First-tier Tribunal (Tax)

You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> The"Spotting the Ball" Partnership & Others v Revenue & Customs [2013] UKFTT 210 (TC) (05 March 2013)
URL: http://www.bailii.org/uk/cases/UKFTT/TC/2013/TC02624.html
Cite as: [2013] LLR 845, [2013] SFTD 1004, [2013] STI 1746, [2013] UKFTT 210 (TC)

[New search] [Printable PDF version] [Help]

The"Spotting the Ball" Partnership & Others v Revenue & Customs [2013] UKFTT 210 (TC) (05 March 2013)
VAT - EXEMPT SUPPLIES
Betting, Gaming and Lotteries

[2013] UKFTT 210 (TC)

TC02624

Appeal numbers:

TC/2011/0562 & TC/2011/00736-0744 (inc)

VAT – exemption for games of chance – whether it applies to “Spot the Ball” – whether “Spot the Ball” is a “game” – held yes – whether it is a “game of chance” – held yes –exemption from VAT under group 4, Schedule 9 Value Added Tax Act 1994 and predecessor provisions therefore applies in principle

FIRST-TIER TRIBUNAL

TAX CHAMBER

THE “SPOTTING THE BALL” PARTNERSHIP (1)

IFX COMPETITIONS LIMITED (2)

TOWN AND COUNTRY FACTORS LIMITED (3)

VERNONS COMPETITION COMPANY (4)

SPORTECH PLC (5)

IFX INVESTMENT COMPANY LIMITED (6)

THE FOOTBALL POOLS LIMITED (formerly LITTLEWOODS PROMOTIONS LIMITED) (7)

VERNONS GAMES LIMITED (8)

THE LITTLEWOODS/ZETTERS PARTNERSHIP (9)

FOOTBALL POOLS COMPETITION COMPANY LIMITED (formerly LITTLEWOODS COMPETITION COMPANY LIMITED) (10)

Appellants

-and-

THE COMMISSIONERS FOR HER MAJESTY’S REVENUE & CUSTOMS

Respondents

TRIBUNAL:

JUDGE KEVIN POOLE

SHAHWAR SADEQUE

Sitting in public in Victoria House, Bloomsbury Place, London on 7, 8 and 9 October 2012

Jonathan Peacock QC instructed by Deloitte LLP for the Appellant

Andrew McNab of Counsel instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

DECISION

Introduction

1. This appeal concerns the correct VAT treatment of a version of the activity known as “Spot the Ball” (referred to in this decision as “STB”). The version it is concerned with is “panel STB”.

2. In all versions of STB, a photograph is taken of a football match. The football is removed from that photograph (along with much of the rest of the background). The doctored photograph (which therefore includes the players, some part of the original background and a great deal of white space) is published on a printed coupon. Participants pay to submit a coupon with a number of crosses placed on it by them, each of which usually represents an attempt to identify as accurately as they can the most logical location of the centre of the missing football.

3. In the “panel” version of STB, the participants’ efforts are not compared with the actual position of the ball on the original photograph. Instead, there is a panel of football experts appointed by the promoter, comprised of former professional footballers and/or others involved in the administration of the game. After all the participants have submitted their coupons, the panel meets and decides where, in their opinion, is the most logical position for the centre of the missing ball. They do not see the original “undoctored” photograph until after they have done so. All the participants’ coupons which have been submitted are checked to see which of them include crosses closest to the position selected by the panel. Cash or other prizes are awarded to the participants who have placed their crosses closest to that position.

4. The reason for involving the panel was in order to structure the activity in a way which, it was believed, would avoid a particular part of the gaming regulatory legislation; it was not to do with VAT or other taxes.

5. The Appellants claim that they (or entities whose rights of action in that behalf have devolved upon them) have wrongly accounted for output VAT on the payments received from the participants over the period from 1979 to 2006, and they now seek a repayment of that output VAT (net of the associated input tax which was, they agree, wrongly deducted). In short, they claim the supplies were in fact exempt supplies. The sums involved are in the region of £72.5 million plus interest.

6. There are a number of issues between the parties, but for the purposes of this hearing we are asked to address only one of those issues and a decision in principle is required on it.

7. The point of principle to be determined is whether this version of STB is a “game” and, if so, whether it is a game of chance or a game of skill; for this purpose, a game of chance and skill combined (or a game that involves both an element of chance and an element of skill) is treated as a game of chance. The answer to this question will, for the reasons given below, determine whether the supplies in question were or were not properly taxable and may therefore remove the necessity for any discussion of the other issues between the parties

8. In particular, we are not asked to determine at this stage whether the Appellants’ claims should be precluded, in whole or in part, by reason of the lapse of time between the relevant events and the delivery of the Appellants’ voluntary disclosures by which they made their claim, or on the basis of unjust enrichment.

9. The precise detail of the route by which each of the Appellants argues that it has (or may have) become entitled to make a voluntary disclosure the subject of this appeal was also not before us as a matter for decision. The parties are agreed that for the purpose of this decision we may assume for the time being that all the Appellants have such an entitlement.

10. The detail of precisely which of the Appellants operated a relevant STB activity at any particular time was not fully clarified before us, though there was some documentation showing that, from 1975 to 17 June 1994, the activity was conducted by the first-named Appellant partnership which involved various of the other Appellants (or their predecessors); and from 18 June 1994 to 2001 it was conducted through a different partnership, involving different partners, but also separately by an associated company of one of the former partners.

11. However we are not asked to consider the position of any individual operator of the activity over the period 1979 to 2006. Instead, we heard evidence as to the general way in which the activity itself had developed and been operated as a whole over that period, and the parties are agreed that our view of the activity as a whole will be determinative of the point of principle as it applies to all the operators over the relevant period in respect of which these Appellants are claiming.

The facts

Introduction

12. We received witness statements from the following individuals:

(1) John Brooke, Product Manager of The Football Pools Limited, with over thirty years’ experience of STB in various roles within the Littlewoods and Sportech organisations. He described the general process by which the STB coupons were collected and sifted and the winners identified. Unfortunately Mr Brooke was unable to attend the hearing through illness, so there was no opportunity to enlarge upon or test his written evidence. None of it was flagged as being controversial, however.

