Horse versus machine: battles in the betting shop

Changes in regulation and taxation during the past decade have had a profound effect on the experience of betting in dedicated shops in the United Kingdom. This article explores how betting shop customers and staff in London have responded to the introduction of gambling machines depicting roulette and other casino-type games in an environment that was traditionally dedicated to betting on horse and dog racing. The rise of machine gambling has been presented as a transition from ‘social’ to ‘asocial’ forms of gambling by researchers working in the UK, Las Vegas, and Australia. Traditional bettors and betting shop staff also present betting and machine play as discrete and value them differently. I show that while some of the experiential qualities of machine play observed elsewhere have been replicated in shops, the differences between traditional betting and machine play are overstated, for structural reasons. Traditional bettors and staff are interested in distinguishing their activities from those of newcomers. Responses to new gambling media in betting shops are socially, as well as experientially, mediated, a crucial insight for the wider study of gambling and gambling regulation.

Introduction: the gambling industry in the UK In the past decade, profound changes in taxation and regulation have transformed the UK gambling industry. The Gambling Act 2005 replaced ad hoc legislation, the majority of which dated from the 1968 Gaming Act, which was designed to rid the casino industry of criminals. Arcane rules based on the low-technology environments of bricks-and-mortar casinos, betting shops, and members' clubs proved inadequate in a new era of diverse offerings, including remote gambling. Perhaps most importantly, in 1994, the Conservative government introduced the National Lottery and the state became a participant in the gambling industry as well as (indirectly) responsible for its regulation (Miers 2004). Paternal legislation aimed at limiting the growth of markets for gambling seemed inconsistent at best when mass participation in the National Lottery was encouraged by advertisements inciting potential consumers with the tantalizing suggestion that 'It could be you!' Regulatory changes were not only a reaction to changes in the gambling industry but also an explicit attempt by successive governments,Conservative and Labour,to make the UK an attractive base for the global gambling industry. During a series of lengthy consultations, beginning with the Budd Report in 2001, gambling was re-cast as entertainment, an important part of the leisure industry and 'UK plc' more generally. Tessa shops, and in the six months before October 2005 FOBs generated 89 per cent of profit in shops belonging to Stanley Leisure (Stevenson 2005). The annual reports of gambling operators show that gross gaming win by machines is continuing to grow as OTC declines. 3 I begin by describing betting shops in London (also known as Licensed Betting Offices, LBOs, turf accountants, or bookies) before comparing the two kinds of betting and the forms of engagement and meaning-making they facilitate. I show that some aspects of the transformations described elsewhere and particularly in the United States and Australia are reproduced in the UK. In addition, I consider structural differences between traditional bettors and machine players, showing how particular technologies are refracted through the betting shop as a working-class space in cosmopolitan London. I show how divisions between newcomers and old-timers resurface in new guises, including in the distinction between new and different kinds of betting. In this sense, I use fieldwork data to resist the idea that technology necessarily leads to gambling that is asocial, let alone less socially meaningful.

Bookies
When they first opened in 1961, betting shops were required by law to be austere in order not to stimulate demand for a working-class activity that was regarded as morally and economically dysfunctional (Dixon 1991). This did not prevent their rapid growth, as they tapped into the established practice of illegal betting on streets and in factories (Chinn 2004). In 1962, The Sunday Telegraph devoted special features to what it referred to as 'Britain's newest, least documented and in many ways most depressing growth industry -the Betting Shops ' (9 September 1962). By 1966, there were 16,000 betting shop licences in Britain, and Labour Chancellor Jim Callaghan introduced the first betting tax. In 2007, alarmed by the implementation of new licensing practices contained in the Gambling Act of 2005, the Evening Standard warned of a 'betting shop boom in London' declaring that 'London is suffering an explosion of betting shops, further fuelling fears that gambling is spiralling out of control ' (30 April 2007: 11). In 2009, following changes in the market, regulation and taxation, there were 8,862 such shops in Britain, 67 per cent of which were owned by companies known colloquially as 'The Big Three' (Ladbrokes, William Hill, and Coral) (Gambling Commission 2011a). 4 Non-remote betting is by far the largest sector of the British gambling industry, as regulated by the Gambling Commission, accounting for 53 per cent of £5.7 billion annual gross gambling yield in 2009-10. 5 Casinos are the next largest, accounting for 14 per cent, with bingo third at 12 per cent (Gambling Commission 2011a). The majority of my research participants are habitual betting shop users, called 'regulars' in the trade, who spend several hours in shops each day. The 2010 British gambling prevalence study identified those who spent more than seven hours a month (the top 10 per cent of their sample in terms of time spent gambling) as 'high time' gamblers and found that high-time only gamblers were more likely (compared with all regular gamblers) to be from semiroutine/routine households (39%), to live in the lowest income households (53%), and to have no educational qualifications (35%). They were less likely to be in paid work (40%) and more likely to be unemployed (7%). This group therefore consisted disproportionately of those with the poorest socio-economic indicators (Wardle et al. 2011: 65).
