Subscribe to the Journal:


If you would like to support the Journal, please do so here.  The Asia-Pacific Journal is available free to all. But your contribution allows us to improve and expand our service  in the wake of 3.11.

Donate - $25, $50, $100

Join Us:JapanFocus Twitter page  APJ Facebook Page  

Display Your BOOK, FILM, OR EVENT here

 Peace  Philosophy  Centre

Dialogue and learning for creating a peaceful, sustainable world.


Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
Click a cover to order.
The Asia-Pacific Journal: Japan Focus
In-depth critical analysis of the forces shaping the Asia-Pacific...and the world.

The Asia-Pacific Journal, Vol 11, Issue 42, No. 2, October 21, 2013.

Those Restless Little Boats: On the Uneasiness of Japanese Power-Boat Gambling

落ち着かない小舟: 日本競艇賭博不安論

Tom Gill

Send to Kindle

Gambling fascinates, because it is a dramatized model of life. As people make their way through life, they have to make countless decisions, big and small, life-changing and trivial. In gambling, those decisions are reduced to a single type – an attempt to predict the outcome of an event. Real-life decisions often have no clear outcome; few that can clearly be called right or wrong, many that fall in the grey zone where the outcome is unclear, unimportant, or unknown. Gambling decisions have a clear outcome in success or failure: it is a black and white world where the grey of everyday life is left behind.

As a simplified and dramatized model of life, gambling fascinates the social scientist as well as the gambler himself. Can the decisions made by the gambler offer a short-cut to understanding the character of the individual, and perhaps even the collective? Or are people different when they gamble? Gambling by its nature generates concrete, quantitative data. What can we learn from that data? In this paper I will consider these issues in relation to powerboat race (kyōtei1) gambling in Japan: an industry with sales of ¥920 billion in financial year 2011.

Fig. 1: Restless little boats. Contestants at Heiwajima jockeying for position before the start. Photo by author.

Kyōtei is just one branch of the mighty Japanese empire of gambling. It is the second most popular form of race gambling after horse-racing with sales of ¥2,625 billion in 2011. Another ¥620 billion was spent on bicycle racing, and ¥84 billion on motorbike racing. The total spent on race gambling in 2011 was ¥4.25 trillion.2 The spend on race gambling has been declining steadily and is now at about half its peak level of ¥8.5 trillion yen recorded in 1990, the year the bubble economy collapsed. Nonetheless, by international standards Japan is still a big gambling center. Let us never forget that race gambling is dwarfed by pachinko, which is classified as ‘amusement’ rather than ‘gambling’ but reported gross revenue of ¥18.9 trillion yen in 2011.3 Japanese people also spend about 1 trillion yen a year on the national lottery (Takarakuji; literally ‘Treasure Lottery’)4 and smaller amounts on sports lotteries etc.

This empire exists despite the fact that gambling is illegal in Japan, under article 23 of the Penal Code, which prescribes up to three years with hard labor for any “habitual gambler,” and three months to five years with hard labor for anyone running a gambling establishment. Yet laws passed over the years allow the state to ignore its own moralistic prescriptions.


A general analysis of gambling in Japan, or international comparative analysis, is beyond the scope of this paper. Instead, having established that gambling is a major phenomenon in contemporary Japan, I will keep a narrow focus on powerboat racing – itself a massive phenomenon. I have four field sites. The first is the powerboat stadium at Heiwajima, on the southwest side of Tokyo, near Shinagawa. The second is the ‘Boatopia,’ or off-course betting center, in central Yokohama. The third is a certain café in Yokohama where the proprietor, a good friend of mine, has been studying boat-racing form for 20 years. The fourth is a dog-racing track in Oxford, U.K., which I have visited several times to generate comparative data. I hope to mix fieldwork-based qualitative research with quantitative data in a bid to elucidate what powerboat gambling can tell us about class, gender and personality in contemporary Japan.

Fig. 2: Yokohama Boatopia. Photo by author.

