Betting Shows and the PA Betting Shows was greater than the correlation between the SIS Betting Shows and the Exchange prices.

A considerable amount of expert evidence was adduced, on behalf of the claimants by

The evidence was called in order to address two matters. First, to explore whether SIS created its final Betting Shows before PA published its final Betting Shows to a statistically relevant extent (the “leader-follower issue”). Second, to explore the correlation between, on the one hand, the final SIS Betting Shows and the final PA Betting Shows and, on the other hand, the last betting prices published by the Exchanges and the final PA Betting Shows (the “correlation issue”).

So far as the leader-follower issue is concerned, the experts agreed, in their joint report, that “the proportion of times for which TRP sets its Final Price first is found not to be statistically significantly different from the proportion of times for which SIS sets its Final Price first. This finding applies when considering only the runners for which SIS and TRP have the same Final Price as well as, separately, when considering only the runners for which SIS and TRP have different Final Prices. In all cases, the proportion of instances in which SIS or TRP set its Final Prices first is not statistically significantly different from 50%.”

The data set on which that joint conclusion was reached excluded prices which either TRP or SIS had reposted, that is, where the final price posted by them was a reposting of an earlier posted price. In a supplemental report professor McHale carried out the same analysis but including reposted prices, in which he concluded that in those circumstances the percentage of times when TRP was the first to post its Final Price increased to 53.1%. In the course of cross-examination he acknowledged, however, that there was nothing in this further analysis on the leader-follower issue which suggested SIS was following, in any meaningful sense, the TRP price, saying “you can’t draw any conclusions either way”.

I accept the submissions made on behalf of SIS that the conclusion in the supplemental report of Professor McHale on the leader-follower issue takes the matter nowhere. First, in the absence of a similar analysis including reposted prices for a statistically significant sample of prices including where the final prices of TRP and SIS were different, no meaningful conclusion can be drawn simply from the fact that when the prices were the same, 53.1% of the time the TRP final price was posted first. Second, it was pointed out in Mr Hunt’s response to Professor McHale’s supplemental report that the SIS reposted prices which caused the move from 50% to 53.1% arose (in all but a handful of cases) in the circumstances where SIS had posted an Early Price, TRP began its Betting Show first, SIS then reposted its Early Price as the first Betting Show and the price remained unchanged in the run up to the off. For both these reasons, the fact that the percentage of times when TRP posted its final price first increased to 53.1% when re-posted prices are taken into account does not provide any sound basis for concluding that SIS had copied the TRP price.

As to the correlation analysis, an overall limitation is that, given that there is no evidence as to the choice of races included in the data set analysed by the experts, it is impossible to draw any statistically relevant conclusions. Professor McHale acknowledged that the conclusions he could draw in relation to the sample would only be of statistical significance in relation to any greater set insofar as the sample is representative of that set and that, given that he was unaware of the basis on which the

Professor McHale was in any event careful to point out that he could not, and was not purporting to, draw conclusions as to causation from his correlation analysis. The most he could say was where he felt – given his understanding of the way in which SIS produced its Betting Shows – that the differences in correlation between TRP/SIS and TRP/exchange prices had not been explained to his satisfaction.

The essential difficulty in seeking to draw any conclusions from the statistical evidence as to correlation is that on SIS’s case as to how its traders produced the SIS Betting Shows, there are a number of reasons why there would in any event be significantly less correlation between the SIS Betting Shows and the Exchange prices, than there would be between the SIS Betting Shows and the PA Betting Shows. These include: (1) the extent of the overround added when converting the Exchange prices to fractional odds varied depending on the number of runners in a race, the length of the odds for a particular horse (a greater overround being added for longer odds runners) and the nature of the race (certain types of race, referred to as bad eachway races, where there was a clear favourite, such that each-way bets offered on other horses would not reflect their chances of winning); (2) the rounding of decimal prices, due to there being fewer rungs on a fixed-odds ladder; (3) volatility on the Exchanges, which called for judgment by the SIS Traders as to which prices to ignore; (4) the fact the last price on the Exchange (which is the price relied on in the correlation evidence) may not be reflective of the trend of pricing on the Exchange; and (5) the fact that even if average prices over the last few moments are taken as the last price on the Exchange, this still includes (and thus will be affected by) the outlying prices, whereas a SIS trader may well have simply ignored the outliers altogether.

Professor McHale agreed that each of these factors would have an effect on the expected degree of correlation between the different sets of prices. He had not, however, taken into account the cumulative effect of such factors when producing his calculations.

Accordingly, I find nothing in the expert evidence to indicate that, whether in relation to the Four Horses or more generally, SIS traders made reference to PA Betting Shows in creating the SIS Betting Shows.

As to the claimants’ reliance on the (few) occasions when the final SIS Betting Show is equal to the final Betdaq price, I similarly find nothing from which I can draw the conclusion the SIS traders were referencing the PA Betting Shows. The claimants’ argument is that on SIS’s case, the SIS Betting Show would necessarily be lower than the Betdaq final price, because the SIS traders would always add an overround, so that if there are (even a few) cases where the final SIS Betting Show is the same as the final Betdaq price, that demonstrates that SIS traders were not creating their prices from the Betdaq price. This argument is based on the evidence that in the dataset which was supplied with Professor McHale’s supplemental report, in 36 instances out of 654 (5.5%), the final SIS Betting Show was indeed the same as the Betdaq price. The dataset itself was not in evidence, nor was there any evidence as to the movement in Betdaq prices over the period of each of the relevant Betting Shows.

As I have explained above, there may be a number of reasons why a SIS trader, while following the prices on Betdaq, may choose to ignore the last in time price posted on the Betdaq Exchange (for example, because it was unrepresentative of the general trend of pricing on the Exchange). Without more information, it is not possible to conclude, merely because in 36 cases out of 654 the final Betdaq price was the same as the final SIS Betting Show, that the SIS traders were not creating the SIS Betting Show by reference to the Exchange prices. Moreover, this is no more surprising than out of the same dataset, in 40 instances (6.1%) the final PA Betting Show was the same as the final Betdaq price, given that the PA Betting Shows are based on oncourse bookmakers’ prices which include the bookmakers’ profit. There is no evidence that, on the 36 occasions when the SIS Betting Show was the same as the final Betdaq price, it was also the same price as the PA Betting Show.

The remaining factual matter in issue on this part of the case is whether the SIS traders ceased to make any reference to the PA Betting Shows in January 2017 or whether, as the claimants contend, they continued to do so thereafter. In light of my conclusion in relation to the Four Horses (being the only subject matter of the direct infringement claim) this is relevant only to the allegation that one of the unlawful means employed was infringement of copyright. Moreover, in light of the claimants’ submission that it is enough for their purposes that SIS admitted making use of the PA Betting Shows in January 2017 (and my conclusion that the use which SIS is now alleged to have made of the PA Betting Shows does not in law amount to infringement) it is strictly unnecessary to determine this issue. In case the matter goes further, however, I set out in Appendix 1 my review of the contemporaneous evidence relating to this issue.

On the basis of that evidence, which corroborates the evidence of Mr Tompkins, Mr Conway and Mr Ames that SIS traders had been told not to look at the PA Betting Shows from some point in January 2017, I make the following findings:

The claimants claim that TRP is the owner of a sui generis database right in the RDT Database pursuant to Directive 96/9/EC and the Copyright in Databases Regulations 1999 (the “Regulations”) and that SIS infringed that right. This is an alternative claim to the claim for infringement of copyright in the PA Betting Shows (which are not themselves to be found in the RDT Database): if sufficient skill and judgment went into the creation of the PA Betting Shows so that it is an original work and not merely a copy of entries in the RDT Database, then the act of referring to the PA Betting Shows could not amount to an infringement of a sui generis database right in the RDT Database.

TRP also claims ownership of a sui generis database right in the Racing Store, being a database of information including pre-raceday data, Raceday Data and Betting Shows. SIS admits, however, that TRP owned such a database right and that (without prejudice to its defence to the other claims) it infringed that right. Accordingly, I need not consider this issue any further.

Pursuant to the Regulations:

SIS accepts that TRP is the owner of a database right in the RDT Database, but denies that its actions constituted an infringement of a database right.

The claimants rely on the same conduct on the part of SIS to establish infringement of the database right, as they do to establish infringement of copyright in the Betting Shows. SIS submitted that its conduct was incapable of constituting infringement of the database right in the RDT Database, for three reasons.

First, the relevant conduct was confined to making reference to the PA Betting Shows. Each PA Betting Show was a new price, which although generated by the selection of sample entries in the RDT Database, was separate and distinct from the entries in the RDT Database. Even if there is no copyright in the PA Betting Shows, the manner of their creation means that making reference to a PA Betting Show is not making reference to any part of the RDT Database. Accordingly, such conduct could not amount to an extraction or re-utilisation of any part of the RDT Database.

Second, SIS’s actions constituted at most consultation of the PA Betting Shows. Such consultation is not an extraction or re-utilisation of the PA Betting Shows even if they formed a part of the RDT Database: see British Horseracing Board Ltd v William Hill Organization Ltd (Case C-203/02) [2005] 1 CMLR 15 at [54].

Third, in any event the extent of the consultation made of the PA Betting Shows was insufficient to constitute an infringement within the meaning of Regulation 16. In the British Horseracing Board case (above) the CJEU held that the prohibition on systematic extraction or re-utilisation of insubstantial parts of the contents of a database is an exception to the general rule that a substantial extraction or reutilisation is required. Accordingly, as interpreted by the CJEU, this prohibition is directed at “acts of extraction made by users of the database which, because of their repeated and systematic character, would lead to the reconstitution of the database as a whole or, at the very least, a substantial part of it.” SIS contends that even at its highest, the conduct alleged against it by the claimants does not reach that threshold.

I accept these submissions, to which there was no real answer from the claimants save that, in relation to the second of the above points, the claimants contended that consultation ceased to be consultation if the relevant information is then adopted and retransmitted as part of a commercial product. For reasons which reflect those I have already given in finding that there was nothing which amounted in law to copying the PA Betting Shows for the purposes of the copyright infringement claim, I reject the contention that SIS adopted and retransmitted any part of the Betting Shows. Accordingly, I find that there was no infringement of TRP’s database right in the RDT Database.

I have referred (at paragraphs 36 to 40 above) to the history of the Tote, the nature of its business (pool betting) and manner in which it supplied SIS with Raceday Data from 1 January 2017. The claimants contend that while the Tote was permitted by Arena to collect and distribute on-course data to bookmakers for pool betting purposes, it was prohibited from collating and distributing on-course data for any other purpose.

Accordingly, the claimants contend, the following actions of the Tote constituted a breach of its contract with Arena: (1) collection and distribution of data that was not required for pool betting; and (2) using and distributing to SIS for the purposes of fixed-odds betting the data that it was permitted to collect only for pool betting.

It is first necessary to determine what, if any, contract existed between the Tote and Arena. There are two possible candidates: first, an agreement (the copy of which in evidence is unsigned and dated “2013” (the “Tote Agreement”); second the Arena Terms, either the general terms and conditions contained in them, or the special terms and conditions relating to those attending for purposes relating to media rights.

The Tote Agreement

The recitals to this agreement refer to an earlier agreement between the Horserace Totalisator Board and Arena (the “2009 Agreement”), and to the transfer of the Tote to Betfred in 2011. It incorporates all of the terms of the earlier agreement, and sets out a number of new terms. So far as those new terms confer any rights on the Tote, they relate solely to sponsorship, advertising and the like. Similarly, the rights granted to the Tote by the 2009 Agreement related only to matters such as branding, advertising, marketing, media and sponsorship. The only reference to on-course betting activities was a provision under which Arena promised to ensure that alternative betting facilities and/or providers would be restricted to operating in existing designated betting areas. Clause 15.1 of the Tote Agreement provides that it constitutes the entire agreement between the parties.

There is manifestly no express prohibition in the Tote Agreement on the Tote’s use of data acquired on the Arena Racecourses. Given that there is nothing in it which purports to regulate in any way the Tote’s activities on the Arena Racecourses there is also no basis for implying such a prohibition into it. Although an implied term was pleaded by the claimants, this was not pursued in closing argument.

Accordingly, while the claimants did not abandon their claim that the Tote’s conduct involved a breach of the Tote Agreement, attention focused in closing on the Arena Terms.

The Arena Terms

Notices posted at the entrance to Arena racecourses stated that: “Entry to this racecourse and grounds … is granted subject at all times to the Arena Racing Company Terms and Conditions of entry”, and identified where those terms could be found.

The notices stated that special conditions may apply if “you are attending the Racecourse as a member of staff … or otherwise than in your personal capacity as a spectator.”

The notices drew attention specifically to two provisions, prohibiting attendees from conducting any commercial activities, including distributing data relating to races, and from using any devices for, among other things, transmitting any data relating to a race, unless they were expressly permitted to do so.

I was shown the terms and conditions applicable to Wolverhampton racecourse, it being common ground that materially the same terms apply at the other Arena racecourses. The preamble states:

“These Conditions apply to all persons entering the Racecourse. If you attend the Racecourse in any capacity other than as a private individual, you must also agree to a set of Special Conditions which will apply in addition to the Conditions.”

The conditions include the following:

“4.6

No Ticket gives any Attendee any right to receive, use or exploit any racing data, audio, visual or audio-visual coverage of any race or fixture. All such rights are reserved. Any coverage or data made available at the Racecourse is made available for the personal use and consumption of Attendees only and redistribution to any person outside the Racecourse is strictly prohibited.”

“10.1

You shall not use any device or technology to capture, record, store, transmit or broadcast any data relating to any race, fixture or other race-related activity unless you are expressly permitted to do so by Condition 10.4.” (Clause 10.4 permitted mobile devices for personal use.)

“10.2

In particular, you must not capture, record, store, transmit (including live stream): … (d) any data relating to any race … e.g., odds, Going, colours and colour changes, jockey changes, results…” (Clause 10.3 stated that the list in 10.2 was indicative of the types of activity prohibited, but not an exhaustive list.)

“10.6

Unless specifically permitted by these Conditions, you shall not use mobile telephones or any other communications device while on the Racecourse to communicate with anyone outside the Racecourse for the purpose of or in connection with any betting.”