(2) Peter Dent, Technical Manager in the Project Management Office of Sportech PLC. Mr Dent had over 25 years’ experience of the technical aspects of the detailed processes used in the measurement of the precise location of the panel’s mark of the ball centre, the location of the corresponding marks on participator coupons and the precise measurement of the distances between the two. Mr Dent also gave oral evidence.

(3) Carl Lynn, Finance Director for the UK operations of the Sportech PLC group (which had acquired the Littlewoods share of the STB business in 2000). Mr Lynn had over 20 years’ experience of STB, first as a self-employed collector of coupons and entry fees from members of the public, then on the finance side of Littlewoods Pools (including involvement in its VAT returns from 1997). Mr Lynn also gave oral evidence.

(4) Ashley Little, the Officer of HMRC (which in this decision includes the predecessor body Her Majesty’s Commissioners for Customs and Excise) who formally rejected the Appellants’ claim. Her evidence was uncontentious and of the most formal nature and neither side considered it necessary for her to be called to give oral evidence.

13. We also received copies of various documents associated with the history of STB, including:

(1) a “disclosure request” from HMRC and an associated set of answers from the Appellants dated, respectively, 24 October 2011 and 30 November 2011. This related to the identity of the entities operating the STB businesses over the period from 1979 to at least the end of 2000 and clarified the connection between those entities and the Appellants in this appeal;

(2) copies of correspondence with HMRC in 1979 when the appropriate VAT treatment of STB was the subject of extended debate; and

(3) copies of various rules, coupons and other documents relating to the actual conduct of STB activities over the period from 1970 to date.

14. Finally, we viewed extracts from two television programmes on the subject of how STB competitions actually worked at the relevant times – a “Nationwide” programme of 6 March 1975 and a “How Do They Do That?” programme of 2 September 1994.

15. From the above evidence, we find the following relevant facts.

16. The genesis of STB is a little unclear, but we were told it evolved out of the Littlewoods Pools business which started in the 1920s. It was certainly an established activity by 1970, though it was always much less popular than the football pools.

17. The financial peak of both the football pools and STB was in 1994 for Littlewoods and the Spot the Ball Partnership (a partnership in which a Littlewoods company held a majority partnership share). In that year, revenues from the football pools were over £800 million and revenues from STB were around £65 million. In March 1994 the STB weekly revenue peaked at £1.5 million. At that time, approximately 1.6 million STB coupons were submitted for the weekly competition. There were approximately 4,000 people employed in the pools and STB business, and each week they would process approximately 8 million coupons for both games. The employees were spread across five buildings, three of them in and around Liverpool and the other two in Cardiff and Glasgow. In addition there were around 40,000 to 50,000 local self-employed collectors, who collected the coupons and the payments from the participants and submitted them to the operator through a chain of main collectors and concessionaires. They were paid by commission.

18. The introduction of the National Lottery in November 1994 had a drastic effect on both the pools and STB businesses. In the following year, the revenues of STB fell by 45%, and by the third year they had fallen by 70%. Both businesses had to reduce in size and by the end of 1997 the combined football pools and STB operation was down to one building in Liverpool and less than 1,000 employees. The STB business still continues, but an online variant has been launched and now only about 30,000 paper coupons are received each week (down from 1.6 million at its peak in 1994).

19. This appeal relates to the period from 23 April 1979 to 31 December 2006. In terms of the way STB was actually operated, this breaks down into two periods when there were quite different methods employed of checking the coupons received, though the basic rules remained the same. We summarise first the basic rules (and the way in which participants acted when filling out the coupons) and then the different methods that have been used to find a winner each week.

The rules and modes of participation

20. There has been no single set of rules governing STB over the period we considered. There have been many slight variations, and no exhaustive exercise was undertaken to identify the precise rules applicable from time to time over the whole period, indeed such an exercise would probably not be possible now, given the lapse of time.

21. We were however presented with a sprinkling of documents from 1970 up to the last year or so – sometimes rules or extracts of rules printed on coupons themselves and sometimes separate “rules” documents, and the key provisions of all of them are substantially to the same effect.

22. A typical coupon would invite a participant to “use your skill and judgment to decide from all the information contained in the picture, the spot where you think the centre of the ball is most likely to be and indicate the spot by making a cross....” or some variation on that language. The common elements of the wording we saw on all the coupons was that the participant should use his/her “skill and judgment” in order to decide “where the centre of the ball” was, or was most likely to be, on the photograph. Although participants were therefore instructed to guess the actual location of the missing ball, the rules generally made it clear in one way or another that the winner would be decided not by reference to the actual position of the ball on the original photograph, but by reference to the opinion of the panel of experts as to which entry was most “skilful” or was closest to the panel’s opinion of the most logical position of the ball.

23. There were however some cases in which we were provided with possibly incomplete copies of coupons and no associated rules, or copies of coupons which appeared to be complete but which lacked part of the “normal” content. This applied in particular in relation to:

(1) a “Vernons” coupon (which we were told must originate from the period 1994-96, when Vernons operated its own STB activity), where the participants were clearly deemed to be so familiar with the basic rules of the activity that they were not explained – so that coupon contained no instruction as to the placing of crosses at all (apart from a prohibition on overlapping and a maximum restriction on the number of crosses), nor was there any reference to the role of the panel except as deciding the outcome (with no detail of how they would do it, whether by reference to the original photograph or their own opinion). We were however also provided with a separate standalone document headed “Vernons Spot the Ball Rules” which we were informed applied to the 1994-96 period of operation by Vernons (though the evidence as to how this was done was unclear). This document did refer to the role of the panel in more detail, as being to pick entries that “in their opinion, most accurately show the most logical and probable position of the ball”;

(2) a Littlewoods coupon which we were told was issued in 2008, which referred to a prize of £250,000 “when you mark the centre of the ball” and contained no reference to any panel of experts (though it did refer to “conditions” which could be obtained from Littlewoods; we were separately provided with a document entitled “Littlewoods Spot-the-Ball Conditions – Effective from...17^th January 2004” which did include wording about the panel deciding on the logical position of the ball). We note this coupon actually relates to a date after the end of the period relevant to this appeal.

24. Overall we are satisfied that the exercise that the participants were asked to undertake throughout the period 1979–2006 was an exercise of attempting to mark the centre of the missing ball, in the knowledge that their entries would not be “marked” by reference to the actual position of the missing ball but by reference to the experts’ opinion of the correct location of the missing ball. We are also satisfied that this was precisely what actually happened.