Twenty-seven per cent of 'high time' gamblers were retired and 60 per cent were not in paid work. Although there was considerable variation between and within shops, the majority of my research participants were either retired or not in paid work, with time, but not necessarily money, to spare. The regressive properties of taxes on working-class gambling, including lotteries in the UK and electronic gaming machines in Australia, have been illustrated in a comparative study by Pickernell, Brown, Worthington, andCrawford (2004) (see also Clotfelter 1993;Clotfelter & Cook 1987).
The prevalence study provides a national snapshot of gambling; however, betting industry professionals view the London market as distinctive, characterized by higher stakes, 6 a mobile population (a shop can be highly profitable one day and much less so the next), and greater competition for sites. 7 Development officers crave locations with high 'slip counts' (the number of bets taken each day) and big punters in groups or singly. Traditionally, the most valuable sites were near pubs which showed televised horse-racing (called 'wet sites'), often with a strong Irish contingent. During the 1980s and 1990s, sites near Chinese restaurants were favoured. Today, development officers identify the most lucrative markets for betting shops as areas of multiple-occupancy housing where traditional bettors, including Irish, Greek, Turkish, and West Indian, combine with more recent migrants from Africa and Eastern Europe. These are the demographics of the areas of central and southeast London where I conducted fieldwork. 8 I have worked with betting industry professionals, punters (customers), and staff in betting shops in England since 1996. In 1996, most shops I visited were small, dark, and pungent. In 2002, I wrote that in the betting shop, 'not smoking was frowned upon' (Cassidy 2002: 72). By the time I began work as a cashier in London in 2009, smoking had been banned and the shops were far more varied. The effects of changes in technology and regulation since the 1990s have been unevenly distributed. One of my local shops had to have the floor replaced because people had been urinating in the corner when the toilets were shut following a drugs raid. Victor Chandler's shop in Mayfair, on the contrary, resembles a gentlemen's club or casino.

'Mugs only'
On my first day working as a cashier in a central London shop I explained to Ian the middle-aged manager that I was interested in understanding how people used and thought about betting shops. He was enthusiastic and said that I would need to wait until the afternoon as 'nothing happens until the racing starts' . I looked across the counter to the shop floor, where half a dozen men were shouting and playing a FOB as he continued, 'Yeah. No one comes in till about lunchtime, then we can find you some people to talk to. Show you the ropes, that sort of thing. Shame no one's about just now' . While we had a cup of tea and waited for people to arrive, I pointed to the men in the shop and said, 'What about them? Is it okay if I go and chat to them?' Ian looked surprised and said, 'They don't know anything about betting. Better off waiting for the others' .
It is not just that betting and machine playing are different, as I was continually told by OTC bettors and staff, but that betting is rational and respectable, while machine playing is irrational and embarrassing. Mark, a 40-year-old regular bettor on long-term sick leave from his job as a council worker, was vehement, 'You will never catch me on those bloody machines. Strictly mugs only. That's a mugs-only zone over there. Be careful you don't catch muggery when you're over there talking to them' . The organization of the space in the shops where I worked reflected this perceived difference in interests and attitudes. As Ian told me: Every shop I've actually been in now that I've worked in it's like that, you know, the machine customers are separated they are by themselves they stick around the machine and the betting customers separately. Only now and again you might have someone who's betting come over like, if someone's winning a vast amount of money they come over to see, have a look, but he's not going to be here for long 'cos he's trying to do something on the other end.
Amit, a 30-year-old manager who had been working in southeast London for the past four years, described the situation in more graphic terms: In my shop we have a two-state solution and peace. We had war once! And I wanted peace, so I said to the machine players stop yelling and shouting and smashing up the machines and you racing men you stick to your corner and don't bother my friends over here.
Why should OTC bettors and staff choose to distinguish so firmly between OTC betting and machine playing? Is OTC betting skilful? Is FOBs play irrational? Does evidence from the betting shop support Griffiths's argument that 'technology is essentially turning gambling from a social pastime to an asocial one ' (1999: 279)?