First, however, some basic information. Professional powerboat racing in Japan dates from the promulgation of the Motorboat Race Law in June 1951 (Miyoshi 2009:72). The Ministry of Transport was given supervisory authority. Other forms of public gambling were authorized in the same period – horseracing in 1948, under the auspices of the Ministry of Agriculture; cycling in 1948 and car/motorbike racing in 1950, under the Ministry of International Trade and Industry. This was all part of a government campaign to finance postwar reconstruction.The national lottery, relaunched in 1948 after a brief experiment just before the end of the war, dates from the same period. The official objective of powerboat gambling was to raise funds for “projects related to motorboat development and other kinds of shipbuilding, the dissemination of maritime thinking,6 tourism enterprises, promotion of public interest projects in the field of physical education etc., and strengthening of regional finances” (Tanioka 1996: 115). As Tanioka points out, the great changes in the Japanese economy in the last 60 years call into question the need for public gambling income and strengthens the voices of those who wish to see these exceptions to the anti-gambling law abolished (ibid. 114-116).

From modest beginnings, kyōtei grew rapidly from the late 1960s until 1991, and from 1972 to 1983 it outstripped the nationally run horse races to be the biggest form of race gambling in Japan. But it went into decline after the Bubble Economy burst: after peaking at ¥2.21 trillion in 1991, sales declined to ¥0.97 trillion by 2005 – a 56% decline (Miyoshi 2009: 142-3, 151). That is roughly where they are today.

All the contestants are professional racers and graduates of the Yamato Boat Racing School run by the Japan Motorboat Racing Association (JMRA; Nippon Mōtābōto Kyōsōkai), located since 2001 in Yanagawa city (Fukuoka prefecture), which issues each of them with a four-digit registration number. There are 24 boat race stadia in Japan, all of them under the auspices of the JMRA, and also some 30 off-course betting centers. The racing is a completely closed shop – there are no unlicensed courses or contestants, nor any rival organization to the JMRA. The gambling is a slightly different story. There are some unlicensed, illegal betting shops that offer closed circuit television coverage of races and use the same computer-generated odds as the legal betting centers, as I will discuss below.

Famously, the JMRA used to be run by the late Sasakawa Ryōichi (1899-1995), a controversial figure who was arrested for war crimes in 1945 but never indicted, and it is now run by his son, Yōhei. Under the Motorboat Racing Law, 2.6% of the profits from boat racing (reduced from 3.3% when the law was reformed in 2007) is administered by a related foundation, the Nippon Foundation. Formerly known as the Japan Shipbuilding Industry Association (Nippon Senpaku Shinkōkai), and better known around the world as the Sasakawa Foundation, it was informally ‘nicknamed’ the Nippon Foundation in Japan in 1995 and officially took that name in 2011 when it achieved the status of charitable foundation (kōeki zaidan hōjin).7 The various branches of the Nippon/Sasakawa Foundation are highly influential in Japan and abroad, for example bankrolling many Japanese studies programs at universities around the world.8 The present paper will not go further into this admittedly fascinating topic, instead focusing more tightly on the boat races and punters themselves.

Individual stadia are run by local authorities, and for some they are a significant source of income.9 However, the administrative structure of the industry reflects the highly ambiguous attitude to gambling in officialdom, which is always uneasily aware of the gross hypocrisy of the state making money from a practice it condemns and outlaws. This has led to some surprising outcomes. For instance, local authorities may run race courses, but they do not generally advertise them. Miyoshi, in her interesting account of local government-run race courses, records a scandal that occurred in Nara prefecture in 2002, when the prefectural government newsletter carried, for the first time ever, a brief announcement of race days at the local velodrome. The prefectural government was widely accused of corrupting public morals by advertising gambling in an official publication, though it had been quietly running the race track for 52 years (Miyoshi 2009: 145-6).