Additional terms and conditions are provided for a variety of categories of commercial third parties. These are: (A) participants in racing; (B) media contractors and licensees; (C) photographers; (D) press; (E) on-course bookmakers; and (F) contractors, trades and exhibitors. The only one of these that the claimants contend is relevant is (B). At the start of the special terms and conditions, the following appears:

“Attendees will only be required to comply with, and will only be entitled to any of the rights and privileges set out in, the following Special Conditions where specifically and expressly admitted to the Racecourse on the basis of one or more of the categories of Special Condition.”

Special Conditions B apply to “Media Attendees”, being those attendees who are present under the terms of a “Third Party Media Agreement” or “for purpose(s) equal or similar to those permitted by a Third Party Media Agreement”. A Third Party Media Agreement is defined as an agreement with a third party under which they, their sub-contractors and licensees were permitted to enter the Racecourse “for the purpose of producing still images or audio/visual/data coverage of races and fixtures and/or transmitting, broadcasting or otherwise distributing [the same]”.

By clause 24.3 of Special Conditions B, save where otherwise agreed expressly in writing:

“…Media Attendees will not be entitled to exercise any rights to, or to produce and exploit, audio/visual/data coverage of races and fixtures on their own behalf or for any purpose other than that permitted under the Third Party Media Agreement.”

The first question arising is whether the special terms and conditions relating to Media Attendees applied to the Tote. There is no evidence of any agreement having been entered into between the Tote and Arena (or any predecessor of Arena in relation to the Arena Racecourses), whether in writing or otherwise, which regulated the activities the Tote in fact (as is common ground) carried out, over many decades, in respect of pool betting on the Arena Racecourses. The Tote’s presence on racecourses for the purposes of pool betting was originally pursuant to a statutory right, and there is no evidence of its presence on the Arena Racecourses having been made the subject of any agreement (other than the Tote Agreement, dealt with above). There is accordingly no evidence of there being any agreement that might constitute a Third Party Media Agreement within the meaning of Special Conditions B.

It remains possible that the Tote’s presence on the Arena Racecourses was for purposes “similar to those permitted by a Third Party Media Agreement”. Even if so, however, there is no evidence that the Tote was “specifically and expressly admitted” to any of the Arena Racecourses “on the basis of” the Special Conditions B. Unless it did, then the preamble to the Special Conditions makes it clear that the Special Conditions do not apply to it.

Accordingly, I reject the contention that the Tote was contractually bound to comply with Special Conditions B in the Arena Terms.

So far as the general conditions in the Arena Terms are concerned, those cannot have been intended by Arena to apply to the Tote. The provisions I have set out above (from clauses 4.6 and 10) are in absolute terms. They would preclude the Tote from carrying out any activity at all, yet it is common ground that the Tote was carrying on its pool betting activities at the Arena Racecourses lawfully. I reject the claimants’ contention that it is necessary to read clause 4.6 of the general conditions subject to an implied term “save insofar as expressly authorised”. Arena recognised that it was necessary to set out in express terms (i.e. the Special Conditions) the extent to which a third party would be entitled to make use of information acquired at the racecourse. It is possible to imply a term (among other requirements) only if it can be said with certainty what the term would have been. That is not possible given the inherent complexity in seeking to identify precisely what use the Tote could make of information acquired on course and to which of the restrictions in the general conditions it should remain bound.

Accordingly, in my judgment, there was no contractual restriction on the Tote feeding Raceday Data to SIS, whether that data was contained in the data collected by the Tote for pool betting purposes, or otherwise.

The law

The parties are agreed that the requirements for a free-standing claim in breach of confidence are those set out by Megarry J in Coco v AN Clark (Engineers) Ltd [1969] RPC 41, 47:

As to the second requirement, SIS accepts that it involves an objective assessment. It requires that the recipient of the information knew or ought to have known (viewed by reference to a reasonable person standing in their position) of the confidentiality attaching to the information: see Matalia v Warwickshire County Council [2017] ECC 25, per David Richards LJ at [46], citing with approval the decision of Arnold J in

Where, as here, the original acquirer of the information was the Tote, and SIS was a third party recipient of the information, it is common ground that a claim may lie against SIS if it knew or ought to have known of the confidentiality attaching to the information. If so, then SIS would be directly liable to the claimants for unauthorised use of the information, and not merely liable as an accessory to any breach of confidence by the Tote.

The claimants rely on the decision of the House of Lords in Douglas v Hello! Ltd (No.3) (heard together with OBG v Allen) [2008] 1 AC 1. In that case OK! magazine

Although the Douglases themselves maintained a claim for invasion of privacy, the claim against Hello! was based on breach of confidence. Lord Hoffmann emphasised the importance of the fact that the information had a commercial value, which the Douglases had exploited by their agreement with OK!, saying (at [117]):

“The point of which one should never lose sight is that OK! had paid $1m for the benefit of the obligation of confidence imposed upon all those present at the wedding in respect of any photographs of the wedding. That was quite clear. Unless there is some conceptual policy reason why they should not have the benefit of that obligation, I cannot see why they were not entitled to enforce it. And in my opinion there are no such reasons. Provided that one keep’s one’s eye firmly on the money and why it was paid, the case is, as Lindsay J held, quite straightforward.”

The information in question in the Hello! case did not arise in a commercial context but related to the Douglases’ personal lives. That was, however, irrelevant to the claim by OK!, as Lord Hoffmann explained at [118]: “What matters is that the Douglases, by the way they arranged their wedding, were in a position to impose an obligation of confidence. They were in control of the information.” He saw no conceptual difficulty in the fact that the obligation of confidence was imposed only in respect of a particular form of information, namely the photographic images, commenting (at [119]):

“If OK! was willing to pay for the right to be the only source of that particular form of information and did not mind that others were free to communicate other forms of information about the wedding, then I think the Douglases should be able to impose a suitably limited obligation of confidence”.

Lord Nicholls of Birkenhead and Lord Walker of Gestingthorpe dissented partially on the grounds that the information in the photographic images was not intended to be kept secret, but was intended to be (and was) published to the world by OK! at about the same time that the rival pictures were published in Hello!. They referred to A.G. v Guardian Newspapers (No.2) [1990] 1 AC 109, 282C-D, per Lord Goff of Chieveley for the proposition that “…once [information] has entered the public domain (which means no more than that the information in question is so generally accessible that, in all the circumstances, it cannot be regarded as confidential) then, as a general rule, the principle of confidentiality can have no application to it.”

That did not trouble Lord Hoffmann however, who said (at [120] and [122]):

“But I see no reason why there should not be an obligation of confidence for the purpose of enabling someone to be the only source of publication if that is something worth paying for.”

At [122] Lord Hoffmann addressed the problem that once information is in the public domain then it can no longer be the subject of confidence, saying:

“…it is certainly the case that once information gets into the public domain, it can no longer be the subject of confidence. Whatever the circumstances in which it was obtained, there is no point in the law providing protection. But whether this is the case or not depends on the nature of the information. Whether there is still a point in enforcing the obligation of confidence depends on the facts. If the purpose of publishing the pictures was simply to convey the information that the Douglases had married, the bride wore a wedding dress and so forth, then the publication of any photographs would have put that information in the public domain. So would a description of the event. In this case, however, the point of the transaction was that each picture would be treated as a separate piece of information which OK! would have the exclusive right to publish. The pictures published by OK! were put into the public domain and it would have had to rely on the law of copyright, not the law of confidence, to prevent their reproduction. But no other pictures were in the public domain and they did not enter the public domain merely because they resembled other pictures which had.”

Does the Raceday Data have the necessary quality of confidence?

I have set out in sections B and C above the circumstances relating to the Raceday Data. The key features relevant to this first requirement for a claim in breach of confidence are as follows.

The racecourses are the private property of Arena. While a horserace is a public event, in the sense that members of the public are invited to attend, Arena is in a position by reason of its ownership of the property to impose restrictions on what members of the public who attend the races can do with the information they see. It is (in the words of Lord Hoffmann in the Hello! case) “able to control the information”. This it did through the Arena Terms, which prohibited anyone who attended the race from transmitting off-course any information relating to “any race, fixture or other race-related activity”, and from communicating with anyone outside the racecourse “for the purpose of or in connection with any betting.”

The information was commercially valuable. Arena exploited that value by selling the exclusive right to collect and distribute the information for the purposes of fixedodds betting to TRP. Although the relevant amounts are not in evidence, given that this trial is not concerned with any issues of quantum, it is common ground that they are commercially significant.

The restrictions imposed on the use of the information on anyone else attending the racecourse is an essential adjunct to the ability to exploit the information for commercial value through the grant of exclusive rights to TRP.

So far, the parallels with the Hello! case are strong. A further parallel is that the information was intended, in both cases, for publication. The timescale is different here, in that information relating to events on-course is put into the public domain almost instantaneously, including by way of TV coverage. In my judgment, that does

The commercial value which Arena sought both to protect and to exploit lies in the fact that the information is wanted by off-course bookmakers as soon as possible upon it coming into existence. Much of it ceases to have any value once the race has started, save for the result, which ceases to have value shortly after the end of the race. To take a specific example, the information that a specific horse has refused to enter the starting gate is something which bookmakers need to know as soon as possible so that they can cease taking bets on that horse.

SIS was fully aware of the time-critical nature of the information, as Mr Smith acknowledged in an email to Mr Siers on 17 August 2016 in which he said: “The data necessary to take and settle bets has to get into the public domain at some point at which point anyone can collect it and use it but it is not clear whether we (or anyone else) can source the data quickly enough to be useful to bookmakers.”

Thus the claim for breach of confidence is focused solely on the short period in which (1) TRP required exclusivity and (2) SIS sought to circumvent that exclusivity. The undoubted commercial value of the information during this period, notwithstanding that it was collected for the purpose of public dissemination, is demonstrated by the fact that both TRP and SIS were willing to pay substantial sums for their respective rights to collect and disseminate it.

For this same reason, the issue dealt with by Lord Hoffmann at [122] of Hello! (quoted above) does not arise here. The issue there addressed is whether the confidentiality survived the publication of the Hello! pictures. While Lord Hoffmann held that it did, he noted that if the purpose of publishing the pictures was to convey underlying information about the wedding, then the publication would put an end to confidence. In this case, publication is indeed for the purpose of conveying underlying information, which means that confidentiality is lost upon publication. Since, however, the claimants only rely on an obligation of confidence that exists for the short time sufficient to enable TRP to enjoy its exclusive right of publication, this is irrelevant.

The fact that the races are broadcast live on television does not undermine this conclusion. It is common ground that there is substantial value in the data element of the live feed from racetracks in addition to (or, in the case of the SIS feed, instead of) live pictures. That is partly because (as noted above) the data is time critical to the period before a race starts, so that televising the race is, by definition, too late. Even if the television pictures capture certain events before the start of the race, it is highly unlikely that they would capture in sufficient detail the elements which go into the Raceday Data, and certainly not packaged in a form which would provide bookmakers with the requisite information fast enough.

SIS contends that because the information comprising Raceday Data was visible to everyone (potentially thousands) on the racecourse, it did not have the necessary quality of confidence. Moreover, because other racecourses (in particular the Independent Racecourses) did not contain similar restrictions to those set out in the Arena Terms, it is impossible to conclude that the information is subject to any “general obligation of confidence”. That, however, misses the point. Raceday Data is not confidential because it is of an inherently confidential nature, but because there is a substantial commercial value in the information, provided it is disseminated to offcourse bookmakers as soon as possible, and Arena has the ability to, and does in fact, control its dissemination via exclusive channels so as to exploit that value. As in the Hello! case, the fact that TRP (and, before it, SIS) was willing to pay substantial sums for the benefit of an exclusive right to collect and disseminate the information to offcourse bookmakers engaged in fixed-odds betting is a strong indication of its confidential nature.

Leaving aside for the moment, therefore, the question whether the Tote was itself bound by any obligations of confidence, in my judgment there was commercial confidentiality in the Raceday Data. For any other attendee at the Arena Racecourses, it is unlikely that Arena or TRP would need to resort to the equitable principle of confidence, because any dissemination of Raceday Data by those other attendees would almost certainly amount to a breach of contract. The need only arises in the present case because of the unique position of the Tote which was undoubtedly free to collect Raceday Data and disseminate it for pool betting purposes and which (as I have found) was under no contractual restrictions on the use it could make of information it acquired on course. This falls to be decided in relation to the second requirement, which addresses the circumstances in which the Tote acquired the information.

Was the information imparted to the Tote in circumstances importing an obligation of confidence?

SIS contends that this second requirement is not satisfied because: (1) the information was not provided to the Tote by the claimants, but collected by the Tote’s employees who were lawfully on the Arena Racecourses for that purpose; (2) the Tote could reasonably conclude that any value residing in the package of data that it assembled belonged to it, and that it was entitled to sell it as it wished; and (3) Neither Arena nor TRP took any steps (on the assumption that there is no contractual restriction on the Tote) to prevent the Tote from using the information it acquired.

As I have already noted, the issue is determined by asking whether a reasonable person, in the position of the Tote, would have appreciated that the Raceday Data was confidential.

There are three preliminary points. First, the absence of a contractual restriction on the Tote is not determinative of the question. Indeed, the equitable obligation of confidence is of practical relevance precisely because there is no contractual restriction. Second, I do not think it makes any difference that the Tote is allowed on to the Arena Racecourses in order to collect the information, as opposed to the information being provided to it by Arena. In either case it is information over which Arena asserts control and decides to whom it should be made available. Third, the fact that the Tote was lawfully permitted to collect Raceday Data and use it for one purpose (pool betting) does not in itself mean that the information cannot have been acquired by it in circumstances importing an obligation of confidence. It is perfectly possible that where A provides information to B knowing and intending that it be used

No evidence was led from the Tote by either party. I have no direct evidence, therefore, as to what it actually knew. Evidence of the Tote’s belief as to its ability lawfully to pass on Raceday Data to SIS is to be found in what Mr Siers (who represented the Tote in negotiations with SIS) said at the time. On numerous occasions between September and December 2016 he gave assurances to SIS that the Tote was allowed to sell data it collected on Arena Racecourses to SIS. This is reflected in the warranties provided in the Tote HoT.

It was not suggested that Mr Siers gave assurances to SIS, or included the warranties in the Tote HoT, knowing that they were false. In the absence of any evidence to suggest to the contrary, and not having heard from Mr Siers or anyone else on behalf of the Tote, I find that those assurances and warranties reflected Mr Siers’ state of mind, and that accordingly Mr Siers (and thus the Tote) honestly believed that there was no restriction imposed on the Tote’s ability to provide Raceday Data to SIS.