Invitations by OTC bettors and staff to view machine playing as irrational and trivial evoke Geertz's distinction between 'status' and 'money' gambling on cockfights and 'plunging' on games of chance. Geertz's article on cock fighting in Bali in the 1960s has been criticized for a variety of reasons (Carrithers 1988;Hobart, 1986;1990;Shankman 1984). None the less, it provides a useful example of how different types of betting coexist and are differentially judged. According to Geertz, there are two kinds of bets or toh: between the two owners of the cocks set to fight (toh ketengah) and the 'cloud' of peripheral bets which take place between members of the audience (toh kesasi) (2005 [1972]: 65-6). Geertz's argument is that deeper fights (those with higher toh ketengah) correlate with 'status gambling' and shallower ones with 'money gambling ' (2005 [1972]: 73). Individuals express their equivalence to and unity with kin through collective bets at even money and reflect on the anti-social status of individuality by being reticent about peripheral bets. Even more strikingly, Geertz contrasts both toh kesasi and toh ketengah with the activities of peripheral gamblers, described as 'plungers ' , 'addicts' , and 'vulgarians' (2005 [1972]: 72). In Geertz's analysis, betting between cock-owners is 'meaningful' (2005 [1972]: 72) in a way that betting by the poor is not: There are, around the edges of the cockfight area, a large number of mindless, sheer-chance type gambling games (roulette, dice throw, coin-spin, pea-under-the-shell) operated by concessionaires. Only women, children, adolescents, and various other sorts of people who do not (or not yet) fight cocks -the extremely poor, the socially despised, the personally idiosyncratic -play at these games, at, of course, penny ante levels. Cockfighting men would be ashamed to go anywhere near them (2005 [1972]: 73).
It is not quite clear whether Geertz is describing the upper-class consensus, or expressing his own views, but he does not consider these peripheral activities in any more detail. Why should they be less revealing than betting on cockfighting? By betting this way, the least powerful eloquently evoke their social world by participating in games the odds of which they know to be stacked against them, for low stakes with people they know they cannot trust.
A similar division of labour can be found in recent studies of gambling. Table games are analysed as 'play' that is expressive of important social distinctions (Hayano 1982;Malaby 2003;Papataxiarchis 1999). Work which focuses on machines, by contrast, places more emphasis upon the interior states generated (Griffiths 1999;Schull 2005;Woolley & Livingstone 2009). Unlike Geertz, these studies relate the interiority generated by machine gambling to wider processes, including the unequal impact of gambling liberalization on different socio-economic groups (Kingma 2004;Layton & Worthington 1999;Pickernell et al. 2004). However, machine gambling is explored primarily in experiential terms. In the case of betting shops, as I will show, this way of describing machine play reflects the ideas of OTC bettors and staff rather than machine players themselves.
Long-term fieldwork with OTC bettors and machine players shows that the contrast drawn between the two activities by traditional bettors and shop staff is overstated. Firstly, betting OTC sometimes generates the kind of experience more commonly associated with playing machines. Secondly, betting on machines is not always asocial. Thirdly, neither betting OTC nor betting on machines is profitable in the long run, for similar structural reasons. The distinction between the two activities is not, therefore, based on the difference between skill and luck. Although there are professional horserace bettors making a profit through online betting exchanges, there is no such thing as a professional betting shop bettor. Despite being engaged in an activity that is not profitable in the long run, machine players may ameliorate the effects of short-term gambling losses and more general shortages of money not just by sharing winnings, but also by plugging into networks that provide access to loans and favours. Finally, machine players sometimes employ strategies that are similar to those used for betting OTC. In the next section, I will develop these four points in more detail. In the final section, I describe how the two kinds of betting are envisaged as opposites by OTC bettors and staff interested in presenting their activities as respectable in contrast to those of newcomers to the shops who 'give the bookies a bad name' .

Zoning out
The FOBs interface supports a kind of play that has been observed in other contexts, particularly among poker machine ('pokie') players in Australia (Livingstone 2001;Woolley & Livingstone 2009) and Las Vegas (Schull 2005). Bets are placed via a touch screen that depicts the roulette table. Once a player is satisfied with the bet, he or she presses 'spin' and the table is replaced by a graphic of a roulette wheel, around which spins a ball which eventually settles on a number. After each spin, the screen returns to the table. Play takes the form of a rapid tapping of the screen. Each tap produces a noise, and can be undone and redone before the bet is finalized. Though some players agonize over individual bets or taps, most do not. They simply push notes into the machine, sit down, and tap away. When I asked players to explain this technique, they described a world in which whether or not they win is decided elsewhere by an unknown force that is outside their control. As Donnie explained: You just feel the numbers and the machine. You need to block out everything else and just let your hand be guided by the future. Thinking stops that flow you need, right. It's there, you just need to tune into it. That's the hard bit, though (Winston, 18, machine player, southeast London).
This way of playing FOBs appears to share experiential qualities with 'getting into the zone' , achieving a Zen-like absence from everyday life by playing pachinko (Richie 2003), losing oneself in role-playing on-line games (Lee 2007), and playing the lottery in Havana (Holbraad 2007). Could this withdrawal into interiority (Livingstone 2005), an extreme form of self-denial, provide the basis for the rejection of this form of gambling by OTC bettors?
OTC bettors in my local shop referred to machine players by saying 'tap, tap, tap' while tapping the sides of their heads in a gesture used to indicate mental illness. They described machine players as 'zombies' , 'manic' , 'problem gamblers' , and, using the current terms of the debate about gambling in the UK, 'addicts' . 9 In keeping with this discourse, they were often scornful towards them, occasionally sympathetic, sometimes both at the same time, as when retired train driver Reg gazed over to the machines and said, 'Stupid twats. Can't help it, mind' . The absence sought by Donnie and Winston caused the greatest offence to those in the betting shop who chose to present betting as an intense intellectual engagement the rewards of which accrued to the individual. This ideal was maintained despite the wide range of practices it overlooked.