Consider, for another example, my two main field sites. The Heiwajima boat stadium is in Tokyo’s Ōta ward; the Yokohama off-course boat betting center is in Yokohama’s Naka ward. Yet both these facilities are run by the city of Fuchū, on the western outskirts of Tokyo, nowhere near either venue. This stems from the 1969 decision by Tokyo’s famous left-wing governor (1967-1979), Ryōkichi Minobe, to abolish all forms of municipal gambling, which he viewed as immoral and as a tax on the poor (Miyoshi 2009: 117-128). Minobe was taking a traditional left-wing position: he was strongly supported by the left-leaning Asahi Shinbun, which consistently campaigned against gambling from the Meiji era to the 1980s (Nagashima 1998: 354). Minobe’s stand was strongly supported by the mayor of Yokohama (1963-1978), fellow socialist Ichio Asukata (Miyoshi 2009: 120), and although the city of Yokohama never quite pulled out of the gambling business until 2010, when the Kagetsuen bicycle track in eastern Yokohama was closed down after sixty years, it has not taken charge of any new gambling ventures since then.

But that was not the demise of gambling in Japan’s two biggest cities.10 The facilities are there, but they are run by a medium-sized town 40 kilometers away, which brings with it the experience of hosting the biggest horseracing course in the Tokyo area. To complicate matters still further, Fuchū city is not the official promoter of the Fuchū racecourse, which is run by the Japan Racing Association. Nor does it run the Tamagawa boat race stadium, which is actually in Fuchū – that is run by the city of Ōme, some 25 kilometers further along the Tama river.

So when you walk into a powerboat stadium or off-course betting center, you enter awkward, contested territory. On the one hand, a site of entertainment and virtuous contributor to the common weal; on the other hand, a place where criminal behavior is licensed – a place whose existence may not be mentioned by the body in charge of running it.

The moral waters are cloudier still in Yokohama, where the Boatopia off-course powerboat betting center, a smart five-story citadel of gambling, opened in December 2007 in Ōgi-chō, right next door to Kotobuki-chō, Yokohama’s main skid-row district (see my book on Kotobuki-chō, Gill 2001). The men who live there, many of them welfare recipients, include large numbers of habitual gamblers. Thus a cynic might compare this bit of urban planning to constructing a cocaine shop next door to a town full of cocaine addicts. The city fathers’ defense argues that this will take customers away from the numerous illegal gambling dens run by yakuza in Kotobuki-chō, who take bets on bicycles, boats and horses, using the same closed-circuit TV service as legal betting centers to display odds and televise races. They then pocket the 25% of stake money supposed to go to the government in betting tax, as well as setting a 100 to 1 limit on payouts.11 Then again, government critics retort, if the city of Yokohama was that concerned about illegal gambling in Kotobuki-chō, they might perhaps have closed down the twenty-odd gambling dens that have been operating there, known to all local residents and to the authorities, for several decades. One of them, after all, is round the back of a yakuza office less than a one minute walk from the Kotobuki-chō police station.

As I say, this is murky, morally compromised territory.

My Basic Theory of Gambling

In the brief discussion of gambling in my book Men of Uncertainty (Gill 2001: 165-166) I proposed a taxonomy of gambling, based on the four principal issues of (1) chance of winning; (2) scale of payout if successful; (3) whether the opponent is an individual or a professional house; and (4) the relative importance of skill and luck in determining the outcome. These elements, I argue, tend to vary in a common pattern, generating a kind of spectrum. Thus poker or mah-jongg are games with a high chance of winning, a relatively small payout (except for high rollers), played against other individuals, and with a large degree of skill involved. At the other end of the scale, when buying a national lottery ticket – in Japan’s case, for the Takarakuji – the chance of winning is infinitesimally small; the possible payout is colossal; it is played against a professional house (the national government); and the outcome depends almost entirely on luck.