That, however, is not an end to the enquiry, given that the question is an objective one. For this purpose it is first necessary to determine what facts were known to the Tote, and then to consider what a reasonable person in the Tote’s position would have appreciated in light of those facts.

Given that the Tote employees were regular attendees at Arena Racecourses, I infer that they knew of the Arena Terms. It is inconceivable that they were unaware of the notices posted at the racecourse entrances. The Tote knew, therefore, that Arena sought to impose obligations of confidence on anyone attending the Racecourses.

The documents disclosed by SIS in this case clearly demonstrate that the Tote was aware that Arena sought to exploit the commercial value in the information generated on-course by granting TRP the exclusive right to collect and sell it for the purposes of fixed-odds betting. For example, in an email from Mr Smith of SIS to Mr Siers, in his capacity as Chief Commercial Officer of Betfred, dated 17 August 2016, informing Mr Siers that as a result of steps taken by Arena/TRP as from 1 January 2017 SIS did not anticipate being able to provide Raceday Data or the prices of FRB bookmakers for Arena Racecourses from 1 January 2017, Mr Smith said:

“…we do know that … both issues (data delivery for TRP and the FRB situation) demonstrate clearly that ARC and TRP are very determined to control all the data for their events and we can expect them to make it very difficult for companies like SIS to provide data only products on their races.”

Mr Siers was also sent a copy of TRP’s press release dated 29 September 2016 announcing its right to provide “exclusive coverage” to licensed betting offices at Arena Racecourses, and stating that no other organisation will have access to those courses to collect prices “or other race day data needed to provide a complete data service to the betting industry”.

In any event, given the Tote’s long-standing involvement with British horseracing, I find it inconceivable that it was not aware of the commercial importance of Raceday Data for the purposes of fixed-odds betting and the steps that Arena had taken to protect its ability to exploit that information through an exclusive arrangement with SIS for some years prior to 2017.

In addition, there can be no doubt that the Tote knew that the only activities it had historically ever carried out on Arena Racecourses were those related to pool betting (save only where it had entered into specific agreements to enable it to carry on additional activities).

In summary, a reasonable person in the Tote’s position would have known of the steps taken by Arena and TRP in order to preserve confidentiality in Raceday Data, to

In these circumstances, I find that a reasonable person in the Tote’s position, appreciating that the information was acquired by it for the purposes of pool betting only, would also have appreciated that it was acquired in circumstances importing an obligation of confidence, so that it could not be used (whether by the Tote or by third parties to whom the Tote sold the information) for the purposes of fixed-odds betting.

Did SIS know (or should it have known) of the confidentiality attaching to the information?

In considering what a reasonable person in the position of SIS would have understood, as to the confidentiality attaching to Raceday Data, it is first necessary to establish what SIS actually knew.

The key individual, in this respect, is Gary Smith, its chief executive officer and the person with responsibility for the decision to enter into the Tote HoT. SIS did not call Mr Smith to give evidence. Although Mr Smith had left SIS’s employment in 2017 there was no suggestion that he could not have provided a statement or been called. The claimants invite me to draw, where necessary, adverse inferences against SIS from the absence of evidence from Mr Smith. I accept that if there is prima facie evidence as to Mr Smith’s state of mind adverse to SIS’s case, then SIS’s failure to call him would be an important factor in determining whether he had that state of mind. It is necessary in the first place, however, to focus on what was said or written at the time as recorded in the contemporaneous documents.

The starting point is a racing strategy paper prepared by SIS in July 2016. This was prepared against the background that SIS was to lose its rights to transmit data from Arena Racecourses. The paper stated the “clear recommendation” that SIS should look to remain in the business of providing horseracing data. In order to do so, it identified potential partners as a source for alternative horse racing content, one of which was the Tote.

The paper acknowledged that SIS was not fully “rights compliant” in the way that it collated certain data. This was a reference to the practice of “scraping” data from official sources without the permission of the rights owner. SIS had been doing this for eight years by obtaining information from the “Turf TV feed” (television coverage from certain racecourses, none of which were the claimants’ racecourses). After considering alternatives, the paper concluded that “SIS should move swiftly to become completely compliant with existing rights agreements.”

Under the heading “Totepool”, it was stated that SIS currently produced the data that triggered the Tote’s raceday data feed, and that it was feasible that this might be reverse-engineered so that SIS utilised a data feed driven by the Tote’s on-course representatives. It noted that “further discussion is required to ascertain whether or not Totepool data can be used for this purpose while SIS’s commitment to compliance is fully satisfied. Totepool is permitted to collect data on course, although there may be a risk that Totepool’s contract with the UK’s racecourses specifically prohibits the use of that data to drive fixed-odds betting. It is understood that this risk is not considerable; but further conversations with Totepool would be required to be reassured on this point.”

This paper demonstrates five things. First, SIS appreciated the risk that the Tote could not lawfully supply SIS with Raceday Data for use in fixed-odds betting. Second, for some years SIS had indulged in unlawful behaviour involving ‘scraping’ data from TV feeds. Third, that it determined for the future that it would be fully rights compliant. Fourth, in line with that determination, it recognised the need to seek assurances from the Tote that it was legally entitled to supply Raceday Data to SIS. Fifth, at this point those within SIS who produced this document (which included Ms Coulson) clearly thought that the source of the Tote’s rights (and thus the potential source of limitation on those rights) was a contract or contracts with the owners of UK racecourses.

It was the evidence at trial of Ms Coulson and Mr Witten that thereafter SIS sought, and obtained, assurances from the Tote that it had the right to collect Raceday Data from the Arena Racecourses and to pass it on to SIS. Moreover, it was their evidence that SIS believed this to be true.

This evidence is corroborated by contemporaneous documents. For example, a note of an internal SIS meeting on 6 September 2016 records that Ms Coulson and Mr Dolan of SIS had met with Mr Siers that morning, and that he had confirmed that the

SIS’s state of knowledge at the time is revealed by internal documents in which the planned arrangement with the Tote is referred to, and in which SIS is described as being “compliant” or “fully compliant” (no doubt referring back to the determination reached in the earlier strategy paper). For example, in a note of a meeting held between Ms Coulson and Mr Brent Dolan and Mr Siers on 23 August 2016, in order to investigate the possibility that SIS might use on course data generated by the Tote to facilitate an alternative supply of race day data to customers, it is recorded that “It is apparent that if an arrangement can be made with [the Tote], SIS will be able to offer an alternative – but fully compliant – data feed to customers. This is not black or grey data. Strategically, such a move would ensure that SIS remain in the field of UK horse racing data.” Also, in the note of meeting of 6 September 2016 referred to above, it was stated that the arrangement with the Tote would allow SIS to continue in the data collection space with regard to all UK horseracing “while being fully rights compliant”. Finally, in a document dated 20 October 2016 headed “Totepool – an opportunity for SIS to acquire an alternative, compliant source of Race-Day Data” (under Mr Booth’s name but commented on by Ms Coulson) the intention to enter into a deal with the Tote was set out and referred to as a “fully rights compliant offering”. The document (as amended by Ms Coulson) stated that “Totepool is allowed both a) to be on course to collect race day data and b) to sublicense that data for resale to a partner such as SIS.”

I find that the expressions of belief found in these documents are an accurate reflection of the state of mind, at the time, of those within SIS with responsibility for them. There is no reason to think that the writers of these internal documents were incentivised to be anything other than truthful.

It is clear that by this stage Ms Coulson, at least, still understood that the basis of the Tote’s rights was contractual. By 29 November 2016, however, Mr Smith knew that was not so. Following a discussion he had with Mr Siers and Mr Scanlon of the Tote on that date, Mr Smith relayed to his colleagues (including Mr Witten, Ms Coulson and Mr Ames) the contents of his conversation about the provision of Tote Raceday Data for Arena courses, saying:

“I know that Paul and Rachel have already been researching this and have already had discussions with Phil Siers. I have been rather pessimistic about the opportunity so far because of the perceived legal risk. After talking it through with Joe and Phil today I am somewhat more comfortable because they are offering us Tote data (“Tote Off”; “Tote results”; “Tote NonRunners”) and are not offering official race-day data. Apparently the Tote versions are usually very similar or the same as the official data but there are some minor differences … They are also taking into account there are no detailed contracts in place between racecourses and the Tote, probably because of its roots as a government franchise. Please can we schedule some time to discuss this to decide whether to proceed?”

This email demonstrates that Mr Smith (like his SIS colleagues) had initially had concerns, that he recognised enquiries should be made in relation to them, and that having made those enquiries of Mr Siers and Mr Scanlon he was sufficiently comfortable to proceed with the transaction.

The assurances that had been received from Mr Siers were formalised in the Tote HoT, negotiated and signed at the end of December 2016. Mr Ames, who had rejoined SIS as interim Product Director at the end of November 2016, was deputed to lead the negotiations with the Tote over the Tote HoT. SIS’s legal department were involved in the drafting of the Tote HoT.

By clause 3.2 of the Tote HoT the Tote warranted that:-

“(a)

it has (and so far as it is reasonable aware, will continue to have during the Term) save to the extent that it has notified SIS that it no longer has any of the underlying rights) all necessary rights from any and all third parties and the power and full authority to grant the Rights to SIS in relation to all Events from the Tracks and for SIS to lawfully receive, access and use the data contained in the Race Day Data Feed as permitted under these HOT.

(b)

use of the Race Day Data Feed by SIS in connection with the performance of each party’s obligations and/or exercise of their rights under these HOT will not breach any third party rights, including any Intellectual Property Rights (as defined in clause 3.5 below) of any third party.”

By clause 3.3, the Tote granted an indemnity to SIS for any loss arising out of a breach of the warranties in clause 3.2. By clause 3.4 the indemnity was limited to the sums actually paid by SIS under the Tote HoT.

In light of the above evidence, the most significant hurdle facing SIS, in relation to the second of the Coco v A.N. Clark requirements, is that it knew essentially the same information that was known to the Tote, on the basis of which I have concluded that a reasonable person in the position of the Tote would have appreciated that the information was acquired by them in circumstances which imported an obligation of confidence.

Thus SIS knew: (1) that Arena imposed restrictions on the use of Raceday Data upon all attendees; (2) that Arena had granted exclusive rights to exploit Raceday Data for fixed-odds betting purposes to TRP; (3) that there was considerable commercial value in being able to disseminate Raceday Data to off-course bookmakers as soon as possible, and that the exclusivity given to TRP was in order that TRP and Arena could exploit that value; (4) that the Tote had no contractual arrangement with Arena regulating its entitlement to collect, or sub-licence to others, Raceday Data; and (5) while the Tote had the right to be on the Arena Racecourses, that had only ever in fact been exercised for the purposes of pool betting.

In addition SIS was also aware that its own agreement with Arena granted it the exclusive right to collect and distribute Raceday Data. I infer that any reasonable person in SIS’s position would appreciate that the rights which TRP had been granted were similarly exclusive, even if SIS had not appreciated the significance of the press release from TRP in September 2016 which made express reference to that fact.

In these circumstances, I consider that the crucial factor, in determining what a reasonable person in SIS’s position (as opposed to the Tote’s position) would have understood, is that SIS received assurances from the Tote that it had the right to provide Raceday Data to SIS: would a reasonable person in the position of SIS, having received these assurances, have understood that the Tote was free to use (and pass on) Raceday Data for purposes unconnected with pool betting?

In my judgment, the answer to this question is no. It is important to pay close consideration to the details of the assurances given. As I have already explained, Mr Smith, as of 29 November 2016, knew that the Tote did not claim a contractual basis for its rights to use the information. Had the basis of the Tote’s rights been contractual, then it would be difficult to argue that SIS’s reliance on an assurance by the Tote that it could lawfully provide the Raceday Data was unreasonable. Instead, Mr Smith knew that the Tote was relying solely on its “roots as a government franchise”. I conclude that a reasonable person, knowing the matters I have identified above in paragraph 184, would have appreciated that it was highly unlikely that the Tote’s roots as a government franchise, for the purpose of pool betting, would have enabled it to circumvent the commercial confidentiality which Arena imposed on Raceday Data so that it could exploit the value in it in connection with fixed-odds betting. With such an appreciation, a reasonable person would have sought assurances from Arena itself before proceeding to take Raceday Data from the Tote.

Accordingly, I conclude that a reasonable person in the position of SIS would have appreciated that the Tote acquired the information in circumstances imposing obligations of confidence, which would be breached by use of that information otherwise than by the Tote for the limited purpose of pool betting. The second of the Coco v A.N. Clark requirements is therefore satisfied.

The claimants go further and contend that, notwithstanding the assurances sought and obtained from the Tote, SIS was nevertheless aware of, or turned a blind eye to, the fact that the provision of Raceday Data to it by the Tote was unlawful. This is irrelevant to the direct claim in breach of confidence, but might be relevant to the claim in conspiracy. I deal with it here for convenience.

First, the claimants rely on inconsistencies in the evidence of SIS’s witnesses as to their understanding of the source of the Tote’s rights. It is true that there was some confusion among the witnesses as to whether the Tote’s rights were contractual (as

Second, the claimants contend that the following email exchange between Ms Coulson and Mr Dolan on 31 October 2016 demonstrates that Ms Coulson, at least, lacked an honest belief in the lawfulness of the Tote’s actions. Ms Coulson emailed Mr Dolan, saying “Do you think Phil [Siers] will let us have a redacted version of his contract which allows him to sub licence for any purpose?” Mr Dolan replied “I can ask but he may not wish to provide it but I know we are okay… The only issue to cover for us is that he carries the can!!”. It is suggested that an honest person would have pressed for sight of the contract. I reject this contention for two reasons. First, even if there was anything suspicious in Ms Coulson not pressing for sight of the contract as at 31 October 2016, this became irrelevant a month later when it was apparent that there was no such contract. Second, and in any event, I do not regard it as suspicious that SIS did not press for sight of a contract or contracts it then believed the Tote had. Mr Witten described it as normal commercial due diligence to ensure that the person purporting to sell rights has the ability to do so, but it would be unusual to ask to see source documents of the seller and that he did not think to do so. Ms Coulson agreed that it was unusual in her experience for third parties to provide copies of their contracts. Mr Ames (who was brought into the negotiating process only late in the day) said that he had not asked to see a copy of any contract with the Tote and that it never occurred to him that he would need to do so. (It is true that, at one point in his evidence he said that the question of seeing a copy of the Tote’s contract did come up in conversation with Mr Siers and it was clear that Mr Siers was not going to provide it. He soon corrected himself, however, to say that he had meant to refer to the Tote’s legal advice, not its contract. I prefer his corrected evidence, which is inherently more likely given that Mr Ames, before he became involved in negotiations with Mr Siers, had been one of the recipients of Mr Smith’s email saying that the Tote had no contracts with the racecourses.)