Certain styles of betting OTC have as much in common with machine play as they do with the work of betting undertaken by Reg and other regulars. Typical of these mechanical bettors are lunchtime customers who enter the shop and glance at the screens to see what races are available before writing a number on a slip and handing it over the counter. This process is repeated until their lunch break ends. These bets are unsystematic. They don't involve studying form and are placed without discrimination between media. The event frequency is lower than the bettor would experience on the machines, but racing functions as a results mechanism, in the same way as does a random number generator. This is how Jay described his betting behaviour when I approached him on a day off, also spent in the shop: Lunchtimes when I have a punt, I just pop in and bet on whatever I see. Usually just picking numbers, doesn't matter what racing is on, I just pick a number. Try my luck. I don't really think about it. I usually just want to forget about work! So it's a bit of a distraction. I get out of my head for a bit (Jay, 25, roadsweeper, southeast London).
Another punter, Frank, told me that if he was agitated he would 'bet a bit' to 'blow off steam' before joining his friends to discuss the racing, combining several styles of betting in one afternoon: Take yesterday, right. I come in 'ere, what, before lunchtime wunnit. South African racing. Perfect. I just sits down and starts punting. Numbers only. I look at it. Number four. Look again. Number two. Whatever. Emptying me head. From working nights and being woke up by next door. Then I get me head on and I come over 'ere with the lads and I talk about the racing (Frank, security guard and nightclub doorman, 50, central London).
As well as routinized, unreflexive betting of this kind which facilitates withdrawal, rather than engagement with the world, OTC bettors attempt to 'channel' , like machine players, by placing their bets as late as possible before the 'off' . This feature of betting is well known to all cashiers and particularly trainees, who sit and watch no one do anything for minutes at a time until twenty seconds before a race starts, when everyone runs to the counter waving his or her slip and (literally) shouting the odds. When I complained about this behaviour, Wally explained his desire to leave his bet until as late as possible, in a way that recalled Winston and Donnie's description of playing the machines: Why am I last on? Because that way I'm closest to the race. Any little hint I'm going to get from the race itself, see? I study form, but sometimes I get a feel for a race, and that's even better. The race tells you something and the closer you can get to the off the clearer the message (Wally, 70, retired tug-boat driver, southeast London).
Wally described how he also sought to 'get out from my own way' like Donnie and Winston: 'It's not like working out the race, like a puzzle, it's a feel, or a signal ... It's not thinking. Or ... it's trying not to think' . Sometimes, OTC betting is an attempt to get as close as possible to an event in time while leaving it unaffected by that proximity. This description of a momentary and fickle shadow cast by a future event necessitates Wally's late bets and shows that his engagement with certain races resembles that of players with their machines.

Socializing on the machines
The FOBs interface and OTC betting are each capable of facilitating a disengagement similar to that described by machine players in Australia and Las Vegas. In this section I describe a form of engagement with FOBs that depends upon relationships between people rather than between people and particular technologies. This form of engagement became familiar to me after long periods of time spent in betting shops. My guide to the FOBs scene was Donnie. In my local shop, Donnie identified four main groups of players: Serbs, Albanians, Chinese, and young black British men. 'These are the main groups in this shop that you will see playing a lot. Not always. You might get someone else try them out, maybe a young guy from around here [by this Donnie meant white British], but this is not so normal' . During a typical afternoon, a large crowd of perhaps fifty people would mill around the four machines and drift in and out of the shop. Some of the people were playing as individuals, others were nominated to play on behalf of their friends. Rival teams with different playing styles and times of playing were easily distinguished once I had a guide. Side bets and the circulation of money, goods, and favours were part of every session. As Donnie told me: In the mornings we watch and see who is playing, but we don't play ourselves. The machines need to fill up, to get ready to pay out. So you want someone to sit and play with their money, before you get on with your own.
Money-handling was one of Donnie's many responsibilities as de facto organizer of the group. During our morning tour of various bookies, Donnie and I would encounter acquaintances either who wished to contribute to the day's play and approached us, or whom Donnie felt able to approach to ask for funds. Sometimes men would stop their cars or jump from passing buses to give Donnie money for the day, holding up traffic, eager to be part of the action. On other occasions, people would spot us and either studiously ignore us or run away. In either case, Donnie kept a complete picture of who owed what in his head, calculated according to his own unique accounting system, in which money and favours (jobs, lifts, tip-offs of trouble) found equivalence. Cigarettes and other sundry goods served as the small change that lubricated this system.