In that book, I noted that race-betting and pachinko, the most popular kinds of gambling among day laborers in Japan’s skid-row-like yoseba districts, are both located around the middle of the spectrum. Let me elaborate a little on what I mean by that. In race betting, (1) the chance of winning is generally somewhere in the range between 1 in 2 (as in a game of cards with an evenly-matched friend) and 1 in millions (the Takarakuji); (2) the payout will tend to be more than the few thousand yen in a friendly card game and less than the many millions of yen in the Takarakuji; (3) one’s opponent is somewhat ambiguous, with a professional house (a publically-run race venue or off-course betting center) but a system of betting that effectively pits punters against each other; and (4) there is a degree of skill and knowledge involved in picking winners, though luck is undoubtedly the biggest factor.

Let me explain point 3, the ambiguity of the opponent. An important element in any kind of race gambling is the setting of the odds. This can be done in two ways: (1) a professional bookmaker sets odds and takes bets; (2) a tote, or pari-mutuel system subtracts the house cut (in the case of Japanese race betting, that means the stinging 25% government betting tax12), and then divides the remaining stake money between the holders of the winning ticket. In the former case, the gambler is competing against the bookmaker, who is himself taking a gamble when he sets the book. Bookmakers will offer longer odds than their rivals to attract more custom, or shorter odds on outcomes that they fear may actually materialize. A bookmaker can make a mistake, and make a loss on a race. Too many mistakes will bankrupt him. But in the latter case, the house will never lose. The odds are automatically generated according to the amount of money bet on each outcome. If the favorite wins, the house will pay a small amount of money to a large number of punters; if an outsider wins, the house will pay a large amount of money to a small number of punters. But in either case, the computer-generated odds ensure that the house keeps 25% of the staked money. Hence the punter is effectively gambling against his fellow punters, trying to find a winning outcome that will not have been spotted by too many of the other punters so that it will still offer attractive odds.

Let me note the complex psychological territory we are entering here. The Japanese race gambler will usually lose, since he faces unfair odds; on average he will lose 25% of his stake because of the betting tax. He can blame his losses on poor performance by his selected contestants, while ascribing his occasional wins to his own sagacity; and he may feel that his opponent is a huge government-run institution, whereas in fact that contest is decided in advance in favor of the latter, leaving the punter to struggle against his fellow punters for the remains of the cake after the government has taken its hefty 25% slice.

So far my comments apply to all forms of race gambling in Japan – on horses, bicycles, powerboats and motorbikes.13 Now let us gradually narrow the focus to and consider the properties that distinguish powerboat racing from the other kinds of race gambling.

In class terms, powerboat and bicycle races attract a more dominantly working-class clientele than horse-racing does.14 Admission is almost free (50 or 100 yen is common), and I have noticed that these stadia are quite often frequented by homeless men looking for a sheltered place to lie down, perhaps with some free entertainment on the side. If you do not bet, it is an extremely cheap day out. Tanioka (1996: 83) cites a study made in the early 1990s which found that 69% of university students had bet on pachinko, and 53% on horse races, but only 7% on powerboat or bicycle races. Clearly the latter two are further marginalized from mainstream culture than the former two. A 2006 survey by a Japanese public opinion pollster found that 58.2% of 1,379 men and women over the age of twenty reported having gambled in some way in the last year. Of those, 75.1% had bought a lottery ticket, 39.7% had played pachinko, 15.7% had bet on horse-racing, 3.1% on bicycle racing, and just 2.4% on power boats (CRS 2006: 1). Again, bicycle and boat gamblers appear as a relatively small fringe element of Japanese society. The data I cited earlier shows that the annual spend on powerboat gambling is about a third the total for horseracing. If this survey is reliable, there are only about a sixth as many punters, indicating that the per-capita spend is higher for power-boats than for horseracing. The picture is of a smaller population but consisting of more serious gamblers. (An alternative possibility is that more people lie about boat-race gambling than about horse-race gambling.) For bicycle gambling, about a fifth as many punters staked about a quarter of the money wagered on horseracing, so the gap is less pronounced than for power-boat gambling.