Third, the claimants contend that SIS’s behaviour in seeking an indemnity from the Tote indicates that it was at least subjectively reckless as to whether the Tote’s actions were unlawful. The argument is that SIS asked for an indemnity because it knew (or turned a blind eye to the fact that) the Tote could not lawfully provide the information to SIS, so it wanted an unlimited recourse against the Tote. This is demonstrated, they say, by Mr Dolan’s response to Ms Coulson’s suggestion that Mr Siers be asked for a copy of the Tote’s contract with racecourse. That response was: “…he may not wish to provide it … the only issue for us is that he carries the can…” I do not accept that SIS’s conduct in seeking a warranty, backed by an unlimited indemnity, from the Tote is indicative of turning a blind eye to the fact that the Tote did not have the lawful right to provide SIS with Raceday Data. It is a normal part of the allocation of risk between contracting parties to include such warranties and indemnities. The mere fact that SIS recognised such risk, and provided for it in this way, does not equate to turning a blind eye to whether the Tote was acting unlawfully.

The fact that SIS was prepared to go ahead without an unlimited indemnity in any event points against SIS being content to proceed only on the basis that the Tote “carried the can” for unlawful behaviour.

Fourth, the claimants point to email exchanges between Mr Ames and Mr Smith on 29 December 2016 in which they express nervousness or suspicion about Mr Siers. In an email to Mr Smith and others (copied to Mr Andrew Graham, head of legal at SIS), Mr Ames reported that the Tote HoT was with Mr Siers, who had said he could not talk today because he was too busy. Mr Ames noted that this might be “classic Phil in which case we will have a frustrating day tomorrow starting at 10:00 – phone call in the diary – going back-and-forth on detail and commercials”, or, worse, it might be that Mr Siers was working on a TRP contract “in which case the discussion will be more difficult and complicated as he does not have a desperate need to sign?!?!?!?” Mr Smith responded: “… I am always nervous dealing with Phil Siers – your scenario two is a very real possibility. Do we have anyone else involved in the tote data feed arrangements from bet Fred? Joe Scanlon for example?” Mr Ames replied saying he was nervous of getting involved in internal politics but it was already clear that the two teams within Betfred and the Tote were not talking to each other. In a subsequent reply of the same date he said “hopefully just my suspicious mind doing cartwheels.”

The claimants link this to a finding in unrelated litigation in 2012 that Mr Siers was “unconvincing” as a witness, and contend that it is not “consistent with honest business practices to take such an important, potentially controversial, and inherently unlikely fact” on trust from someone SIS knew to be untrustworthy. I have quoted from the emails in full, because the full context shows that Mr Ames and Mr Smith were referring to the possibility that the Tote might, at a very late stage in negotiations, be doing a deal behind SIS with TRP, which would have left SIS (and LC) with no alternative data feed form Arena racetracks in January. Neither the fact that they may have feared that Mr Siers was dealing with TRP behind their back, nor their awareness (assuming they were aware) of judicial comments as to Mr Siers’ reliability as a witness, provides any basis in my judgment for concluding that SIS did not in fact trust his assurances that the Tote had the ability to provide Raceday Data to SIS, particularly when those assurances were enshrined in a formal document.

Fifth, the claimants contend that anybody with horseracing knowledge would have known that the Tote could not sublicence the information. They rely on the fact that the Tote had never sought to do so and that SIS had, until December 2016, the exclusive right to distribute Raceday Data from Arena Racecourses. It would – if it seriously thought that its rights were vulnerable to being undermined by the Tote (particularly given they viewed Mr Siers as a “canny and aggressive” businessman) – have taken steps to close off that possibility. Moreover, SIS knew (at least from the time of TRP’s press release in September 2016) that TRP believed that it had exclusive rights. The claimants also rely on the fact that SIS’s internal documents from the outset show that it had a concern that the Tote was not in a position lawfully to supply the data, and that this was a potentially “controversial” matter. In my judgment, these points all go to confirm something which is evident from the documents I have already referred to, namely that SIS did indeed have concerns that the Tote was unable to provide Raceday Data to it. That, however, is only the starting point. If SIS, having realised at the outset that there was risk that the Tote could not lawfully provide Raceday Data to it, had thereafter done nothing to satisfy itself on the point, then it would be difficult to avoid the conclusion that it had, at least, blindeye knowledge of the breach of confidence. But, on the contrary, SIS made enquiries of the Tote and received reassurances, repeated in a formal legal contract, that the Tote was lawfully entitled to provide the information.

Sixth, the claimants place particular reliance on Mr Smith’s comments on 30 December 2016 relating to concerns as to messaging clients and the risks of litigation. Mr Ames emailed his SIS colleagues identifying an “interesting tactical question” concerning whether to inform the market if they reached an agreement with the Tote. On the one hand, if they did not announce the deal, he envisaged that if TRP objected when they saw SIS providing a data service to bookmakers on 1 January 2017, then they would be able to “push back” by showing they had a legitimate source of data. He commented that this would then “play out behind the scenes with whatever legal or industry politics there might be.” On the other hand if SIS announced the deal, “thereby confirming to our customers and industry parties that they can expect to see legitimate feed and continued service, there seem to be a lot of benefits”. These included that customers would see SIS have cut a deal, TRP, PA and RMG would have two targets to think about protecting their position and “if things were to go wrong, it would be well known that we have bought the feed in good faith and therefore played the game with a ‘straight bat’”.

Mr Smith responded, on the question of publicity, saying that he would only announce a Tote Raceday Data deal if Mr Ames and Mr Graham “feel that we would reduce our legal risk by announcing it. I feel that there is a significant risk of injunctions flying if we announce a tote deal before it takes effect.”

While this evidence clearly demonstrates that SIS was aware of the risk that the Tote was not in a position lawfully to provide Raceday Data from Arena Racecourses, it falls short in my judgment of demonstrating knowledge (including blind-eye knowledge) sufficient to establish liability in conspiracy, when seen in the wider context I have set out above.

It is true that Mr Smith undoubtedly continued to recognise the litigation risk of proceeding, and was keen to find ways to mitigate that risk, but there is in my judgment an important distinction between recklessness as to whether conduct is unlawful, turning a blind-eye to suspicions to the contrary, and believing conduct is lawful but recognising the risk that someone affected by it might take a different view, and even seek to vindicate that view by taking legal proceedings.

Accordingly, I conclude that although a reasonable person in SIS’s position would have appreciated that the Tote acquired the information in circumstances of confidence which precluded its use for fixed-odds betting, it (whether through Mr Smith or others involved in the transaction with the Tote) neither actually knew that, nor turned a blind eye to whether, the Tote was unable lawfully to provide Raceday Data to SIS.

Was there an unauthorised use by the Tote/SIS to the detriment of the rights holder?

The claimants contend that the Tote breached the obligation of confidence in two ways: first, by providing to SIS the information that the Tote was allowed to collect (i.e. that needed for the purposes of pool betting) and, second, by collecting and providing to SIS additional information (which it did not otherwise need, or collect, for the purposes of pool betting).

While SIS accepts that information in the first category was provided to it, it disputes that the Tote provided to it any information over and above that which it collected for pool betting purposes.

Given my conclusion that the obligation of confidence restricted the Tote’s ability to use the information that it was authorised to collect, to use in connection with pool betting, I need not resolve this dispute. If the former constituted a breach of the claimants’ rights of confidentiality, then the latter must have done. In my judgment, both uses of the information were not only unauthorised but were to the detriment of the claimants. Although no evidence was led at trial as to the extent to which TRP or Arena suffered damage by reason of the fact that SIS supplied a rival (unauthorised) data feed to bookmakers, I accept that to interfere with TRP’s exclusive right to distribute Raceday Data in this way was necessarily to the detriment of the claimants.

In case the matter goes further, I nevertheless find that the information provided by the Tote to SIS did go beyond that which it collected for pool betting purposes. My reasons, stated shortly, are as follows.

I accept that some of the bases on which the claimants advanced this part of the case were unfounded. For example:

The most compelling piece of evidence, however, is a document setting out what should be done by the SIS employees who attended the Tote’s premises in Wigan for the purposes of providing information to SIS’s traders that was not within the Tote feed. This was a document which Mr Ames said dated from around January 2017. It is the best evidence of what information was actually provided by the Tote to SIS (as opposed to what was potentially to be provided according to documents created during the negotiations with the Tote during the latter part of 2016). Mr Ames was taken through this document in detail and admitted, in relation to a number of the items referred to in it, that it was information additional to that which the Tote collected for the purposes of pool betting. SIS contends that these admissions are not to be taken at face value, because they are contrary to his earlier evidence to the effect that the “additional” information provided by the Tote was merely that which could not be integrated with the Tote feed for technical reasons. It is true that there is an inconsistency in Mr Ames’ evidence on this point, and (as SIS accepts) he was not particularly close to the detail at the time, but his admission that “the feed [from SIS to bookmakers] required not just the information in appendix 2 but a vast array of other information which is also needed but was not provided as part of the Tote data” was clear, and he was not re-examined on this.

SIS says that Ms Coulson’s evidence (that it was her understanding at the time that the need for information to be telephoned through from Wigan was because of technical issues with the data feed) should be preferred, as she was closer to the detail. Her evidence on this point is, however, undermined by her acceptance that she did not

Various of the items of information that appear in the operations manual are additional both to (i) the items in appendix 2 to the Tote HoT and (ii) the items marked as green on the list provided by Mr Hanlon on 7 October 2016. In light of this, and Mr Ames’ admission, I find that the data provided by the Tote to SIS did extend beyond that which it collected for pool betting purposes.

Conclusion

For the above reasons I conclude that SIS is liable to the claimants for having breached the claimants’ confidence in the Raceday Data supplied to it by the Tote.

In order to access the Betdaq and Betfair websites, with sufficient speed to be able to make use of the prices displayed, it was necessary to sign up to their terms and conditions. Those terms and conditions prohibited any use of the data (in the case of Betfair for commercial purposes) without written consent.

It is common ground that SIS breached those terms and conditions, at least for the first couple of months of 2017. During that period, I find (based principally on Mr Tompkins’ evidence) that SIS was predominantly using the Betfair website. It never sought, or obtained, any permission from Betfair to do so.

It is accepted by SIS that it knew that it was sourcing prices on the Betfair website in breach of contract, and that to the extent that it sourced prices on the Betdaq website until at least March 2017, it knew it was acting in breach of contract.

The claimants contend that SIS’s continued use of Betdaq data after March 2017 was equally in breach of Betdaq’s terms and conditions. SIS contends, on the other hand, that from March 2017 it had Betdaq’s permission to use Betdaq data.

Since the claimants have no independent claim against SIS for breach of contract as against the Exchanges, these breaches of contract are relevant only if they constitute

“unlawful means” for the purposes of the tort of conspiracy. For the reasons I set out below in section G of this judgment, I have concluded that they do not.

If that conclusion (as a matter of law) is wrong, given that SIS accepts that it was knowingly in breach of contract for at least the first two months of 2017, the claimants would have succeeded on this element of the claim. Whether that breach of contract continued for two months, or longer, would make no difference to the finding on liability. However, it might be said (if this were the only basis on which there was a conspiracy to injure by unlawful means) that the measure of damages in that event would depend on the length of time SIS’s use of the Betdaq prices was unlawful. I therefore address that issue in the following paragraphs.

The first question is whether SIS had permission, at all, from Betdaq to use its data. It is necessary to distinguish between two different ways in which its data could be accessed. The first was by signing up to the terms and conditions of its website. The second was by accessing its “application programming interface” (“API”), by using software known as “Geeks Toy”. While signing up to the website allows faster access than merely viewing the public pages on the website, the fastest way to access the data was via the API.

Initially, SIS sought (and obtained) Betdaq’s consent to access its website. In an email dated 21 February 2017 Ms Coulson asked Mr McLaughlin for confirmation that he was “ok with us using Betdaq website for the purposes of creating prices for the Arena tracks.” Mr McLaughlin replied that he believed this was with Richard Lang (of LC) and Mr Witten “to agree commercials”. Ms Coulson, in an email of 28 February 2017, pointed out that the commercial discussions related to access to the

It is common ground that, from about March 2017, SIS obtained the Betdaq prices by accessing its API.

Mr Witten’s evidence was that he reached agreement with Mr Lang of LC (on behalf of Betdaq, which was wholly owned by LC) for access to the API in March 2017. The fact that they were negotiating is evidenced by an exchange of emails on 1 and 2 March 2017. On 2 March Mr Lang proposed a special discount rate for the provision of the Betdaq feed for a 12-month term, and asked Mr Witten to let him know if he was “happy to proceed on these terms.” There is no email in response, but it is clear from the terms of the email that they were holding discussions. Taking into account the fact that, from about this time, SIS paid a licence fee to Betdaq, I find that an agreement was indeed reached permitting SIS to access the Betdaq API.

The next question is whether that permission extended to the use of the Betdaq data for producing Betting Shows to provide to SIS’s customers.

Mr Witten’s evidence is that it had been intended to formalise the agreement he reached with Mr Lang in a written contract, but that this did not materialise until long after the event. It is apparent from an email from Mr McLaughlin of Betdaq to Mr Witten on 22 September 2017 that a draft of a licence agreement had been provided to SIS at some point prior to that. That draft, by clause 2.3(g) prohibited SIS from using the price feed to provide market information to a person other than SIS for business or commercial use. Further, by clause 2.1, SIS was to be granted a licence to access the

Both sides served a witness statement from Mr McLaughlin but he was called by neither party, and I put out of my mind the contents of both of his statements.

It is unclear when SIS had first been sent the draft licence agreement. On the balance of the evidence, however, I find it was originally sent sometime in March 2017. It is likely that LC, having reached a commercial agreement, would have followed that up with a draft contract relatively quickly. It is also consistent with the existence of a draft of the agreement bearing the typed date 4th March 2017.

The draft licence agreement was not signed, however, until long after the event, on 12 March 2018. On that date LC wrote to SIS confirming that they had entered into an agreement in relation to SIS’s use of the Betdaq data with effect from 4 March 2017. The letter stated: “We understand that SIS has been using the Betdaq Data purely as an internal reference tool only, in conjunction with various other sources of information and data, together with the knowledge and skill of SIS’s employees, to support SIS in the provision of services to its customers. We confirm that we consider such use, as described in this letter, to be in accordance with the terms of the agreement.”