When groups like Donnie's were playing the machines, although one individual played at any one time, this was a collective experience in every sense. Players would swap seats if an opportunity arose that was deemed to better suit one person's skill over another. As machines 'turned' against or in favour of a player, they may swap, to 'change the luck' or because 'Chris is good at sitting quietly and playing gently until the machine turns' . More and less aggressive play was required at different times and each player's style was to be deployed at optimum moments in the rhythm of the game. 'Horses for courses' was how Donnie described this communal labour. Players were surrounded by other members of the group who offered advice and provided a constant commentary on gains and losses. The people engaged in the session thus included players (of whom there may be several), contributors, and those who were regarded as potential contributors but who may not have been part of the action on this occasion. Playing to exhaustion or running out of money brought a low-key end to a session and opened discussions of mistakes made and future strategy. Decisions to cash out were made by Donnie and trusted deputies after animated discussions. Cashing out either limited losses or ensured a profit to be distributed in accordance with Donnie's calculations.
Even those who are not part of a playing group are not immune to the attention of others. I was amazed at the brazen interventions by regulars in my shops while newcomers played the machines. One persistent young man called Ishan would reach across unknown players and press the screen, something that might annoy even the closest of friends. Sometimes he was told to get lost, but occasionally his bet coincided with success and he might receive a gift of money from the player. In this way Ishan became the shop clown, licensed to intervene where others would not dare and a source of good luck -until he fell off a bus and broke his leg. A group of twelve of us went to see him in hospital to give him the shirt that Donnie had stolen especially from Marks & Spencer as a 'get well soon' gift. Some of Ishan's visitors asked him to touch the money they planned to play with later that day.
In betting shops, machine play is not always asocial and cannot be opposed to OTC betting on this basis. Donnie and his group have appropriated FOBs and turned them into a wholly social form of gambling in the sense not only that play is not always single-user, but also, more profoundly, that play is informed by communal decisions and strategies, and the sharing of stakes and winnings. Donnie is acting as a kind of bookmaker: by collecting stakes and sharing winnings, he is dealing in risk. Conversely, betting OTC is often a solitary activity conducted surreptitiously and without engagement with others. Conducted in this way, it is less social than FOBs betting of the kind I've just described.
Although machine play enables a form of engagement which has been characterized as solitary and asocial, in betting shops the realization of this potential is limited by two factors: the physical layout of the shop; and the different etiquettes associated with OTC betting and machine play. Both factors reduce the potential for spontaneous social interaction between OTC bettors and increase that potential between machine players. It is almost impossible to prevent people watching someone who is playing a machine in a shop where hanging around not betting, or 'lurking' as it is known in the trade, has always occurred. More importantly, looking at machines, unlike looking at slips, is not, under all circumstances, forbidden. Attempting to read another person's slips in a betting shop is taboo, both in the body of the shop and especially at the counter, where cashiers immediately learn that slips should always be handled face down. Peeking at slips starts fights. In this sense, OTC betting is private and therefore individualized; machine play is not. In betting shops in London, it may be argued that owing to their contrasting material culture and the conventions governing their use, betting on machines is more open to communal play than is betting OTC. 10 'You literally cannot win' I can't believe anyone would be stupid enough to play a machine controlled by the bookmakers, raking off a great chunk of cash, where you literally cannot win. This is a betting shop, but that's got nothing to do with betting whatsoever.
Steve, builder, 47, central London Steve's description of the distinction between OTC betting and machine playing prompts a number of questions. First of all, is it possible to make a profit betting on horses in shops? Secondly, is the distinction between OTC betting and playing machines an embellishment of the distinction between games of skill and of chance (Oldman 1974;Stevens & Young 2010)? Steve describes betting as 'work' and the betting shop as his 'office' . He works at betting by studying the form, planning bets, comparing early prices, watching races in order to spot fast finishers, and betting systematically. He also keeps regular hours. Spending time in betting shops with punters, I became accustomed to the question, 'Where have you been?' if I arrived a few minutes late to my usual seat. I became so familiar with their routines that I could predict the movements of regular customers. If Reg was not at his 'desk' in Ladbrokes by ten thirty in the morning, for example, he would be in the snooker hall at five in the afternoon, watching salsa classes and drinking Guinness. As he explained, 'I like to take a day off every now and then, to stay fresh' .
Despite the work that OTC regulars put into their activities, the distinction between OTC betting and machine playing is not simply an elaboration of the difference between profitable betting based on skill and unprofitable betting based on luck. There are a number of factors that militate against profitable betting in shops. When bookmakers construct odds for events, they build in an 'over round' . In simple terms, the over round ensures that if the bookmaker receives equal amounts of money for each horse, they will make a profit regardless of the outcome. In practice, markets for events are more complex, and outcomes may be winners or losers for bookmakers. Despite this, one of the principles of bookmaking is that the price of doing business is built into the odds offered. Odds, also referred to as 'prices' , offered by bookmakers are thus poor value relative to those used by professional gamblers available on-line through betting exchanges, where markets are more efficient. 11 Big wins are theoretically possible from small stakes placed on multiple bets or 'accumulators' with striking names and gigantic odds. 12 The terms of these bets, constantly promoted in the shops, favour the bookmaker because as a number of bets are combined, the over round is also multiplied.