25. It occurred to us that if a panel of experts were being asked to assess the “skill” with which a particular entry had been made, then other criteria apart from simply proximity to the point chosen by the panel might be relevant. For example, the panel might take the view that a coupon with a single cross on it which was very close to the right point might be considered more “skilful” than a coupon with a pre-printed sticker of 1,000 crosses of which one was slightly closer – the comparison being between sniping and carpet-bombing. Mr Lynn confirmed however that so far as he was aware from his personal involvement and discussions with the employees who had actually sat with the panel when they deliberated, the panel never strayed from the simple path of treating a closer cross as “beating” a more distant one.

26. We were shown various coupons which demonstrated the different ways in which participants approached the challenge, some of them rather bizarre. Originally the rules required crosses to be marked on the coupons by hand in ink or biro (without overlapping) but from about 1994 the operators made available various patches of stickers in different sizes – so that participants could simply stick, say, 1,000 small crosses on very easily. Most coupons are completed in this way. We were told that rubber stamps were also made available for the same purpose.

Defining the panel’s location for the centre of the ball

27. In any event, once the participants had placed their crosses on the coupons and the coupons had been collected together for checking (they were all physically brought to the promoters’ offices by Saturday morning each week), the panel made their decision (generally on the Saturday morning) as to the most logical location for the centre of the missing ball at a meeting at the promoters’ premises. This was done by the panel considering a blank coupon and finally deciding where the ball should be. A pinhole was made in that coupon by or under the direction of the chairman of the panel. A manager from the operator placed a Letraset dot over the pinhole to mark it and then the location of the centre of that dot was measured by using a device called “the Final Mark machine”.

28. The Final Mark machine used pre-printed marks on the coupon as a reference point and by means of fine adjustment with a small joystick a hairline cross was moved by the promoter’s operations manager to what he judged to be the exact centre of the panel’s chosen spot (using a magnifying lens to ensure a high level of accuracy). The Final Mark machine then generated precise X and Y co-ordinates for the chosen spot and printed out those co-ordinates.

Comparing the panel’s location with the coupons submitted

29. During the period covered by the appeal, the process used for picking out the winning coupons changed. This was on an unknown date in 2002. We therefore set out both versions of the process used, first the process up to 2002 and second the process thereafter (which is still in use today).

(a) Process up to 2002

30. The co-ordinates produced for the panel’s selected spot by the use of the Final Mark machine were used to produce Perspex “sifting masks”. The co-ordinates were telephoned through to the other sifting offices, each of which had its own punch co-ordinate machine. This was a machine which was developed by the promoters solely for the purpose of creating sifting masks. The co-ordinates were entered in the machine and it produced a number of sifting masks, which were simple sheets of Perspex with circular holes cut in them at the location of the spot chosen by the panel. They had marker lines to align with the edges of the coupon. Each sorting office cut about 100 sifting masks for the “first sift”. Each sifter was given a sifting mask and a batch of coupons and checked each coupon to see if a cross was visible through the hole in the sifting mask. If a cross appeared and was roughly in the centre of hole, then the coupon was passed on to the next sift. If not, it was discarded.

31. This process was repeated for a second time, using a sifting mask with a smaller hole. This was called the “second sift”. The second sift was usually complete by the end of Saturday afternoon.

32. By the end of the second sift, there were normally about 3,000 coupons across all of the sifting offices which had not been discarded. These potential winning coupons were then all taken to the Littlewoods head office for a third and fourth sift, which usually started on Sunday morning.

33. The third and fourth sifts followed a similar process to the first and second sifts, with the hole size in the sifting mask reducing each time. Magnifying eyeglasses were used as absolute precision became more important. By the end of the fourth sift (which was usually completed on Sunday afternoon) there were generally about 300 coupons left which had not been discarded.

34. In order to identify the winning coupons from this remaining 300 or so, four or five junior managers from Littlewoods would examine each of those coupons on the six Final Mark machines that were kept there. For each coupon, they would use the joystick to move the cross hairs to the centre of the nearest cross to the panel’s chosen location, then the machine would calculate the deviation between that location and the centre of the participant’s cross. The deviation on each coupon was printed out and attached to it. Each coupon carried printed index marks which would also be registered by the machine, which was programmed to apply a correction for any shrinkage or stretching of the coupon (as measured by reference to the index marks).

35. This process was generally completed by the end of Sunday. On Monday the chairman of the panel would come back to the head office and he was presented with the top coupons, from which the deviation measurements had been detached. An independent accountant would attend, to oversee the integrity of the process. An assistant manager and the panel chairman would re-measure the deviation on the top coupons and the chairman would rank the coupons in order of deviation, signing the winning coupon to identify it as the closest.

(b) Process from 2002 to date

36. The sifting process has been replaced by computer scanning and calculation. Once the panel have selected their location for the centre of the ball (“the target”) on Saturday and the Final Mark machine has identified its co-ordinates (all in the same way as before), all the incoming coupons (which have already been pre-scanned by computer) are checked.

37. First, all the coupons are re-aligned and resized, if necessary, against the theoretically accurate size and alignment which are programmed into the system. We are told that the scanning process itself inevitably involves some “noise”, as a result of which the scanned image may be as much as 3 pixels (0.255 mm) “out” from the original coupon. Although this error sounds small, it is potentially very large when compared to the accuracy of the subsequent calculations – see below.

38. The software identifies for closer consideration every coupon which has some kind of mark in an area 18 pixels by 18 pixels, centred on the target. As the distance between pixel centres is approximately 0.085 mm, the area searched is 1.524 mm x 1.524 mm. In spite of the small size of this area, in a sample occasion from 2011 approximately one quarter of all the coupons submitted had a mark in this area.

39. The computer then identifies the cross which is closest to the target on each of those coupons and calculates the location of the centre of that cross.

40. It is important to remember that the scanning equipment works on a digital and not an analogue basis. It breaks up the picture which it sees into a very large, but finite, number of individual elements and assigns a location and a value to each of them. Each scanned image is therefore made up of a large number of individual elements, known as dots or pixels. The scanning equipment uses a combination of 300 dots per inch optical resolution and software enhancement using a sub-pixel resolution technique. This results in the system mathematically recognising 1200 dots per inch (each dot measuring approximately 0.021 mm across, with 47.24 dots per mm), though optically it only “sees” 300 dots per inch. Given the size of a standard coupon, the system therefore effectively divides each coupon mathematically into some 31.5 million unique spots.