Shortly after receipt of Mr McLaughlin’s email of 22 September 2017, SIS ceased to access the Betdaq API and began to provide the official prices. At the same time it commenced negotiations with TRP for an agreement to provide independent bookmaker customers with official TRP prices, although that agreement was not signed until May 2018. In an internal email dated 5 October 2017 Mike Jones, director of software engineering, said “we are no longer allowed to use Betdaq so we are unable to produce any prices for TRP tracks ourselves”.

On the basis of the above evidence, I find that between March and September 2017 SIS’s use of Betdaq data to create Betting Shows for its customers did not constitute a breach of contract. First, there is no evidence that the terms of the draft licence agreement had been agreed before March 2018. SIS’s access to the Betdaq API from March 2017 was on the basis of Mr Lang’s email of 2 March 2017, which simply asked Mr Witten to confirm that SIS was happy to proceed on those terms. Second, I conclude that LC (who negotiated the matter on behalf of Betdaq) was aware that SIS was using, and intended to continue to use, the Betdaq prices in order to create Betting Shows for its customers, and not just for LC. Mr Witten’s evidence is that he made that clear to Mr Lang. This is corroborated by the terms of an email to Mr Lang of 20 February 2017, in which Mr Witten referred to the fact that SIS was looking to have access to Betdaq prices in order to assist in producing Betting Shows, describing this as “a service” that LC was currently taking from SIS. Third, the licence agreement, when signed, must be read in conjunction with the letter of 12 March 2018 from LC, which made it clear that SIS’s use of Betdaq prices, for internal purposes, in order to support SIS in the provision of services to customers was not a breach of the terms of the licence.

The claimants rely on the fact that SIS stopped using Betdaq data soon after Mr McLaughlin’s email of 22 September 2017. In my judgment, however, SIS’s reaction is not inconsistent with the conclusion that its conduct up to that point had been in accordance with the terms of the permission granted by Betdaq. It did not have the benefit of any written agreement, so did not have the security of any term having been agreed, and whatever view may have previously been taken by Betdaq it was clearly now refusing to permit SIS to access the API for the purpose of producing Betting Shows without a signed agreement.

Even if that conclusion is wrong, so that it becomes necessary to consider (for the purposes of the claim in conspiracy) whether SIS knew that (or turned a blind eye to whether) it acted in breach of Betdaq’s terms and conditions, I find on the basis of the evidence set out above that SIS believed that it had permission to access the Betdaq API in the way that it did between March and September 2017. Leaving aside the confirmation received from LC in March 2018 that SIS’s use of the Betdaq API had been in accordance with the agreement, there was an inherent tension, within the terms of that agreement, between clause 2.1 which permitted SIS to access the API for its own “internal business purposes” and clause 2.3(g) which prohibited SIS from using the content to provide market information to a person other than itself for business or commercial use. In circumstances where, to SIS’s knowledge, LC knew that SIS’s “business purposes” in accessing the API were so that SIS could create prices so as to provide a service to its customers – including LC – I infer that SIS believed that its use of the API from March 2017 was consistent with the informal licence it had obtained from LC.

Which of the claims asserted by the claimants constitute

Areas of common ground

SIS did not take issue with the contention that the following claims, if made out, would constitute unlawful means for the purposes of the tort of conspiracy: infringement of copyright and infringement of the sui generis database right; breach of contract by the Tote and breach of confidence by the Tote.

SIS disputes, however, that a breach of the Exchanges’ terms and conditions would constitute unlawful means.

Breach by SIS of the Exchanges’ terms and conditions

This raises for determination the question whether in a claim for unlawful means conspiracy the unlawful means can be constituted by a breach of contract against a third party which is not actionable by the victim of the alleged conspiracy.

The claimants accept that there is no authority bearing directly on this question. They submitted, however, that in light of the decisions of the House of Lords in Revenue and Customs Commissioners v Total Network SL [2008] AC 1174 and the Supreme Court in JSC BTA Bank v Ablyazov [2018] 2 WLR 1125, “unlawful means” encompasses anything that the defendant is not lawfully entitled to do (including, therefore, something which constitutes a breach of contract actionable only by a third party), provided there is a sufficient connection between that act and the aims of the conspiracy.

In Total Network the defendants had participated in a “carousel fraud” thereby depriving HMRC of VAT to which it was entitled. Their actions were held to constitute the common law offence of cheating the public revenue. The question for determination by the House of Lords was whether conduct amounting to a crime could constitute unlawful means, for the purposes of the tort of conspiracy, notwithstanding that the conduct was not actionable at the suit of the claimant as some other tort. The House of Lords determined, unanimously, that it was. The leading speech was delivered by Lord Walker. Given the state of confusion in the recent case law, he went back (at [89]) to general principles to be derived from the older cases in which the economic torts had been developed. He concluded (at [93]) that

“…all the statements of general principle in the classic cases seem to me to be consistent with the proposition that unlawful means, both in the intentional harm tort and in the tort of conspiracy, include both crimes and torts (whether or not they include conduct lower on the scale of blameworthiness) provided that they are indeed the means by which harm is intentionally inflicted on the claimant (rather than being merely incidental to it).” 235. At [94] he concluded:

“In my opinion your Lordships should clarify the law by holding that criminal conduct (at common law or by statute) can constitute unlawful means, provided that it is indeed the means (what Lord Nicholls of Birkenhead in OBG Ltd v Allen

[2008] 1 AC 1, para 159 called “instrumentality”) of intentionally inflicting harm. In Lonrho Ltd v Shell Petroleum Co Ltd (No 2) [1982] AC 173, the sanctions order against Southern Rhodesia was part of the story, but it was not the instrument for the intentional infliction of harm.”

At [95], he said

“Having said that I would accept that the sort of considerations relevant to determining whether a breach of statutory duty is actionable in a civil suit (Cutler v Wandsworth Stadium Ltd [1949] AC 398) may well overlap, or even occasionally coincide with, the issue of unlawful means in the tort of conspiracy. But the range of possible breaches of statutory duty, and the range of possible conspiracies, are both so wide and varied that it would be unwise to attempt to lay down any general rule. What is important, to my mind, is that in the phrase “unlawful means” each word has an important part to play. It is not enough that there is an element of unlawfulness somewhere in the story.

In a concurring speech, Lord Mance (at [116]) agreed that “at least some criminal acts, not amounting to torts, may suffice to ground the tort.” He expanded on this (at [119]) as follows:

“Caution is nonetheless necessary about the scope of the tort of conspiracy by unlawful means. Not every criminal act committed in order to injure can or should give rise to tortious liability to the person injured, even where the element of conspiracy is present. The pizza delivery business which obtains more custom, to the detriment of its competitors, because it instructs its drivers to ignore speed limits and jump red lights (Lord Walker in OBG Ltd v Allan [2008] 1 AC 1, para 266) should not be liable, even if the claim be put as a claim in conspiracy involving its drivers and directors. And - as in relation to the tort of causing loss by unlawful means inflicted on a third party - there is a legitimate objection to making liability “depend upon whether the defendant has done something which is wrongful for reasons which have nothing to do with the damage inflicted on the claimant” per Lord Hoffmann in OBG Ltd v Allan, at para 59.”

On the facts of Total Network, there was no question of the unlawful means having been “merely incidental” (in the words of Lord Walker) or having “nothing to do with” the damage inflicted on HMRC (in the words of Lord Hoffmann, quoted by Lord Mance). The unlawful means deployed were undoubtedly directed at HMRC: that was obvious from the title of the offence (cheating the public revenue).

As Lord Hope put it (at [43]-[44]), having noted that caution was needed where the cause of action is parasitic on the unlawful means used by the defendant against a third party (and reserving his opinion in relation to it):

“But in this case, there was no third party. The means used by the conspirators were directed at the claimants themselves. This is a case where the claimants were persuaded by the unlawful means to act to their own detriment which, in para 61 of OBG, Lord Hoffmann said raises altogether different issues. One has to ask why, in this situation, the law should not provide a remedy … The situation that is contemplated is that of loss caused by an unlawful act directed at the claimants themselves.”

In the Ablyazov case, the claimant in the conspiracy action had also been the claimant in earlier proceedings against A, in which a freezing order had been obtained against

Lords Sumption and Lloyd-Jones (with whom the rest of the court agreed) referred (at [10]), to the “absence of just cause or excuse” as a useful concept to explain what makes the conspiracy actionable as such:

“A person has a right to advance his own interests by lawful means even if the foreseeable consequence is to damage the interests of others. The existence of that right affords a just cause or excuse. Where, on the other hand, he seeks to advance his interests by unlawful means, he has no such right. The position is the same where the means used are lawful but the predominant intention of the defendant was to injure the claimant rather than to further some legitimate interest of his own … In either case, there is no just cause or excuse for the combination”.

They continued (at [11]):

“Conspiracy being a tort of primary liability, the question what constitute unlawful means cannot depend on whether their use would give rise to a different cause of action independent of the conspiracy. The real test is whether there is a just cause or excuse for combining to use unlawful means. That depends on (i) the nature of the unlawfulness and (ii) its relationship with the resultant damage to the claimant. This was the position reached by the House of Lords in [Total Network]”

At [12], they cited Total Network for the proposition that “means were unlawful for the purpose of founding an action in conspiracy, whether they were actionable or not” and went on to refer to the distinction drawn by Lord Walker (and Lord Mance), in the passages I have cited above, between those unlawful means which were purely incidental, and those which were the means (or instrumentality) of inflicting harm upon, or were directed at, the claimant.

As in the Total Network case, the unlawful conduct (contempt of court) in Ablyazov was undoubtedly directed at the claimant, it being the party for whose protection the Orders had been made. Lords Sumption and Lloyd-Jones expressly limited their comments to a case of criminal conduct. At paragraph 15 they said:

“The reasoning in Total Network leaves open the question how far the same considerations apply to non-criminal acts, such as breaches of civil statutory duties, or torts actionable at the suit of third parties, or breaches of contract or fiduciary duty. These are liable to raise more complex problems. Compliance with the criminal law is a universal obligation. By comparison, legal duties in tort or equity will commonly and contractual duties will always be specific to particular relationships. The character of these relationships may vary widely from case to case. They do not lend themselves so readily to the formulation of a general rule. Breaches of civil statutory duties give rise to yet other difficulties. Their relevance may depend on the purpose of the relevant statutory provision, which may or may not be consistent with its deployment as an element in the tort of conspiracy. For present purposes it is unnecessary to say anything more about unlawful means of these kinds.”

While the House of Lords, in Total Network, was dealing only with criminal conduct as unlawful means, and the Supreme Court in Ablyazov expressly excluded other conduct including breach of contract from their considerations, in my judgment if a civil wrong, not actionable at the suit of the claimant (such as a breach of statutory duty or breach of contract with a third party) is capable of constituting “unlawful means”, then it must at the very least be on the right side of the distinction drawn by Lords Walker, Mance and Hope in Total Network referred to in paragraphs 234 to 237 above. That is, it must be the “instrumentality” by which the claimant was harmed, and not merely incidental.

The dividing line itself is not an easy one to draw. The nature of the distinction is, however, given further colour by two examples given in Total Network of cases falling on the “merely incidental” side of the line.

The first example is that given by Lord Nicholls in OBG, cited by Lord Walker at [95] of Total Network. At [159] of OBG, Lord Nicholls said:

“The answer lies in keeping firmly in mind that, in these threeparty situations, the function of the tort is to provide a remedy where the claimant is harmed through the instrumentality of a third party. That would not be so in the patent example.”

By the “patent example”, Lord Nicholls was referring to what he said at [157]:

“Take the case of a patent. A manufacturer seeks to steal a march on his rival by employing a novel, patented process. In order to sell his product more cheaply, he does so without paying any licence fee to the owner of the patent. By means of this patent infringement he undercuts his law-abiding rival. He has damaged his rival’s business by an unlawful means. But this conduct, however reprehensible, cannot afford the rival manufacturer a cause of action for damages for interference with trade by unlawful means. Parliament has specified the nature and extent of the remedies available for infringement of patents. Remedial relief for infringement of a patent is available to patentees and exclusive licensees. It would be inconsistent with the statutory scheme if the common law tort were to afford a remedy more widely.”

The second example, also given by Lord Walker at [95] of Total Network, is the case of Lonrho v Shell. Lonrho was the co-owner of an oil pipeline from Beira in Mozambique to a refinery in what was then Southern Rhodesia. The UK government had imposed sanctions on Rhodesia, after the declaration of UDI, under the Southern Rhodesia Act 1965. That prevented Lonrho from operating its pipeline. Lonrho claimed that Shell and its co-defendant BP had breached the sanctions and thus (by prolonging the period during which Lonrho could not use the pipeline) caused loss to Lonrho. Lord Walker cited this as an example of unlawful means that were not the instrument for inflicting harm, and thus would not qualify as unlawful means for the purposes of the tort of conspiracy.

Further guidance is found in the comment of Lord Walker, at [96] of Total Network, to the effect that the question whether a breach of statutory duty could amount to unlawful means may well overlap or coincide with the sort of considerations relevant to determining whether a statutory breach of duty is actionable in a civil suit, citing Cutler v Wandsworth Stadium Ltd [1949] AC 398. Those considerations include whether the statutory provision was enacted for the benefit of a particular class of persons, and whether there is anything in the statute which indicates it was intended to confer a right of civil action.

While the wrongdoing in both the infringement of patent example (OBG) and the breach of sanctions example (Lonrho) involved statutory rights and obligations, I consider that the reasoning employed would have led to the same conclusion had the relevant wrongful act been a breach of contract. If that is so, I do not see how the breach of the exchanges’ terms and conditions can be described as directed at the claimants, or the instrumentality of the harm caused to them. While (assuming all other elements of the tort are present) the use of the prices on the exchanges’ websites is the means by which SIS was able to publish its own prices and thus compete with TRP, the unlawfulness involved in that conduct was directed solely at the exchanges. The exchanges’ terms and conditions were there solely for the exchanges’ benefit, and the only persons damaged by the breach of those terms were the exchanges. From the point of view of the claimants, it was irrelevant whether SIS’s use of the exchanges’ prices was done by agreement with the exchanges (which would no doubt have involved payment of a fee) or not.