Professional horse-race bettors use on-line exchanges, and not just because of the superior markets they generate. In betting shops, winning gets you noticed. Every shop has its own 'monitored customers' . Bets made by these individuals are recorded under nicknames, usually based on their appearance so that they can be easily recognized by incoming staff. 13 As well as learning to recognize monitored customers in order to decline or reduce their bets, every cashier must learn their 'permission to lay' terms. An each-way bet of over £50 on a second favourite at odds of 6/1 or lower, for example, may require 'permission to lay' . Cashiers asked for a bet of this magnitude are obliged to consult their manager, who will note which customer is asking for the bet and either grant or withhold permission, perhaps suggesting less favourable terms or a lower stake. For larger stakes, shop managers should call head office and ask 'permission to lay' from an area manager. These two systems enable betting shop operators to decline or severely restrict business from customers or on terms they describe as 'unprofitable' and contribute to the impossibility of making a long-term profit betting OTC in shops.
Strategizing: history, form, and pedigree Zaloom (2003) and Knorr Cetina and Bruegger (2002) have described how financial services workers appropriate apparently neutral technology, including screens, in unexpected ways. I observed a similar phenomenon among machine players. Some machine play (and some OTC betting) is based on flow. Too much thinking and strategizing can hinder this process. However, both forms of betting also support elaborate interpretation and the use of a diverse range of evidence to support or disprove predictions. Although winning numbers are generated remotely and randomly, the recent history of each machine is given on screen as a string of numbers. Some players use this 'form' to attempt to predict the next number, perhaps dividing the roulette wheel into halves or quarters and playing numbers that are physically opposite those that have already come up. Others ignore the wheel and focus entirely on numbers, on the basis that the wheel merely represents the output of a random number generator. One 'numbers' system involves covering both 'neighbours' of a number in the next spin. Another dictates that if a twelve has come up, the next spin should cover at least one, two, three, and twenty-one. Some players claim to have identified hardware or software glitches and recruit people to simultaneously press certain buttons on the machines in a shop or number of shops. I was involved in a number of these co-ordinated exercises. Machine players accrue experience of winning sessions and attempt to repeat them, share knowledge about playing, and treat this knowledge as cumulative. In other words, they apply deductive techniques to machine playing.
Reading form as part of OTC betting is also an interpretative act. The graphic representation of a horse's career from the unsuccessful but improving FPU86 (fell, pulled up, unseated rider, eighth, sixth) to the stellar 1111 (won last four races) is not always a predictable guide to bettor action. 'Form' gains meaning in relation to horse, races, and bettors. The first line looks unpromising. However, Alan told me that it was his 'banker' of the day because: I've followed that horse from day one because me and my dad won a lot on its dam [female parent]. It started very weak and I wasn't surprised by the first couple of runs. Very green, like. A bit like the mare, she was a slow learner, but she had an engine. I watched the last three runs and he was unlucky (Alan, 65, former computer programmer, central London).
This bet contains a social logic, described by Bouquet as 'pedigree thinking ' (1993: 219), which anchors decisions in the past through the properties that are imagined to be passed forward by people and animals (Cassidy 2002).
'Pedigree thought' is captured in the form pages of the Racing Post, which provide the parentage of every runner, referring to the name of the horse, the weight they will carry in this race, their age, colour, sex, sire (male parent), dam (female parent), and sire of dam. Among the 'form' men in betting shops, this information variously tells you: 'How far it can run' , 'How good it's going to be' , 'What kind of ground it likes' , and 'How much it costs' . In keeping with this idea of the predictive value of ancestry, in the absence of form, breeding is the best indicator of ability. 14 The information provided for unraced horses illustrates this point: For many of my contributors who have been betting throughout their lives, the horses they are betting on today are the offspring of the horses that they and their fathers bet on in previous years. In this sense, 'form' and breeding are logics of continuity, ways of interpreting the past through the present by noticing patterns that reproduce rather than undermine existing understandings of people and horses as 'top class' , or 'also rans' , or perhaps just 'unlucky' . Form and pedigree enable numerous different outcomes to be predicted from the same basic facts. The relationships between pedigree, form, and the outcome of particular races are not quantifiable. In that sense, in practice, they are susceptible to the attribution of logics in the same way as are series of random numbers.