41. The precise method by which the system, within this framework, locates the centre of each cross was not made clear, indeed it would probably require a deep understanding of the enhancement software used to determine the “x” and “y” values for that location. Suffice it to say, however, that once that location has been determined, the system uses floating point arithmetic to calculate the distance between that location and the target. That distance is apparently rounded up to the nearest 0.001 of a pixel (equivalent to 0.000085 mm).

42. Once the top ten winning coupons are identified, they are presented for verification. First a manager checks that the correct fee has been paid for the number of crosses entered, then a visual check is made to ensure that the centre of the cross has been correctly calculated by the system (errors are not unknown, especially where manually drawn crosses are involved). Once this has been done, the top ten entries are presented to the chairman of the panel for him to agree the winner.

43. During a sample taken in 2006 over four weeks, the centres of the winning and other top ten crosses were calculated as falling in the following ranges of distances away from the target:

Position	Closest – pixels (approx. mm)	Furthest – pixels (approx. mm)
1	0.030 (0.003)	0.062 (0.005)
2	0.053 (0.004)	0.081 (0.007)
3	0.078 (0.007)	0.112 (0.009)
4	0.081 (0.007)	0.137 (0.012)
5	0.101 (0.009)	0.145 (0.012)
6	0.111 (0.009)	0.151 (0.013)
7	0.118 (0.010)	0.178 (0.015)
8	0.123 (0.010)	0.191 (0.016)
9	0.145 (0.012)	0.191 (0.016)
10	0.153 (0.013)	0.199 (0.017)

44. To set this information in context, we were informed that a piece of ordinary paper is approximately 0.1 mm thick, and in another sample taken on one occasion in 2011 (by which time volumes of coupons were obviously very much smaller than in the early 1990s when STB was in its heyday), it was found that the winning cross was calculated to have its centre 0.0057 mm from the target and there were 56 coupons which carried crosses with centres calculated to be less than about 0.0595 mm from the target. We were also shown a massively magnified representation of a very recent competition in which we were told a small circle around the target with a radius of 0.07 mm encompassed some 210 crosses on coupons.

45. The Appellants make the point that this post-2002 process actually represents a drop in accuracy compared to the previous visual measurement, where an accuracy of one thousandth of an inch (or 0.0254 mm) was achievable without the mathematical intervention of converting a 300 DPI scan into a notional 1200 DPI one – the mathematical interpolation of the extra notional dots may well not be wholly accurate when compared with actual optical resolution at the higher level (indeed, there was some evidence to the effect that their mathematical modelling for upgrading the scan results from 300 DPI to 1200 DPI may have given results which the National Physics Laboratory considered to be inaccurate). In addition, we were informed that the scanning process itself was subject to an error of up to 3 pixels (see above) – a huge potential error in the context of the final winning margins.

46. However, the post-2002 scanning system was considered fairer because it removed the human subjectivity of the previous sifting method and it also excluded human error in the sifting process, as well as all other variable errors that could arise in the course of it. All coupons were essentially treated exactly the same.

47. And it could not be said that the pre-2002 process was free from inaccuracies either, even if it were operated completely without human error. One illustration of the possible problems inherent in the previous process was identified in March 1994, when a detailed report was produced into the reasons behind an observed error of up to 0.2 mm in the placement of the holes cut in sifting masks in 1992. This was in the context of a process in which the size of the hole in the fourth sift sifting mask was itself only 0.5 mm in diameter. It was identified that the mathematical formulae built into the Final Mark machine’s software for adjusting to deal with variable coupon size (e.g. because of printing errors or expansion due to moisture) were flawed. They were corrected in June 1994.

48. Finally, on the subject of measurement generally, it was confirmed to us that there was not believed to have been any occasion during the life of STB when a participant had marked his cross on precisely the centre of the target – though in 2004 they had awarded the “jackpot” £250,000 prize for doing so as a promotional effort. The bare fact is that the level of supposed precision with which accuracy is judged makes it to all intents and purposes impossible to mark a cross in precisely the correct spot.

Experimental evidence

49. Mr Dent gave evidence of some experiments he carried out during the course of 2012 which were designed to demonstrate the importance of chance as a factor in the outcome of STB.

50. In the first experiment he described, he asked 20 employees to participate in a “dummy run” of STB, using a real coupon. These coupons were then “marked” in the usual way. He then showed the employees precisely where the centre of the ball had been judged to be by the panel, then asked them to try again, using further blank coupons. His aim was to simulate (in the second attempt) the “superbly skilled competitor”. He arranged for the dummy coupons to be compared with the real coupons for that week’s game. His hypothesis was that if skill was a large component of success, then the dummy participants should, after seeing the actual position of the target, have got as close or closer to it than the general body of participants (whose skill level, however high, could not be better than the actual knowledge of the correct location possessed by the dummy participants). On the other hand, if luck was much more important than skill, then he would expect the 20 dummy participants to score poorly by virtue of being statistically swamped by the much larger numbers (in this case, many thousands) of “real” participants.

51. On the basis of the “before” coupons, the top five of the dummy participants would have been ranked at 7^th, 21^st, 139^th, 2,480^th and 2,700^th in the real competition. On the basis of the “after” coupons, the top five dummy participants would have been ranked at a rather lower 162^nd, 370^th, 1,082^nd, 2,431^st and 2,698^th. Overall, 12 of the 20 dummy participants improved the accuracy of their “best guess” after seeing the correct position of the target. 7 dummy participants achieved a worse result for their best guess after seeing it, and one participant was so far away on both attempts that it was not possible to measure the difference accurately.

52. Mr Dent acknowledges this was a small sample and his interpretation of the results was not informed by any expert statistical analysis or opinion. He made no sweeping claims about the significance of the results, merely observing that in his opinion they demonstrated that “there is an element of luck inherent” in playing STB.

53. He also conducted another experiment, in which the winning co-ordinates of the target on a particular coupon were changed by the smallest measurable amount and then the coupons were re-ranked by reference to this revised target. The result was that the real winner would have been relegated to 9^th, the runner up would have been 15^th and the actual 10^th placed coupon would have been promoted to be the winner. In his submission, this also demonstrated (when considered in the light of the inherent element of chance in the accuracy of the measurements) that there was a “large element of chance” involved in STB.