Accordingly, I find that the use of the exchanges’ prices, in breach of their terms, did not constitute unlawful means for the purposes of the tort of conspiracy.

The remaining elements of unlawful means conspiracy: the law

It is common ground that a claim in unlawful means conspiracy involves at least the following elements: (1) a combination or agreement between two or more persons; (2) to take action which is unlawful; (3) with the intention (but not necessarily the predominant purpose) of causing damage to the Claimant; and (4) the Claimant suffers damage as a result: Kuwait Oil Tanker Co SAK v Al-Bader (No.3) [2000] 2 All ER (Comm) 271 (“Kuwait Oil”) at [108].

SIS contends (but the claimants dispute) that there is an additional mental element, namely that the defendant knew the relevant conduct was unlawful. If there is such a requirement, the parties are in disagreement as to where the burden of proof lies (i.e. whether such knowledge is an element of the claim, or whether it would be a defence to establish that the defendant believed the conduct was lawful).

If knowledge is a requirement of the tort, the parties are agreed that knowledge includes both actual knowledge and ‘blind-eye’ knowledge, i.e. knowledge attributed to a person as a consequence of his ‘wilful blindness’ (per Lord Millett in Twinsectra v Yardley [2002] 2 AC 164, at [112]).

Combination

The “combination” required for the tort of conspiracy need not be established by express agreement: “it is sufficient if two or more persons combine with a common intention, or, in other words, that they deliberately combine, albeit tacitly, to achieve a common end”; moreover, “it is not necessary for the conspirators all to join the conspiracy at the same time, but … the parties to it must be sufficiently aware of the surrounding circumstances and share the same object for it properly to be said that they were acting in concert at the time of the acts complained of”: see Kuwait Oil at [111].

It is a rare case where there is evidence of an agreement and, in most cases, “…it will be necessary to scrutinise the acts relied upon in order to see what inferences can be drawn as to the existence or otherwise of the alleged conspiracy or combination”:

Knowledge of unlawfulness

In British Industrial Plastics v Ferguson [1938] 4 All ER 504, where there were parallel claims for inducing a breach of contract and conspiracy to injure by unlawful means, the Court of Appeal held that it is an essential requirement in the tort of conspiracy that the defendant knew of the unlawfulness of the conduct. It did so on the basis that the requirement was the same as that in the tort of inducing a breach of contract. Indeed, it appears to have been common ground at trial that both torts required a “fraudulent mind”: see the extract from the judgment at first instance quoted at p.515F-G. The requirement was expressed most clearly by Finlay LJ who said, at p.514D-E:

“The essence of conspiracy is the co-operation of the minds of the conspirators in pursuance of the unlawful design. A person could never be liable for conspiracy, either in a civil or in a criminal court, if he had no knowledge that the design was unlawful.”

In Belmont Finance Corporation v Williams Furniture (No.2) [1980] 1 ALL ER 393, however, the Court of Appeal came to the opposite conclusion, holding that ignorance of the law was no defence in the tort of conspiracy. In that case, the defendants participated in a scheme, involving an acquisition by the claimant company from a third party with the sole purpose of putting that third party in funds so that it could acquire shares in the claimant company. This constituted a breach of s.54 of the Companies Act 1948, which made it unlawful for a company to give financial assistance for the purpose of a purchase made by any person of or for any shares in the company. The Court of Appeal held that the defendants were liable for the tort of conspiracy to inflict damage by unlawful means, whether or not they were aware that the transaction constituted a breach of s.54.

At pp. 404-405 Buckley LJ distinguished between a defendant who “sincerely believed in a factual state of affairs which, if true, would have made their actions legal” and a defendant who (for example because they had taken advice) “may have believed in good faith that the transaction did not transgress s.54”. The former would be afforded a defence, but the latter would not. At p.404h Buckley LJ said: “If all the

Ferguson and Belmont were both cited to the House of Lords in OBG v Allen, but only Ferguson was referred to (with approval), and then only in the context of the tort of inducing a breach of contract.

Both decisions were referred to by the Court of Appeal in Meretz Investments NV v ACP Ltd [2008] Ch 244, a case in which claims were pursued both for inducing breach of contract and conspiracy. Arden LJ referred to Ferguson with approval (at [114]), but only in the context of the tort of inducing breach of contract. At [118] she addressed the argument that there is a tension between (1) the fact that if a party is advised that a transaction does not infringe the claimant’s rights, then he may not have the intention necessary to commit “an economic tort” and (2) the principle to be derived from Belmont that “in the tort of conspiracy ignorance of the law is no defence”. That was an argument, she noted however, that had been rejected by the Court of Appeal in Mainstream Properties Ltd v Young [2005] IRLR 964. The Mainstream case concerned a claim for inducing breach of contract. In rejecting the argument that ignorance of the law is no defence (at [84] and [85] of her judgment in the Mainstream case) Arden LJ referred to and disapproved that part of the judgment of Goff LJ in Pritchard v Briggs [1980] Ch 338 in which he concluded that “if a person knows the facts which make the conduct tortious, a mistake as to the legal consequences flowing from those facts is no defence.” Goff LJ was there dealing with a claim in conspiracy. Arden LJ disapproved this passage because it was difficult to reconcile with Ferguson.

At paragraph [127] of her judgment in Meretz, Arden LJ said:

“In this case the defendants envisaged that when FP exercised the power of sale Britel would lose its right to the development sublease. However, as I have explained, they were advised and believed that the exercise of the power of sale would overreach Britel’s rights to the development sublease. This was an inevitable result of the arrangements to which Britel agreed. The mere fact that Mr Tamimi or FP intended that result to occur does not mean that they had the intention to case harm for the purposes of the tort of inducing breach of contract. All they intended to do was to produce a result which they believed, as a result of the contractual arrangements between ACP, FP and Britel, that they were entitled to produce.”

Although, in this passage, she was purporting to deal only with the tort of inducing a breach of contract, she reached the same conclusion, at [146], in relation to the tort of conspiracy to injure by unlawful means, saying:

“It follows therefore that it is not enough that there is an intention to do an act which in fact causes loss. That act must be done with the intention that it will cause loss. In the present case Mr Hawkins gave advice that the leaseback option would be overreached. That advice was inconsistent with an intention to cause harm to Britel. Nor was there any intention to prevent performance of any residual liability in damages. In those circumstances, in my judgment, the requirement as to intention was not satisfied in the case of any defendant.” 265. Toulson LJ, at [174] of Meretz, said the following:

“Although my conclusion on the issue of unlawful means makes it unnecessary to decide the point, I would support Arden LJ's view, at para 127, that it is a defence to an action for conspiracy to injure by unlawful means if the defendant not only acted to protect his own interests but did so in the belief that he had a lawful right to act as he did. Just as the tort of conspiracy to induce breach of contract is not committed if the defendant believes that the outcome sought by him will not involve a breach of contract (the Mainstream case [2005] IRLR 964), so a defendant should not be liable for conspiracy to injure by unlawful means if he believes that he has a lawful right to do what he is doing. This is consistent with Lord Hoffmann's comment in the OBG case [2008] 1 AC 1, para 56, when considering the tort of causing injury by unlawful means, that the common law in this area is designed only to enforce basic standards of civilised behaviour.”

The conflicting Court of Appeal decisions, and numerous other first instance decisions on the point, were the subject of a comprehensive review by Morgan J in Digicel (St Lucia) Limited v Cable & Wireless Plc [2010] EWHC 774 (Ch), at Appendix I to his judgment. Morgan J noted (at [110]) a number of limitations with the apparently clear endorsement by Toulson LJ (at [127] of Meretz) of the proposition that a person should not be liable for conspiracy to injure by unlawful means if he believed he had a lawful right to do what he was doing. The limitations included that Toulson LJ had held it was unnecessary to decide the point and that the cross reference to [127] in the judgment of Arden LJ is to a passage dealing with inducing breach of contract and not the tort of conspiracy.

Nevertheless, Morgan J concluded as follows (at [117]):

“But for the decision of the Court of Appeal in British Industrial Plastics Ltd v Ferguson, I might have been prepared to hold that I was bound by the ratio of Belmont and I should follow that ratio notwithstanding the obiter dicta in Meretz . However, in view of what was said by the Court of Appeal in British Industrial Plastics Ltd v Ferguson together with the dicta in the Court of Appeal in Meretz, my conclusion is that a judge at first instance ought to follow what is clearly stated by Toulson LJ in Meretz at [174].”

In First Subsea Ltd v Baltec Ltd [2014] EWHC 866 (Ch), Norris J, at [150]-[157] considered himself bound by the decision in Belmont, rather than the obiter comments in Meretz, and concluded that knowledge of the unlawfulness of the relevant acts was not an ingredient in the tort of conspiracy to injure by unlawful means. It appears, however, that neither Digicel nor Ferguson was cited in First Subsea v Baltec.

After the conclusion of the trial in this case, HHJ Russen QC sitting as a judge of the High Court handed down judgment in Stobart Group v Tinkler [2019] EWHC 258 (Comm), in which the conflicting authorities in this area were again subjected to a thorough review, and in which he came to the opposite conclusion to Morgan J. I have received supplemental post-trial written submissions from the parties addressing the impact of this decision.

HHJ Russen QC dealt at length with Meretz (at [548] to [554] and [568]). He noted that Arden LJ, in the passages I have quoted above, appeared to be dealing with the requirement in the tort of conspiracy that the defendant intended to harm the claimant, not with whether it was a requirement that the defendant knew that the means were unlawful. He also noted the same caveats, as Morgan J had done, in relation to Toulson LJ’s conclusion at [174].

HHJ Russen QC, unlike Morgan J, was not persuaded by what was said by the Court of Appeal in Ferguson. He considered (at [566]) that the Court of Appeal in Belmont had addressed the question of knowledge of unlawfulness for the purposes of conspiracy “more directly and, on my reading of the two authorities, in much clearer terms than the judgments in British Industrial Plastics v Ferguson”, noting that the Court of Appeal in the latter case proceeded on the basis that the requirements of the two torts under consideration were the same. He reasoned (at [555]) that to add knowledge of unlawfulness as a requirement “would detract from the fundamental point that it is in the fact of conspiracy (with the requisite intention to injure) that unlawfulness resides. It is that fact (conspiracy being reprehensible in any context) that distinguishes this tort from the tort of unlawful interference.” Moreover, it would raise “a whole host of unanswered questions such as the standard of knowledge (or suspicion) required, the scope for reliance upon patently unrealistic (and possibly selfserving) legal advice and (if it goes to a “defence” as Lord Justice Toulson contemplated) the burden of proof.”

In a number of cases, the requirements of the tort have been set out as either including, or omitting, knowledge of unlawfulness. Some of these are referred to in Stobart (at [562] to [572]). Since the issue was not raised for decision in those cases, they do not assist in resolving the conflict between Ferguson and Belmont.

Before addressing the issue in more detail, it is important to identify the (limited) extent to which it arises for determination in this case, given that the claimants accept that (as clearly stated in the passages cited from Belmont above) it is undoubtedly an essential requirement of the tort that the defendant knew of the facts which constitute, in law, the relevant unlawful means. Taking in turn each of the unlawful means relied on in this case:

That leaves only the claim in breach of confidence. The point is, from a practical perspective, largely academic in this context too because, as I have already found, SIS is liable to the claimants under a separate direct claim for breach of confidence. In these circumstances, and given that SIS is the only remaining defendant in the action, it is difficult to see what a finding that SIS was also liable in conspiracy would add. Nevertheless, as the claimants have advanced both causes of action, I address the question in the following paragraphs.

The claimants urge me to follow the decision in Stobart. They contend that HHJ Russen QC was correct to hold that: (1) the requirement of knowledge of unlawfulness did not feature in certain authoritative statements of the requirements of the tort, including Kuwait Oil and JSC BTA Bank v Ablyazov (above); (2) Belmont was binding authority which was not undermined by the absence of reference in it to Ferguson, because the Court of Appeal in Ferguson had not addressed the requirements of conspiracy in their own right but on the basis that they were the same as in the tort of inducing breach of contract; (3) there was no sufficient basis in any subsequent case for departing from the position in Belmont; and (4) in particular, for the reasons I have set out above, the comments of Arden LJ and Toulson LJ in Meretz provide no basis for departing from Belmont.

SIS contends that I should follow Ferguson and Digicel. First, as a matter of precedent, because Ferguson is a binding Court of Appeal decision and Belmont (on which Stobart is based) was decided per incuriam since Ferguson was not cited. Second, as a matter of principle, as Toulson LJ said in Meretz, at [174], “the common law in this area is designed only to enforce basic standards of civilised behaviour”, and it is difficult to see how any party could act contrary to such standards if they did something that they did not know was unlawful. Moreover, as noted by the Supreme Court in JSC BTA Bank v Ablyazov at [6], “the successful pursuit of commercial selfinterest necessarily entails the risk of damaging the commercial interests of others.

In my judgment, the arguments of SIS are to be preferred, insofar as the unlawful means consisted of breach of confidence. The Supreme Court has commented, in the different context of what might constitute unlawful means, that different considerations may well apply in the case of breaches of the criminal law, and breaches of the civil law such as torts or breaches of contract, because “legal duties in tort or equity will commonly and contractual duties will always be specific to particular relationships”: Ablyazov at [15]. I consider that the same can be said in the context of whether knowledge of unlawfulness is a requisite part of the tort. It might well be said that it is consistent with enforcing basic standards of behaviour to find a person who conspires to perform criminal acts liable even if, as in Belmont, he or she did not know that the acts were unlawful. But where, as in Ferguson or in this case, the unlawful conduct consists of infringing the claimant’s private law rights, then I consider it is consistent with enforcing basic standards of civilised behaviour that a person is liable for conspiring to injure through such unlawful means only if he or she knows (to the requisite standard) that the claimant’s rights are being infringed.

Although not spelt out in this way, the distinction is reflected in Toulson LJ’s comment at [174] of Meretz that his conclusion that knowledge of unlawfulness was a necessary ingredient in the tort was consistent with the conclusion in Churchill v Watson (that ignorance of the law was no defence), a case involving conspiracy to commit a strict liability offence.

Moreover, I consider that liability in conspiracy ought not to turn on the fine distinctions that would arise if knowledge of the facts giving rise to a breach of private law rights such as breach of contract or breach of confidence was sufficient. The distinction between knowledge of facts and knowledge that those facts constitute unlawfulness may in such cases be difficult to identify. For example, in the case of breach of contract, whether there exists a term which prohibits particular conduct is in a broad sense a question of fact, but uncertainty over the meaning or scope of a term may give rise to questions of law as to interpretation, where the court’s view of the meaning as understood by a reasonable recipient of the document will determine the outcome.