Policing the betting shop boundaries
The introduction of FOBs, made possible by a change in taxation, caused an upheaval in betting shops in London. It coincided with a reduction in the number of traditional betting shop customers associated with a decline in the popularity of horse-racing as a televised sport and the availability of on-line betting. Reg and Steve describe themselves as members of a 'dying breed' , something that is echoed by managers who have worked in London for decades. Sheryl, a manager in London for twenty-four years, told me, 'There used to be a lot more big punters when I started, but we don't get that kind of money any more. We've lost all them big players to on-line. You can bet anywhere now, can't you?' David, an area manager for thirty years, agreed: The kind of regular who comes in early and puts on his multiples or sits there reading the [Racing] Post is long gone. Now you get your passing trade and your machine player mostly. You want something there for them. There's no money in horses no more. I'd 'ave twice as many machines if I could.
In 2007, Chris Bell, Chief Executive of Ladbrokes, reinforced this message, saying that, 'UK horse racing is not as relevant to our customer base as it used to be. Customers are much more interested in machines and virtual [racing]. They bet on-demand now rather than spending ages studying form in the Sun in the morning. You have to adapt to that' (Bowers 2007). This reduction is relative: 831 million bets were placed on horses in betting shops in the UK in 2009/10, generating £760 million gross profit, half of the £1,443 million total produced in shops (Gambling Commission 2011a). 16 Video poker is the game of choice for local women in Las Vegas (Schull 2002). In London's cosmopolitan betting shops, playing roulette on the FOBs is an activity associated with incomers, people who have no experience of betting on horses or dogs, whether this is young men, women, or recent migrants. Michael, a regular punter who was born in southeast London but has West Indian roots, told me when I interviewed him, 'Eastern European, Chinese, Asians come in to use the machines. They never used to come in the bookies till we got the machines' . The betting shop professionals I interviewed also made this association, as Farah, a 30-year-old manageress working in north London, told me: Foreigners mainly use the machines. Probably they don't bet over the counter because they think they need to know about the horses and they need to know about the dogs. They need the knowledge. For FOBs it's just numbers. It's so simple. You don't need to know anything ... That's why I think, it's simpler. User-friendly, that's why I think they use it.
The group identified as the heaviest users of machines by betting shop managers reflected the local ethnic composition. In Hounslow, for example, Mark told me: Somalians and Eastern Europeans, they use just the machines. They don't have any knowledge about betting on, like, horses, dogs, so they wouldn't do it. Probably sometimes, like once in a blue moon, they would do on dogs, but horses they would never. Because dogs it's just choosing the numbers again (Mark, 32, betting shop manager, southwest London).

Mark's colleague David added:
We get a lot of women playing the FOBs in this shop, more than I've seen before. The foreign legion, we call them. They come in from across the road. The nail bar over there. Don't talk a word of English. A little girl comes in most days and drops three grand [£3,000] on the machines (David, 55, are manager, central London).
Betting on horses, on the contrary, was associated with older men: The other people in the shop, like, most of them are old people. Above 50. And they bet on horses. They, they don't touch the machines. And they are very good people. All of them. No problems with them at all (Monique, 28, betting shop manageress, southwest London).
Monique had just finished telling me about the young Somali men playing the machines: 'We get a lot of trouble with the Somalian kids. Because umm. They, they are underage. If you ask them for ID they get aggressive and they don't want, like, just causing trouble. Most of them. I wouldn't say all of them. But most of them, they do' . Attributing different qualities to OTC bettors and machine players contributes to the conservation of a 'respectable' relationship between leisure, work, and reward in betting shops (Downes, Davies, David & Stone 1976).

Conclusions
Fieldwork in betting shops undermines a formal opposition between betting and machine gambling. Both activities support elaborate interpretations. How different are stories about pedigree, ground, and form from those about mysterious software and relationships between numbers? In both cases an explanation is sought. In neither case is the explanation verifiable. The representation of machine gambling as incapable of supporting the same degree of intellectual engagement as betting is at least partly a reflection of the nature of the betting shop as a social space and the historical significance of betting on horses and dogs in the UK. Staff and traditional bettors, in my local shop a group of West Indian, black and white British, Turkish, Greek, and Vietnamese men, re-cast gambling as work and distinguish themselves from irresponsible gamblers. Machine players in the same shop, mainly Serbian and Albanian men, engage with the idea that playing must be worked at in order to bring reward, but also recognize that it is not unusual for things to be outside one's control, something that bettors are loathed to admit. Both groups employ a range of intellectual and supernatural techniques in order to get in touch with the truth that their hard work is intended to reveal. Steve, the middle-aged builder we met earlier, was the strongest critic of machine playing I encountered. He told me, 'People who play the cartoons are embarrassing. It's embarrassing. I can't look at them. They give the bookies a bad name' . After he had explained that betting was 'a science, like biology or chemistry' , I asked him why he touched his head each time he had a bet and he told me, 'You've got to be flexible haven't you!' 17 Writing about Internet cafés (PC Bangs) in South Korea, Lee argues that interactions may appear post-social or even post-human, but they do not escape the structural inequalities of the wider national society. On the contrary, they are 'dependent on the national electronic space of the Korean Information Infrastructure' (Lee 2007: 724). The same can be argued of betting shops in London, where a change in consumption has been described by OTC bettors and staff as a shift from betting to gambling. According to this discourse, betting on horses is a respectable skill-based activity that people work at, during their spare time. Machine playing is not only mindless but also dangerous, because it provides the kind of stimulation that turns vulnerable people into addicts. Machine players, who are often newcomers to the betting shop, know little about the language of pedigree, form, and odds, play the machines both singly and in groups, sometimes losing themselves, sometimes connecting to others.