54. We do not read a great deal into the outcome of Mr Dent’s experiments. Their outcomes demonstrate little more than can already be ascertained by simple logic, namely that very small changes of position of the target, when large numbers of closely spaced points are considered, can have a major effect on their order of proximity to that target. If we had sought to place any greater weight on them than that, we would have required expert statistical evidence to assist us.

1978-79 correspondence with HMRC

55. We were also provided with copies of some correspondence between HMRC and various interested parties (including Ladbroke Group Limited) in 1978-79, at a time when HMRC had first decided that STB should be standard rated rather than (as they had previously accepted) exempt. Many of the same arguments as were raised in that correspondence were rehearsed in this appeal. For the purposes of his decision, we considered that correspondence to be of historical interest only. We reject Mr MacNab’s submission that we should give some unspecified evidential weight to it for the purposes of this appeal.

The legislation

56. At all times that are material for the purposes of this appeal, the relevant EU law provision was that contained in Article 13B(f) of EC Directive 77/388, which provided as follows:

“Without prejudice to other Community provisions, Member States shall exempt the following under conditions which they shall lay down for the purpose of ensuring the correct and straightforward application of the exemptions and of preventing any possible evasion, avoidance or abuse:

.....

(f) betting, lotteries and other forms of gambling, subject to conditions and limitations laid down by each Member State;”

57. There is no issue between the parties as to the application of this provision; it is common ground that the relevant UK legislation from time to time was consistent with it.

58. Throughout the period covered by this appeal up to (and including) 31 October 2006, the relevant statutory provision (contained successively in Group 4, Schedule 5, Finance Act 1972, Group 4, Schedule 6, Value Added Tax Act 1983 and Group 4, Schedule 9 Value Added Tax Act 1994) provided exemption from VAT for:

“1. The provision of any facilities for the placing of bets or the playing of any games of chance.”

59. From 1 November 2006, this was amended to read:

“1. The provision of any facilities for the placing of bets or for the playing of any games of chance for a prize.”

60. Until 31 October 2006 the successive provisions all defined “game of chance” for these purposes as having the same meaning as in the Gaming Act 1968. The definition in section 52 of that Act remained unchanged from 1968 up to 31 October 2006, as follows:

“(1) ....“game of chance” does not include any athletic game or sport, but, with that exception, and subject to subsection (6) of this section, includes a game of chance and skill combined and a pretended game of chance or of chance and skill combined;

....

(6) In determining for the purposes of this Act whether a game, which is played otherwise than against one or more other players, is a game of chance and skill combined, the possibility of superlative skill eliminating the element of chance shall be disregarded.”

61. From 1 November 2006, the definition of “game of chance” was moved into the Notes to Group 4, Schedule 9 Value Added Tax Act 1994, which reads as follows:

“(2) “Game of chance” –

(a) includes –

(i) a game that involves both an element of chance and an element of skill,

(ii) a game that involves an element of chance that can be eliminated by superlative skill, and

(iii) a game that is presented as involving an element of chance, but

(b) does not include a sport.

(3) A person plays a game of chance if he participates in a game of chance –

(a) whether or not there are other participants in the game, and

(b) whether or not a computer generates images or data taken to represent the actions of other participants in the game.

(4) “Prize” does not include the opportunity to play the game again.”

62. On this basis, the parties were agreed that the sole question for decision at this hearing was whether STB as described above was a “game of chance” within the successive meanings of that phrase. HMRC’s position was that (a) STB was not a “game” at all, but if it was, then it was a game of skill. The Appellants’ position was that STB was undoubtedly a “game”, and it was a “game of chance and skill combined” (or, after 31 October 2006, it was a “game that involves both an element of chance and an element of skill”).

The case law and related discussion

Is STB a game?

63. Mr MacNab referred us first to DPP v Regional Pool Promotions Limited [1964] 2 QB 244. That case was concerned with an appeal against a conviction under the Betting and Lotteries Act 1934 for running an illegal lottery. The way that Act was structured afforded a defence if it could be shown that the activity in question amounted to “gaming” under section 21 of the Betting and Gaming Act 1960. In order to do so, it had to amount to “the playing of a game of chance for winnings in money or money’s worth.”

64. The activity concerned was as follows. Upon joining the Friends of Spastics League, an individual was given a membership card which included four pairs of digits. Various “games” were then run by the promoter, based on those digits. The detail of those games is not material, the important feature of them was that the members were entirely passive, all that was done was that various events took place which resulted in particular members being selected as winners, by reference to the numbers on their membership cards. The winnings were paid out from the weekly subscriptions that were received from the 6 million members of the League.

65. The prosecutor argued that the subscribers were not involved in a “game” and therefore they could not be “playing a game”, thus the 1960 Act exemption could not apply. The High Court held this was correct. Their reason was that a game necessarily involved the player doing something – some “active participation” – and in this case the so-called player did nothing at all apart from pay his weekly subscription once he had joined the League.

66. The prosecutor had put an alternative argument, as follows. He submitted that it was “of the essence of the playing of a game that the players should come together at a particular place at a particular time”. The Court rejected this submission, but in doing so there was some suggestion that a “game” should, at the very least, involve some means of communication between the players. As Parker C-J put it (at p253):

“I can envisage that it is possible for players to play a game without being physically together in one room or at one time, provided, of course, that there are means of communication.”

67. Mr MacNab invited us to apply this proviso in considering the present case.

68. Along the same line of argument, he referred us to Armstrong v DPP [1965] AC 1262, a decision of the House of Lords. This was again concerned with a prosecution for an allegedly unlawful lottery under the 1934 Act, and the same defence under the 1960 Act was potentially available.

69. In that case, the activity in question could best be described as “postal bingo”. Participants joined by applying and paying a small membership fee. They would then receive a membership card which included their membership number and also a selection of 15 numbers between 1 and 90. They could request particular combinations of numbers and, unless they had already been allocated to someone else, they would receive the requested combination. Otherwise, numbers were allocated at random. They could obtain more than one combination of numbers by requesting the issue of further cards, linked to their membership number. Each week, the members decided which of their cards they wished to enter for the draw, and they paid a fee of 3s 6d per card (1s as a “service fee” and 2s 6d as stake money). Of the 300,000 members, only about 100,000 entered in any particular week. There was a weekly draw and prize winners received cash prizes without needing to claim.