Ferguson is an example of the fine distinctions that can arise in this area. The relevant unlawful conduct was a breach of contract by the defendant, Docherty. Ferguson, the other defendant, did not have knowledge of this, but was suspicious that Docherty might be acting in breach of contract in offering a chemical process to Ferguson. Ferguson believed (wrongly) that if the process was patentable then it could not be secret, such that the claimant’s rights would not be infringed. The issue could be described as a question of fact (was Docherty in breach of contract?) or as one of law (as a matter of law, would the fact that the process was patentable mean that it could not be secret?).

Similarly, in relation to the claim for breach of confidence in this case, there is an overarching question of fact: was the information imparted in circumstances imposing obligations of confidence? Since, however, this involves an objective standard, it involves to some extent – whether a reasonable person would have concluded that the information was imparted in such circumstances – a question of law.

Equally, to find persons liable for conspiracy where the conduct which they have agreed to consists of a breach of the claimant’s private law rights risks tipping the balance too far against legitimate business activities unless those persons know that the claimant’s rights are infringed. In this context, while I agree with HHJ Russen QC that it is the fact of combination that lies at the heart of the tort of conspiracy, I think he overstates the position in concluding that conspiracy is “reprehensible in any context”. That is because what sets the tort of conspiracy apart is the “intense focus” on intent: see Lord Walker, in Total Network (at [76]-[77]), citing Lord Dunedin in

The classic statement of the tort in Mulcahy v R (1868) LR 3 HL 306 at 317, cited with approval by Buckley LJ in Belmont Finance (above) at p.404, is as follows:

“A conspiracy consists not merely of the intention of two or more, but in the agreement of two or more to do an unlawful act, or to do a lawful act by unlawful means.”

In other words, a combination is only ‘reprehensible’ if its object is either (predominantly) to injure the claimant, or to use unlawful means in order to injure the claimant. Importantly, a non-predominant intention to harm the claimant is not in itself a sufficient guilty state of mind for the purposes of the tort. I consider that to find a person liable, where that person knows that a (non-predominant) purpose of the combination is to injure the claimant but honestly believes, for example, that on its true construction the contract between the claimant and one of the conspirators does not prohibit the relevant action, would risk trespassing on legitimate competitive business practices.

While it is true that the Court of Appeal in Ferguson concluded that the same considerations applied to both conspiracy and inducing breach of contract, I do not think that is a reason to dismiss the decision’s significance. The issue was central to the Court’s decision that there was no conspiracy. Moreover, in the key passage in

Finally, I do not believe that the remaining objections to this conclusion referred to at

For the above reasons, I conclude that a person’s liability under the tort of unlawful means conspiracy, where the unlawful means consist of breach of confidence, depends upon that person knowing that (or turning a blind eye to whether) the claimant’s rights of confidence were infringed. It might be said that there is, on the face of it, an odd inconsistency between, on the one hand, liability for breach of confidence depending on an objective question (would a reasonable person in the position of the defendant have realised the information was acquired in circumstances imposing obligations of confidence?) and, on the other hand, liability for conspiracy depending on actual or blind-eye knowledge of breach of confidence. But the inconsistency is fully explained, in my view, by the fact that liability for breach of confidence also depends upon the defendant having made unauthorised use of the information, whereas liability for conspiracy can be established against a person without that person having done more than enter into an agreement with the requisite intention. The oddity is highlighted in this case because the person sought to be made liable in conspiracy (SIS) is the person who (according to my earlier findings) is itself liable under a direct claim in breach of confidence, whereas liability in conspiracy needs to be tested by reference to a person who did no more than agree that someone else would take action constituting a breach of confidence. That merely demonstrates the extent to which the conspiracy claim is likely to be academic if my conclusion on the direct claim is correct.

Burden of proof

Whether the requisite mental element is the knowledge of the facts which constitute unlawful conduct, or that the conduct is unlawful, the question remains as to who bears the burden of proof: is it for the claimant to establish the mental element, or is it a defence for the defendant to establish that it lacked that mental element?

In the only case where this issue has been expressly addressed, the point was left open: see Digicel at [119] of Appendix I to the judgment of Morgan J.

In my judgment, the burden of proof rests on the claimants. I reach this conclusion on the basis of the central role played by intention (see above at paragraph 282) and the classic formulation of the tort in Mulcahy v R (above at paragraph 283), which suggests that the important requirement of intention is not merely to do harm, but to do so by means that are unlawful. Lord Hope said something similar in Total Network (at [44], a paragraph cited with approval by Lords Sumption and LloydJones in Ablyazov at [10]): “a conspiracy is tortious if an intention of the conspirators was to harm the claimant by using unlawful means”.

Given the essential role played by intention, I therefore conclude that the claimant in an action for conspiracy must establish the existence of the requisite knowledge (in the Twinsectra sense of including blind-eye knowledge) of unlawful means. In this case, in relation to the unlawful means consisting of breach of confidence, for the

Knowledge of the other conspirators

Neither party referred me to an authority which has expressly considered this point. In view of my other findings, and particularly in view of the fact that the only alleged conspirator remaining in the action is SIS, it is unnecessary to consider this aspect of the case.

Had it been necessary to do so, I would have found that it is an essential ingredient in the tort of conspiracy to injure by unlawful means that at least one other of the alleged conspirators shared with the defendant the knowledge as to the unlawfulness of the relevant conduct. That is on the basis that, looking at the matter from first principles, the essence of conspiracy is, as Finlay LJ put it in Ferguson at p.514D: “the cooperation of the minds of the conspirators in pursuance of the unlawful design.” An otherwise innocent act of agreement becomes actionable when the object of the agreement is unlawful conduct.

As against this, the claimants relied on a passage from the decision of the Court of Appeal in Kuwait Oil at [133], quoting with approval the following passage from the judgment of Moore-Bick J at first instance:

“Once one reaches the conclusion that the defendants combined to steal from their employer by whatever means might present themselves, the question in relation to any particular scheme or enterprise in which only one or some of them can be shown to have directly participated is whether that enterprise fell within the overall scope of their common design. If several people agree to enable each other to steal from their employer, lending their support in different ways at different times and taking different shares of the proceeds (or even each retaining for himself what he takes), each of them is party to the agreement pursuant to which all the thefts take place. In those circumstances there is in my judgment no need for each to be fully aware of the circumstances of each theft in order for him to be liable as a conspirator provided that the theft in question falls within the scope of their agreement.”

Moore-Bick J was, however, dealing there with a different question, namely the specific manner in which the agreed-upon unlawful conduct (theft) was to be carried out. As is clear from the first sentence, the premise was that defendants had combined “to steal” by whatever means might present themselves, a premise that is only possible if each of them knew that the action to be taken constituted theft.

Intention to injure

There was no disagreement between the parties as to the content of the requirement that the defendant intended to injure the claimant. There needs to be an intention, although not necessarily a predominant intention, to injure the claimant.

Even without there being a specific intent to injure the claimant, the requirement may be satisfied where the injury to the claimant is the necessary corollary of benefit to the defendant. A defendant’s foresight that his conduct may or probably will damage the claimant is not sufficient: see OBG v Allan (above) per Lord Nicholls at [166].

This passage was further explained in WH Newsom Holdings Ltd v IMI Inc [2014] 1 All ER 1132, per Arden LJ at [40]:

“Lord Sumner is taking the situation where loss to plaintiff must follow from the object of the conspiracy. He was taking the case where the proved facts exclude every other inference. As Lord Nicholls put it, the gain and the loss are inseparably linked.”

As noted by the Court of Appeal in Kuwait Oil at [120], it will often be necessary to infer intent from the primary facts.

The remaining elements of the tort of unlawful means conspiracy: application of the law to the facts

As none of the other alleged conspirators (Betfred, LC or the Tote) remain parties to the action, and there is no direct evidence as to their involvement, it is necessary to see what can be inferred as to their involvement, their knowledge and their intentions from the other evidence, in particular the contemporaneous documents.

A combination

Mr Siers (who, in addition to representing the Tote in negotiations with SIS, was

The first mention of LC is in an email from Mr Witten to Mr Smith of 3 November 2016. He referred to having spoken to Gary Bettis of LC that afternoon. LC had been proposing to enter into an agreement with TRP, but Mr Bettis had “kaiboshed” it because the cost was too high. Mr Witten said that he and Mr Bettis had “discussed

By mid-November 2016, SIS had abandoned the idea of providing an alternative feed to bookmakers. In an email of 16 November 2016, to Rachel O’Reilly at SIS, Mr Dolan said “SIS have decided not to provide a retail RDD service for independents or the majors at Arena tracks from 1^st January”. The reason for this was not explained in evidence.

Mr Siers (emailing on behalf of Betfred) emailed Mr Witten on 17 November 2016, expressing surprise at this decision, and expressing concern that this would leave Betfred without data “in the shops or digital”. He asked whether SIS would be able to supply Betfred with official race day data, or whether there was an alternative they could offer. This appears to have prompted a re-think on the part of SIS. Mr Dolan emailed Mr Siers on 22 November 2016, asking him: “if SIS were to provide an alternative race day data feed at TRP Arena tracks from 1^st January to the retail and on line market place but primarily the majors are Totepool legally secure in providing such a feed.” There was no written response to that question, but it is apparent from this, and subsequent emails, that Mr Dolan spoke on the telephone to Mr Siers later that day. On the next day Mr Dolan emailed his colleagues within SIS noting that to date no major bookmaker had signed up to TRP, and that Betfred had indicated that they were interested in a limited race day data service to enable them to accept and settle bets at Arena Racecourses. He asked his colleagues to approach “the majors” to see if there was an interest in SIS providing a “limited” raceday data service. On 24 November 2016 Mr Dolan emailed Mr Siers, to tell him that SIS were now canvassing the major bookmakers to gauge their interest in a diluted feed.

On 29 November 2016 Mr Dolan emailed Mr Witten noting that he had spoken to LC the day before, and asking whether there was any interest from them. Later that same day Mr Smith emailed Mr Witten and others reporting on his discussion with Mr Siers (see paragraph 178 above). Mr Witten responded to say that he had been taking soundings from major customers but there was no commitment as yet.

On 16 December 2016, Mr Smith emailed Mr Witten and others, reporting on his conversation with Mark Chambers of LC. Mr Chambers had said that LC had not yet decided whether to enter into an agreement with TRP, that they were frustrated with the cost TRP was asking for. Mr Smith said that he had told Mr Chambers that SIS would do what it could to provide LC with options if it decided not to contract with TRP, including the provision of data services. In reply, Mr Witten reported that LC now did want to “take up the SIS offer of a Tote derived feed (having previously advised that they didn’t)”. Mr Smith responded saying: “just a thought, and its probably obvious, try and get the deal with the Tote on a three-month rolling basis. From my conversation with Mark [Chambers], I think they are considering the possibility of going without TRP for a period of time to drive a better deal with TRP.

On 20 December 2016, Mr Witten emailed Mr Smith and others, reporting on a call he had had with LC the day before, in which LC said that it was now “highly unlikely” that a deal would be reached with TRP and that LC wished to work with SIS “in a partnership arrangement to counter the situation”. Mr Witten referred to various steps which SIS needed to take as part of a “mitigation strategy”. These included identifying and supplying the alternative Tote data feed, confirming all international horseracing content available to LC, and confirming the position in relation to greyhound fixtures. SIS had in fact been in discussion with LC since at least November relating to providing coverage of greyhound meetings to replace the allweather racing from Arena tracks from January 2017.

These discussions, and negotiations over the price LC would be charged for the service, continued over the ensuing days. For example, on 21 December 2016 Mr

It is common ground that SIS provided Raceday Data and Betting Shows to Betfred, from 1 January 2017 until July 2017, notwithstanding that there appears to have been no written agreement, and no payment was received by SIS, in respect of that service.

On the basis of the contemporaneous communications referred to above (and those relating to the negotiation of the agreement between SIS and the Tote – see paragraphs 174 to 182 above), I find as follows:

Knowledge of unlawfulness of means

So far as SIS’s knowledge of the unlawfulness of means is concerned, I have set out my findings in relation to each of the alleged unlawful means at various points above in this judgment. In summary:

So far as the other conspirators are concerned, I find as follows.

Intention to injure

The claimants contend that SIS (and the other conspirators) intended to injure them in two ways. First, it is said that although it and the other conspirators acted in order to pursue their own commercial interests (by providing LC and Betfred with leverage in negotiations with TRP) this would necessarily cause harm to the claimants. Second, it is said that SIS and the other conspirators intended to damage the value of TRP so that SIS could make a bid to acquire Arena.

In relation to the first way the case is put, the claimants’ submissions were primarily directed towards establishing that SIS’s actions were deliberately aimed at promoting the interests of LC and Betfred. In this regard they rely on: the commercial relationship between the entities (noting that LC and Betfred are both customers of, and indirect shareholders in, SIS); the numerous references in the contemporaneous documents to providing LC with an alternative to the TRP feed so as to give LC leverage in its negotiations with TRP (bolstered by references to the assumption that the alternative feed would only be needed for a short time, and the “otherwise bizarre provision” in the Tote HoT permitting termination on one month’s notice); the fact that the supply of Tote data was either loss-making or at best marginal; and other documents evidencing SIS adopting and promoting the interest of LC and Betfred

I accept (as I have found when considering the purpose of the combination: see paragraph 310 above) that it was the intention of LC and Betfred to take Raceday Data and Betting Shows from SIS in order to improve their negotiating position with TRP, and that SIS was aware of this. I am also prepared to accept that, in pursuing the provision of the information to LC and Betfred with knowledge of LC’s and Betfred’s purpose, SIS can be taken to have intended that benefit to accrue to them.