In the UK (and in many other jurisdictions), access to risks of these kinds is structurally determined as well as technologically mediated. The regulation of machine gambling produces a channel with low entry costs and easy access in traditionally working-class spaces. Ethnographic data from betting shops in London show that the effects of changes in regulation as they relate to technology are unpredictable. 'More' technology does not necessarily lead to asocial or less social gambling, and may under certain circumstances contribute to the creation of elaborate communal practices, as among machine players. It is essential that we understand the experience of gambling using new technology as well as the ways in which media, new and old, fit into particular social niches. Both kinds of betting are enabled by regulation that favours a strong industry which includes the state not only as competitor but also as beneficiary, despite the regressive properties of gambling taxation (Pickernell et al. 2004). It is only by paying attention to social and experiential aspects of gambling that we will be able to understand the impact of such policies.

NOTES
Thanks to my research participants, whose names have been changed in accordance with their wishes. This article is based on fieldwork supported by a joint programme of the Economic and Social Research Council and Responsibility in Gambling Trust, reference RES-164-25-0005. Writing up was supported by a European Research Council starting grant, 'GAMSOC' . Thanks to colleagues at Goldsmiths and the University of London who read and responded to this paper and to three anonymous reviewers for their stimulating comments.
1 The intention to provide a fiscal and regulatory environment 'which enables the gambling industry to compete and grow' is referred to by HM Customs and Excise (2002). A leaked note from Richard Caborn, former minister responsible for gambling, to Mark Davies, former managing director of Betfair, reads: 'It is government-wide policy, and that includes HMT [Her Majesty's Treasury], that Britain should become a world leader in the field of online gambling, in order to provide our citizens with the opportunity to gambling [sic] in a safe, well-regulated environment' (Adams 2007). A similar process of re-categorization has been described by Matilainen (2009) in Finland and Kingma (2004) in the Netherlands. Sallaz (2009) provides an illuminating comparison between the regulation and commercialization of gambling in South Africa and Indian gaming in California. 2 The majority of FOBs in betting shops offer B2 content, defined by the Gambling Commission in June 2009 as offering a maximum stake of £100 in multiples of £10 and a maximum prize of £500 (Gambling Commission 2011b). They offer a variety of games, including casino games like blackjack, with roulette being the most popular. Use of these machines is poorly understood, as little research has taken place in betting shops since they became popular in 2002 (but see Europe Economics 2005;Hancock 2006;Reith 2006;Thompson, Hollings & Griffiths, 2009). 3 In 2008, Chris Bell, CEO of Ladbrokes, described overall growth in gross win of 4.7 per cent as 'driven by our ... machines which delivered gross win growth of 15.2%' (Ladbrokes Annual Report and Accounts 2008). 4 In 2010, Ladbrokes owned 2,073 shops, William Hill 2,228, and Coral 1,630 (Gambling Commission 2011a). 5 The Gambling Commission does not regulate either the National Lottery or spread betting. Gross gambling yield (GGY) is the conventional way of measuring gambling in the UK and defined as 'the amount retained by operators after the payment of winnings but before the deduction of the costs of the operation' (Gambling Commission 2011a).
6 Of William Hill's 2,228 shops, around 650 are inside the M25. In 2008, 35 per cent of the total gross win of £802.6 million came from these shops. 7 The extreme competition between new entrants to the market like Metro and Better are evidence of this. 8 The number and distribution of shops in London has changed since 1966, when, of a total of 16,000 shops, Greater London accounted for 2,249, 2.83 per 10,000 members of population. This figure was lower than for the UK as a whole (2.97) and below that for England (2.84) (Newman 1972: 102). By 2007, one in four bookmakers was in London (Evening Standard, 30 April 2007). The distribution of these shops reflects the changing geography of the city and the demography of the betting shop customer. Westminster has the most, with 113, Hackney the second highest number of 95. In 1966, Westminster had the highest number: 150, but Hackney had only 17 (Evening Standard, 30 April 2007: 11;Newman 1972: 102-4). 9 Reith (2007) has meticulously unearthed the archaeology of the idea of problem gambling which frames these debates. 10 Winnings from OTC betting may also be shared in at least two ways: as spontaneous and irregular gifts of 'luck money'; and through networks of friends and relatives. The first time I was given luck money, I wasn't sure whether or not to accept it. A punter collecting £400 winnings gave me £20 and told me, 'There you go,