70. The House of Lords held that nothing being done by the participants could amount to “playing a game”. They held that the question was primarily one of fact and in the circumstances they expressed the view that there was no evidence in that case which could have supported a finding of fact that a game was being played.

71. They went on to consider a particular argument that had been put forward by the prosecutor, to the effect that “an assembly of players” was required before there could be a “game”. On this submission, Lord Pearson (with whom all the other Law Lords agreed) said this:

“Undoubtedly that is the normal way of playing a game, and there are provisions in this Act.... which contemplate that a game will be so played....”

72. He went on to say, however, that the statutory context effectively left the question undecided. He was not prepared to base his opinion on this argument, and the most he would say in support of it was that:

“... the facts that there is no assembly of players, and that the alleged players are not in each other’s presence nor in communication with each other, may well have considerable weight in any case as evidence in favour of a more general argument that there is no playing of a game.”

73. Next, Mr MacNab referred us to Adcock v Wilson [1969] 2 AC 326. That case was concerned with a different offence, namely under section 32(4) of the Betting, Gaming and Lotteries Act 1963. It involved a variant on normal bingo, under which the participants at a normal bingo club played a special “game” in which one half of the stake money, instead of being shared out as winnings between the players who were present at the particular bingo club, was allocated to a national “game” which involved all players in all participating bingo clubs (some 500 in number). In the national “game”, the stake money was allocated in various ways to individual players at the separate clubs, for example to the person or persons who achieved a full house in their local game on the minimum number of calls. The crucial point in the appeal was whether any of the stake money was “disposed of otherwise than by payment to a player as winnings”.

74. The meaning of “player” was defined in section 55 of the 1963 Act as follows:

‘"player", in relation to a game of chance, includes any person taking part in the game against whom other persons taking part in the game stake, play or bet;’

75. Lord Morris, giving the unanimous decision of the House of Lords, took the view that there was no “national game” as argued by the organisers. Once the local game had concluded at each venue, the national pool of stake money was simply allocated according to the rules. There was nothing more for each participant to do, and “as there was nothing further to be done there was nothing further that could be called a game.” It followed that the people in any one bingo club were not “players” in relation to any game with the people in other bingo clubs; therefore the relevant part of the stake money, when paid over to the national pool, was “disposed of otherwise than by payment to a player as winnings” and therefore the offence was committed.

76. Mr MacNab referred us to passages in the judgments of Ashworth J and Widgery J in the High Court, where Ashworth J expressed the view that taking part in a game necessarily involved there being “some means, so to speak, of identifying a competitor and knowing who is taking part” and Widgery J commented that it was:

“...exceedingly difficult to produce those elements which the common man would ascribe to a game if you have the participants in separate places with no communication between them whilst the activity is going on, and thus no sort of opportunity of seeing how their competitors are progressing...”

77. By reason of the lack of any means for a participant to identify the other participants, to interact with them or to monitor their progress, Mr MacNab submitted, on the basis of the above authorities, that STB was not a “game”.

78. Mr MacNab also sought to persuade us that STB ought more properly to be described as a “competition” rather than a “game”. He acknowledged that one activity could amount to both a “competition” and a “game” but he submitted that was not the case here – mainly because of the absence of the “interaction” that he argued was necessary for a game.

79. Our analysis of the cases is as follows. First, we read Regional Pool Promotions and Armstrong as establishing little more than that some degree of active participation is inherent in the concept of a “game” (a proposition that both parties in this appeal agree upon, and a feature which they both agree is present in STB). Beyond that, the only principles we extract from them are (a) that the question of whether an activity amounts to a “game” or not is primarily a question of fact and (b) that there is no rule of law that requires an assembly of players in order to constitute a “game”.

80. Turning specifically to Adcock, we consider it to be unpersuasive in the present case. The bingo players were all clearly playing in their local game of bingo. The question was whether there was also a national game involving all of them. It seems to us that the general observations made in both the High Court and the House of Lords about the nature of a “game” must be understood in the context they were made, that is to say a discussion about bingo. In a game of bingo, a crucial part of the game is the interaction with the caller and the other players, so it is not hard to understand why the view was taken that the absence of that interaction, together with the lack of any requirement to do anything more than participate in the normal way in the local game, should mean that there was clearly no national game.

If it is a game, is STB a game of chance or a game of skill?

81. In considering this question, Mr MacNab submitted that we should draw a distinction between the rules of the competition (in accordance with which the competitors were entitled to have their entries judged) and the internal procedures that were actually followed by the promoters of STB in deciding the winners. Under the rules, the winner was to be the most skilful competitor (as judged by the panel) and the method by which the promoters chose to reach the correct result was an entirely irrelevant matter. As he put it, the rules made it clear that STB was:

“a game of skill, in which the skill, judgment and logic of the competitors is judged ex post facto by the judges (who are also required to exercise their own skill and judgment); and not a game of chance, where competitors guess what third parties might decide in the future. The competition is one of logic, in which the estimated position of the ball is to be determined by reference to “all of the information provided”.”

82. Mr MacNab pointed out that STB was always held out to the public as a competition of skill. The banner heading on the coupons generally advertised it as a “Football Skill Competition”, competitors were almost always invited to use their “skill and judgement”, and they were informed that the winner would be the person whose entry was adjudged to be “most skilful”, and/or whose entry most accurately showed the “logical position” of the ball (as determined by the panel).

83. In the circumstances, he said, it was correct for us to accept this as an accurate description of STB for what it was – a competition of skill – and there was no need (indeed it would be inappropriate) to consider in meticulous detail the process by which the most skilful participant was selected by the promoters. That was an entirely peripheral matter that did not affect the essence of the activity as being one of skill.

84. Mr Peacock, on the other hand, drew attention to the fact that it was impossible to consider the true nature of the activity without considering it in proper detail. From an examination of that detail, it was completely obvious that skill – even superlative skill – would only take you so far in the competition. This fact was clearly understood by the participants – otherwise why would they pay extra entrance fees for the right to place large numbers of crosses on their coupons? And why would the promoters include invitations on their coupons such as “Enter more X’s for more winning chances” and promotional language such as “This could be your lucky day”? Mr Peacock acknowledged there was certainly some degree of skill involved, but he submitted that the evidence made it clear there was also some degree of chance.

85. So was STB, as Mr Peacock submitted, a “game of chance and skill combined”?

86. The parties agreed there was no directly relevant case law to help with this question, but submitted there were some pointers in two reported cases.