In so far as this point is put generally (i.e. relating to all of the alleged unlawful means pleaded) I disagree, however, that this is sufficient to constitute intention to injure TRP for the purposes of the tort of conspiracy. That is because, while damage to TRP is a likely foreseeable consequence of providing LC and Betfred with leverage in their negotiations with TRP, it is not a necessary corollary. There was no evidence led as to what in fact occurred in negotiations between TRP and LC/Betfred, and I do not think it can be said with certainty that – to the extent that any part of the data provided by SIS to LC/Betfred was achieved through unlawful means – that without it TRP would have achieved a higher price for its own data service. I note in this respect that in Mr Smith’s email to Mr Witten and others within SIS of 16 December 2016, Mr Smith reported that LC were expecting to decide shortly whether to contract with

The second way the claimants put their case on intent to injure is that SIS was motivated by the desire to reduce the value of TRP so that it (or more likely an associated company) could purchase Arena. This was based on isolated references in minutes of meetings of SIS’s holding company in the early part of 2017 to the possibility of making an offer to buy Arena, if LC resisted contracting with TRP for a reasonable period. In a CEO report from March 2017 it was noted that the plan to present a speculative offer to try to acquire Arena had progressed with an initial approach to the three bookmaker shareholders, and that all three had indicated interest in discussing the proposal further. The only evidence of this possibility being contemplated dates, however, from after SIS commenced transmitting Raceday Data and the SIS Betting Shows to LC, Betfred and other bookmakers. It was clearly regarded as speculative. It did not progress even as far as making an offer. Accordingly, I find that it was not part of SIS’s or the other alleged conspirators’ intention in deciding to produce SIS Betting Shows or distribute Raceday Data via the Tote to engineer a take-over of Arena. In any event, even if that had been their intention, I find that it would not have constituted an intent to injure the claimants, since injury to them was not the necessary corollary of creating a favourable environment for making an offer to acquire Arena.

It is possible, however, to put the point more narrowly, insofar as the combination contemplated taking steps which were unlawful and which would inevitably cause damage to TRP. If I had found, therefore, that SIS and Betfred/the Tote had the requisite knowledge that using the Tote Raceday Data for fixed-odds betting purposes was unlawful conduct, then I would have been prepared to find that SIS had the requisite intention to injure TRP and Arena. That is because misuse of confidential information which Arena had (for its monetary benefit) granted TRP the exclusive right to use (for its monetary benefit) necessarily caused loss to Arena and TRP, even if SIS’s intention was to use the information for the benefit of it and others.

Conclusion

For the reasons set out above, I conclude that:

In his first statement, Mr Tompkins, having referred to the fact that in early January the SIS traders did make reference to the PA Betting Shows on Oddschecker when creating SIS Betting Shows, said that “Since then I have not used Oddschecker or any other comparison website, or Bet365 or any other bookmaker websites at all when setting [SIS Betting Shows].”

The claimants challenged this evidence in a number of ways. It was first suggested that on the face of his first statement Mr Tompkins had expressly admitted continuing to look at Oddschecker throughout 2017. This was based on the fact that Mr Tompkins had said (referring to the period after the first couple of days in January 2017) that “from this time onwards” he believed that SIS traders continued to operate as they had before, including by checking bookmakers’ prices. While this part of his statement was poorly drafted, I accept Mr Tompkins’ explanation that he was intending to refer only to the following few days in January until it was made clear to the traders that they must only use the exchanges. It is highly unlikely that he intended otherwise, given his unequivocal statement a few paragraphs later, that he has not used Oddschecker at all since some point in January.

The claimants also refer to the fact that Mr Tompkins’ evidence as to precisely when, in January, the practice of looking at Oddschecker for the purposes of creating Betting Shows ceased, was confused. For example, in his witness statement he said that the instruction to stop looking at Oddschecker was received, and acted on, a few days into January, but in his oral evidence he variously referred to the practice having ceased

“somewhere towards the end of January”, “a point in January” and “some time in January”.

It was Mr Tompkins’ evidence that, even after the practice of looking at Oddschecker for the purpose of creating the SIS Betting Shows ceased, the Oddschecker website remained open on one of the screens in front of him throughout the day, and that this would have been the same for other traders. Mr Conway suggested that although Oddschecker remained “available” to the traders throughout the day, it would be minimised during the period a trader was compiling a Betting Show. I prefer Mr Tompkins’ evidence on this, as he was directly involved in compiling Betting Shows, whereas Mr Conway was not.

However, Mr Tompkins said that the reason Oddschecker remained open all day was because it was a source for setting Early Prices. As a trader he would be given responsibility for a particular racecourse on a particular day. His job would be to set Early Prices for all the races at the start of the day, and adjust those Early Prices during the day. SIS traders would review a variety of prices from different bookmakers (available on Oddschecker) when setting Early Prices, and assess where the SIS Early Price should sit by reference to the market. He said that the last time before each race that he would look at Oddschecker for the purposes of setting Early Prices (both for the next race, and later races) was approximately 10-15 minutes before the off. From then onwards, he would look only at the exchanges for the purpose of creating the Betting Shows for that race.

In response to questioning about Oddschecker, Mr Tompkins said that the webpage visible on his computer screen did not automatically refresh, and he would have to refresh it manually. He accepted that although there was a browser add-on facility available to traders to refresh certain websites automatically, that was for something else, and not used in relation to Oddschecker. It was put to him that the prices appearing on the page updated automatically, but he said that that was something that had been implemented only in the last two to three months, and that it did not happen in 2017. There was no evidence to contradict Mr Tompkins’ account of how Oddschecker worked in 2017.

It was put to Mr Tompkins that if he looked at the Oddschecker website for the last time in relation to a particular race approximately 10-15 minutes before the off, then it was likely that he would have seen the first of the PA Betting Shows for that race. Mr Tompkins accepted that it was possible. In fact, the unchallenged evidence of Mr Lamb of PA (called by the claimants) was that the first PA Betting Show was usually transmitted between eight and ten minutes before the off. While it is indeed possible that a SIS trader who refreshed Oddschecker 10-15 minutes before a race might see prices derived from the first PA Betting Show, it is unlikely that this would have happened on anything other than an occasional basis.

The claimants rely on a note of interview Mr Tompkins had with Mr Adam Tracey, set out in an email from Mr Tracey dated 5 October 2017. The note records that Mr Tompkins said he found his job monotonous and mundane and that he did not like

“simply skimming Odds checker [sic] all day for price moves” and that “scanning odds checker [sic]” was not what he joined a horse racing team for. On its face, this suggests that Mr Tompkins was indeed using Oddschecker when compiling SIS Betting Shows. In his second witness statement, this meeting note having been brought to his attention, he sought to explain it by saying that he was referring to what he was doing for most of the time each day, which was looking at Oddschecker for the purpose of setting and adjusting Early Prices for each race that day. In his oral evidence, he said that the period during which Betting Shows for each race were created lasted for something like eight or ten minutes, so that if there were seven races in a day, he would only be spending about one hour actively setting prices as part of a Betting Show. It was the rest of the day that he was referring to as mundane, and spent scanning Oddschecker.

There is support for his explanation in the note itself, which records Mr Tompkins saying that mistakes would always be made in the way the system worked at the moment, because “Scanning Oddschecker when you have 2/3 meetings means you can’t physically catch every price move. You may check race and then it won’t be 20 minutes until you’ve checked it again”. The reference to checking back into Oddschecker after 20 minutes is inconsistent with using it for the purposes of creating Betting Shows, which is done during a period of no more than eight or ten minutes.

The only other witness who spoke to the manner in which the traders compiled the SIS Betting Shows was Mr Conway. He was not directly involved, however, in creating Betting Shows. His involvement in the Betting Shows was limited to advising the traders. The matters on which he could give direct evidence were primarily the instructions given to traders to use only the exchanges when setting prices for the purposes of SIS Betting Shows and the complaints that were received

There were certain respects in which Mr Conway’s evidence was unsatisfactory, for example his suggestion – given for the first time in his oral evidence – that he gave an instruction before Christmas 2016, that the traders should only use the Exchanges in preparing SIS Betting Shows. This was contrary to Mr Tompkins’ evidence, and was inherently unlikely given that it was not until shortly before the start of the new year that it was clear that SIS would be producing its own Betting Shows for the Arena Racecourses. A further example is his insistence that traders minimised the Oddschecker website during Betting Shows. That was (as I have already noted) contrary to Mr Tompkins’ evidence, which I prefer. He was also unclear as to precisely when the traders stopped referring to the PA Betting Shows, referring variously to the first few days of January, 9^th January and mid-January. Not surprisingly, since he did not trade himself and was not continually observing the trading practices of the traders, he could not say with certainty when, if at all, the practice stopped.

Nevertheless, I accept his evidence that instructions were given – at some point in January - to the traders that they should not look at Oddschecker and that so far as he was aware, this was complied with thereafter. Mr Ames (now the CEO of SIS but at the relevant time working as a contractor with the SIS Products Team), in his witness statement, said that he became concerned, on receipt of an email from Mr Conway on 4 January 2017, that the traders were looking at PA Betting Shows and that he telephoned Mr Conway to make it clear that traders should not do so. Mr Ames was not challenged on that evidence.

Significantly, there is contemporaneous evidence that: the SIS traders had been instructed not to refer to the PA Betting Shows when compiling the SIS Betting Shows; that the reason for this was because SIS understood that it was not lawful to refer to the PA Betting Shows; that this led to complaints from customers because of price differences between the SIS prices and the PA prices; and that at least some customers were told that this was because SIS created its prices by reference to the Exchanges and were thus only “guide prices”.

For example, on 16 January 2017, David Richardson (retail support supervisor at SIS) emailed his colleague John Alderson to say: “I have just taken a call from the owner of SR30498 who will be contacting you tomorrow. The 17:30 Wolverhampton race today had No.1 Roy’s Legacy odds differing vastly from TRP’s (PA) prices. I have been told that the trading team use Betfair to provide these prices (shouldn’t mention that to customer) … The same customer also laid a bet last week where the prices were again vastly different to TRP’s prices and this cost him £300 for which I took a lot of abuse from him and other customers. I explained to him that SIS prices are guide prices and not official prices on both occasions but he is not happy [with] the service.”

Mr Conway, who received a copy of this email, asked Mr Richardson or Mr Alderson to give him a call. The content of that call is revealed by an email from Mr Alderson to Mr Conway and others on 17 January in which Mr Alderson said as follows: “after speaking to Adam today, we agree that it would be useful to regularly keep each other updated regarding the issues we are having with our “guide prices”. This is due to the

On 24th of January 2017, Mr Alderson emailed Mr Conway saying that SIS had received three further complaints regarding their ‘guide prices’. He said: “Adam has been very helpful and kindly explained that we are using the betting exchanges to price up our markets but customers are not accepting the reliability of this method. It would also be useful if I could guarantee to the customers that we are coming to some form of solution in the near future”.

It is clear that at least some customers were being told that that the discrepancies arose from the fact that SIS was using the Exchanges to compile its Betting Shows from the following email of 30 January 2017 from Mr Robert Prior of Bet Bet Bet 24/7 to Mr Alderson in which Mr Prior said: “Having a former trader compiling shows from Betfair is frankly nuts. Betfair prices bare [sic] no relation to betting shop shows highlighted by the Gabrial [sic] The Thug incident last week when 4/1 was showing on ISIS [a SIS feed] and the TRP price was 3/1 shortening to 5/2. It would be far more sensible to compile prices from a bookmakers website and I suggest Hills would be the obvious choice as they would, I assume, be getting a direct TRP feed…”

Mr Alderson forwarded the email to Mr Conway, saying “I have repeatedly explained to him that for legal reasons, we cannot monitor TRP’s feed and that our guide prices are actually in line with the market/exchanges and it’s the track prices that are out of line, but this doesn’t satisfy as the track generates the official SP and not the exchanges. He has since suggested alternative (and unsuitable) methods of providing guide prices such as lifting them from the William Hill website.”

Customer complaints continued. The claimants point to an email dated 11 September 2017 from a Mr Simon Lewis, a bookmaker, reporting that “the problems we are now having with customer complaints and staff complaints about wrong prices on our screens seem to be getting worse than ever. I have looked into it and it seems that SIS are still supplying our text screens with prices that do not mirror the prices in our tills or on the live screens in our shops. I have spoken with TRP and they tell me that SIS could use the official prices free of charge but SIS seem not to care about the effect this is having on our trade and continue to leave us with a very poor service.”

As against this, the claimants point to certain documents which might suggest that the practice of looking at the PA Betting Shows continued. On 4 January 2017, for example, in response to an email from Mr Ames asking him to “confirm that the prices we are adding to the tote race day data feed the BetFred are our prices and not show prices?” Mr Conway wrote: “That’s correct we are following the same process as we follow for the other non-SIS tracks which is to provide SIS pricing. However recently there was a negative feedback from some of our customers that the SIS prices weren’t close enough to the official show prices and since then we’ve made more effort to be closer to the shows. This can be dialled up or down…” Mr Conway explained, however, (and I accept) that it was in response to this email that Mr Ames telephoned him to say that no reference should be made to the PA Betting Shows.

The claimants also point to an email from Andrew Kelly (the supervisor of data operations at SIS) to Mark Slater (who worked for Mr Kelly), forwarding a complaint from a customer and asking whether traders were aware of the issues raised in it. Mr Slater replied: “Traders are aware as I spoke to Adam, Matt & David this morning regarding the prices. They informed me that they are under pressure from above not to be the same as TRP ? I said what so similar prices? To which Adam said yes.” The claimants submit that this demonstrates that the SIS traders were under instructions to make the prices similar to, but not the same as, TRP’s prices. Mr Conway denied that was what was meant. He said that was “absolutely not” the sense of the conversation he had had with Mr Slater and that what he had said was that the pressure was not to look at TRP prices at all. On balance, I believe Mr Conway on this point. The claimants accept (indeed it is part of their case) that SIS’s traders were using the Exchanges, at least in the first place, to compile SIS Betting Shows. It would have made no sense for them to do this, if they had been told that they could look at TRP prices, and then adjust them so that SIS’s prices were similar to, but not the same.

Finally, the claimants rely on an email of 27 June 2017 from Mr Craig Budden (who had recently joined SIS as lead horse racing trainer) to the trading team as follows:

“having monitored you guys for the previous weeks, just thought I would iterate how we should be trading during boards: 30 mins before the first race, or after each preceding race, spend 10 minutes checking all your later races/early prices, then you should be trading the next race explicitly from the exchange prices. You should not be having 365 or any other bookie not taking our prices on screen. Only if you have a dirty each way race should you maybe have odds checker open to do a little comparison to ensure that you are not extremely high on the outsiders. I would advise that over the coming weeks, watch the boards for SIS meetings done from the track and compare how those prices sit against the exchange prices in the margins they are betting too as we should be looking to bet similarly.” Mr Conway said that when he saw that email, he immediately realised that Mr Budden had not appreciated that SIS traders should not be looking, at all, at the PA Betting Shows, and that is what he told Mr Budden.