87. The first of these two cases was Ladbrokes (Football) Limited and others v Perrett [1971] 1WLR 110. This was a decision of the High Court and was concerned with an appeal against a criminal conviction arising from the promotion and circulation of an STB competition.

88. The variant of STB was quite different from that involved in the present appeal. A coupon was issued from which the ball had been removed and instead a grid of 40 boxes had been printed. Competitors were invited to place crosses in the boxes in which they considered it was most likely that the missing ball would have been. For the minimum entrance fee, they were allowed to place eight crosses. For higher fees, they could place up to 13 crosses.

89. Once all the coupons had been received, a panel of expert footballers decided which eight squares the missing football was most likely to occupy, and the coupons were then compared with the panel’s selection. The coupons on which the correct eight squares had been marked were the winners.

90. The promoters were prosecuted for offences under section 47(1)(a)(i) and 47(1)(b) of the Betting, Gaming and Lotteries Act 1963, which provided as follows:

“(1) It shall be unlawful to conduct in or through any newspaper, or in connection with any trade or business or the sale of any article to the public –

(a) any competition in which prizes are offered for forecasts of the result either –

(i) of a future event;

(ii) .....

(b) any other competition success in which does not depend to a substantial degree upon the exercise of skill.”

91. The Magistrates had acquitted the promoter of any offence under section 47(1)(b). This decision does not appear to have been appealed, though the High Court observed in passing that the acquittal was correct “because there must be a considerable element of skill in this competition.”

92. In confirming the conviction under section 47(1)(a)(i), the High Court held that there was a future event (namely the deliberation of the panel of experts) and a result (namely the outcome of that deliberation) and the participants were effectively attempting to forecast the panel’s decision.

93. This latter reasoning was strongly disapproved by the House of Lords by a four to one majority in the second case the parties referred us to, News of the World Limited v Friend [1973] WLR 248.

94. Friend was concerned with the same variant of STB as the present appeal. In Friend, the prosecuting authorities had only prosecuted under section 47(1)(a)(i) of the 1963 Act. Their Lordships (Lord Simon dissenting) held in effect that it was unduly artificial to analyse the competition as a “forecast of the result of a future event”. The participants were simply using their skill and judgment to determine the most logical location for the centre of the missing ball. Thus no offence could have been committed and the appeal was therefore allowed. Their Lordships (insofar as they addressed the point) expressly left undecided the question of whether a conviction under section 47(1)(b) would have been upheld – for which the relevant question would have been whether success in the competition depended “to a substantial extent upon skill” (a different question from that under consideration in the present appeal).

95. We were referred to a number of passages in the judgments in Friend in support of the contentions of both parties. In deference to the fact that Friend was at least concerned with the same version of STB as the present appeal and both parties referred to it at some length in argument, we deal at some length with the submissions made by the parties on various statements in it; overall, however, we should make it clear that we did not find much assistance in any of the passages to which we were referred. When set in the full context of the decision, many of the passages both parties tried to cherry pick from it either did not carry the weight sought to be placed on them, or were either (a) only ambiguously supportive of or (b) downright contradictory to the arguments the parties sought to base on them.

96. First, Mr MacNab invited us to decide the question before us by reference to the rules of STB itself, rather than the detailed processes and mechanisms by which the promoters arrived at a winner in accordance with those rules. Effectively he was submitting that under the rules, the only decisive factor was how skilful the participant was in placing his cross and we ought therefore effectively to disregard any element of chance that might in fact determine the final winner.

97. In support of this proposition that we should focus on the rules rather than the underlying processes employed to chose a winner, he referred us to the following passage in the judgment of Lord Hailsham (at p254D, emphasis added):

“The question, and the only question, which arises is whether, in these circumstances, the prizes were being offered for a forecast of the result of a future event within the meaning of section 47(1)(a). I agree myself wholeheartedly with the proposition submitted on behalf of the respondent that in these cases the court will look at the realities of the offer and the competition and will not allow itself to be deceived, whether innocently or otherwise, by delusive appearances or descriptions. But in this case I can see no difference between appearance and reality, and, with respect to the submission of counsel, none was suggested, that is, I cannot find anywhere any suggestion that the offer was intended to convey to the competitor anything other than what was said, or that the competitor did, or thought he was doing, anything other than what he was invited to do in the offer, or that the competition was in any way conducted or judged in a manner different from that which appeared on the surface.”

98. The context for this passage was that his Lordship was analysing whether a participant in panel STB could really be regarded as attempting to forecast the result of the panel’s deliberations, or whether he should simply be regarded as doing what the rules said, namely trying to specify the most logical position for the missing ball. Therefore, the most we glean from this passage is that one should consider the realities of the competition without regard to any “window dressing”.

99. Mr MacNab also sought to persuade us that it was correct to disregard peripheral matters such as the precise measurement processes by reference to Lord Hailsham’s decision at 256A, where he said:

“I am not at all impressed by the fact, which I accept as accurate, that the exact pinpointing of the centre of the ball “to the last millimetre,” as it was put in argument, is, as MacKenna J said, to some extent a matter of chance, if, as I suppose, the most that skill and judgment can do is to estimate its approximate position.”

100. On close examination, however, this passage appears in the context of his Lordship’s analysis of why the participant’s action should be seen not as a forecast of the panel’s deliberations but as an independent attempt to identify the most logical position for the ball. He was simply saying that the existence of an admitted element of chance in selecting the precise location does not affect the fact that the participant is still using skill and judgment to reach his own view of the correct location, rather than in trying to forecast where the panel would determine it to be. It is also somewhat telling that his Lordship referred to “the last millimetre” in a way which suggested that millimetric precision was enough. As we have seen in the present case, margins of one five hundredth of a millimetre can be the difference between the winner and tenth place.

101. Mr Peacock, referring to the same passage from Lord Hailsham’s judgment, said it in fact supported the Appellant’s case because it was an explicit recognition of the central role of chance in the game. To the same end, he referred us to the following passage in the (dissenting) judgment of Lord Simon (at 265D-E):

“skill and judgment and knowledge of football might enable the competitor to put his cross in approximately the right area (whether the area where the ball actually was or where the judges adjudged that it should be), but the final half millimetre which might make the difference between winning and losing must be purely a matter of chance.”

102. Finally, Mr MacNab referred us to the judgment of Lord Reid, who said at 258B-C:

It is, I think, a proper inference from the findings of fact in the case that their [i.e. the panel’s] decision did depend on skill. And the instructions to competitors were clearly intended to represent to them that the winning position would be that determined by the expert skill of the panel. I do not think that anything turns on the precise way in which the panel set to work or the precise way in which the entries were examined. It would be very surprising if it did.”

103. In this passage, his Lordship was in the middle of his analysis as to whether panel STB could be said to involve a forecast of the result of a future event (i.e. the panel’s decision) and he was commencing his explanation of why he considered that analysis to be wrong. In short, it was because any participant would be quite clearly aware that the panel would be using its own skill to decide where it thought the ball should be, and therefore in practice the participant would do the same thing rather than go through some artificial process of trying to second guess or “forecast” the panel’s decision.

104. Mr MacNab also sought to persuade us that in theory it would be possible, by programming a computer to apply the same logic and judgment as the panel, to deduce where the panel would place the ball. (This was subtly different from saying that a computer could be programmed to determine the real location of the ball – he acknowledged that there would not be enough information to enable that to be done.) He described a computer, so programmed, as the “analogue of the superlatively skilled competitor”, eliminating all chance from the process. The fact that most participants did not have that level of skill did not mean that the competition was any less a competition of skill (any more than a game of chess became a game of chance when played by an unskilled player). We regard his premise (that a computer could be appropriately programmed) to be speculative and unsubstantiated and therefore we do not find this line of thought instructive.

105. Mr MacNab developed this line of argument further, by reference to the Court of Appeal decision in R v Kelly [2008] EWCA Crim 137, [2009] 1 WLR 701, in which the interpretation of the phrase “game of chance” in section 52 of the Gaming Act 1968 was specifically considered in the context of “Texas Hold ‘em Poker”.

106. In that case, where it was accepted that the element of skill involved in the game was significantly greater than the element of chance, the appellant sought to persuade the Court that the preponderance of skill meant that the game could not be regarded as a “game of chance”, even though that phrase was explicitly extended by the statute to include a “game of chance and skill combined”.

107. The Court of Appeal disagreed. The central part of the judgment reads as follows (at 711E-F):

“In our view, as Parliament has provided that games of combined skill and chance are to be treated as games of chance without any qualification, then the only circumstances where chance should not be taken to make a game of skill and chance a game of chance is where the element of chance is such that it should on ordinary principles be ignored – that is to say where it is so insignificant as not to matter.”

108. Mr MacNab’s assertion that Kelly supported HMRC’s case depended upon his submission that there was a crucial difference between the poker in Kelly and STB, namely that any player of poker will know that there is an element of chance involved, whereas STB was held out as a game of skill, involving no necessary random element. In effect, he appeared to be arguing that because (a) a notional computer could theoretically be programmed to arrive at the same expected ball position as the panel, and (b) the activity was held out as only requiring skill, we should “on ordinary principles” ignore such element of chance as existed.

109. Mr Peacock on the other hand invited us to recognise the realities of the situation. However skilful a participant was, he or she could not, by skill alone, win at STB. In the words of Lord Hailsham LC in News of the World, “the most that skill and judgment can do is to estimate its [i.e. the ball’s] approximate position.”

Discussion and conclusion

Is STB a game?

110. We do not consider that the case law provides us with a great deal of assistance in answering this question – see our analysis at [79] to [80] above.

111. In addition, it must be remembered that there are quite specific provisions contained in section 52(6) of the Gaming Act 1968 (and latterly in note (3) to Group 4 of Schedule 9 Value Added Tax Act 1994) which were absent in all of the cases referred to at [63] to [77] above. Those provisions expressly contemplate the possibility of a game being played “otherwise than against one or more other players” or “whether or not there are one or more other participants”. In that context, we consider it is quite simply unsustainable to argue that an activity cannot be a game within the meaning of section 52 of the Gaming Act 1968 (or Note (3) to Group 4 of Schedule 9 Value Added Tax Act 1994) unless it includes the features of identifying, interacting with and/or monitoring the progress of other participants.

112. If we find the cases unhelpful, we must fall back on general principles of construction.

113. Like many other words, the word “game” is a chameleon. It takes its colour from the context in which it is used. It has numerous “ordinary meanings”, as highlighted by the entry from the Shorter Oxford English Dictionary which was put before us. Ignoring for a moment meanings which relate to wild animals pursued with guns or rods, it can mean “amusement, fun, sport”, “amusement, diversion”, “a diversion in the nature of a contest, played according to rules, and decided by superior skill, strength or good fortune”.

114. We do not consider that an activity must involve more than one person in some kind of interaction before it can be a “game”. It is normal, for example, to refer to a “game of patience”, which activity involves only the player and a pack of cards. We discount as unduly artificial Mr MacNab’s argument that this is because there is an element of “interaction” in such a game, namely an interaction between the player and the randomness of the cards.

115. In seeking to explore the boundaries of the concept of a “game”, we also considered “puzzles” (involving the application of skill or logic to arrive at a single correct solution, such as in a crossword puzzle or Sudoku), “pastimes” (involving activities of many kinds whose main purpose is to spend time pleasurably) and “competitions” (in which a participant pits himself against another participant or participants with the purpose of achieving victory). It became readily apparent that such consideration did more to illustrate the vagueness of the concept of “game” than it did to clarify it.

116. In Oasis Technologies (UK) Limited v HMRC [2010] UKFTT 292 (TC) at paragraph [65], the First-tier Tribunal said (when considering the meaning of the word “game” for the purposes of determining whether a particular activity was a “game of chance” for the purposes of section 6 of the Gaming Act 2005):

“There is no definition of “game” in section 6, and so we must construe this term according to its ordinary meaning. There is no single meaning that can be attached to this term. According to the Shorter Oxford English Dictionary, it can variously be regarded as meaning an amusement, fun or sport, or as meaning a diversion, whether or not one in the nature of a contest played according to rules and decided by superior skill, strength or good fortune. We consider that this demonstrates that “game” has a wide meaning, to be construed according to its context”.

We respectfully agree (subject to the inclusion of the missing comma after the word “contest”).

117. In the light of all the above, and adopting the approach of the First-tier Tribunal in Oasis, when considered in the context of section 52(1) Gaming Act 1968 or of Note (3) to Group 4, Schedule 9 Value Added Tax Act 1994, we consider it perfectly apt to refer to the activity of STB as a “game”.