Fig. 3 A date at the off-course betting center

(Scanned from Boatopia advertising leaflet)

In gender terms, there is a male majority in all forms of gambling in Japan – in the CRS survey (2006: 1), 51.3% of women said they had never gambled, against 31.2% of men – but again the gap is especially noticeable in boats and bicycles, despite the industry’s best efforts to attract female fans and promote the races as suitable dating activities. On visits to the Yokohama Boatopia, I have generally found several hundred men and virtually no women among the clientele. Note also that boat and bicycle races are held seven days a week, right through the year, whereas horse racing does have a sort of off-season in the winter and is mostly held on weekends and public holidays. So the boats and bicycles are more obviously geared to the habitual gambler than the horses are. The Yokohama Boatopia is open for business 360 days a year.

Battle of Six

In a sense, powerboat gambling is the simplest, most standardized form of race gambling in Japan. In horse races (keiba) there may be up to 18 horses. Bicycle races (keirin) involve six to nine cyclists. Motorbike racing usually has eight riders. But in powerboat racing (kyōtei) the number of contestants is always six. The industry plays on this simplicity of form, which has a certain appeal in the numerological culture of gambling. The players always wear the same colors: #1 white, #2 black, #3 red, #4 blue, #5 yellow, #6 green. This iconic color-coded gang of six is often represented in posters, stickers, figurines, etc. The use of bright primary colors, and the wearing of helmets, recalls the supā-sentai TV heroes (known outside Japan as Power Rangers) that I have discussed elsewhere (Gill 1998). Another set of cousins might be the Shichi Fukujin, or seven lucky gods.

Fig. 4 Battle of six

This and figures 5, 8, 9, 10 from Bōtoreesu Baiburu (The Boat Race Bible), official publication of JMRA.

Likewise, horse-racing courses, velodromes and speedways vary in length, and the distance of races also varies. But all Japanese powerboat stadiums have an identical 600 meter oval course, which the boats must go round three times, for a total of 1,800 meters. It takes a little under two minutes. Gamblers who dislike boat-racing will call it boring, or wan-pataan, literally “one pattern.” To aficionados, however, there is something hypnotic about the endlessly repeated three-circuit races.

The role of female contestants is a striking feature of kyōtei. They constitute about a tenth of all professional boat racers in Japan (roughly 150 out of 1,500). They usually compete in women-only races, but the better ones will race with the men, and some, like the famous Yokonishi Kanae, known as ‘Kana-chan’ or ‘the nation’s strongest single mother,’15 can sometimes win top-level races against men. Though there are always rumors of race-fixing, they do not appear to focus on women more than men. Certainly Chie Takahashi has not been done any favors: she holds the record for the most unsuccessful boat racer ever, losing 333 times before she recorded her first win, and another 268 times before recording her second victory, on January 9, 2004 (Suzuki 2008: 139). Female racers do have a certain sex appeal: they tend to be working-class girls, tomboyish with close-cropped dyed hair, and will attract bets that might not be warranted by purely scientific considerations.

Fig. 5 The oval course.

Since physical strength is relatively less important than skill and technical knowhow in boat racing, it also allows players to remain active for longer than in most race sports. It is quite common for boats to be piloted by people in their forties and fifties, and as I write, popular veterans Mantani Akira and Katō Shunji are still competing successfully in professional races at the ages of 69 and 71 respectively.

Fig. 6 Yokonishi Kanae.

(Photos from fan-club home page).

Another feature that distinguishes boat racing from other kinds of race gambling is the enormous significance of a good start. Boat races begin with a running start: boats already in motion must pass the start line within 1 second of the start time, displayed on a huge clock with only a second hand, or will be disqualified. Many races are effectively decided at the start line, and the majority are won by the boat that gets round the first turn in first place. Very often that boat will then lead the others round the remaining 2.5 circuits, in a pattern called a nige, or ‘escape.’ Once a boat has established a lead, it is very difficult for others to overtake, because (a) the lead boat can take the inside course at every turn, reducing the distance it has to travel compared to the others; and (b) the pursuing boats will have their progress impeded by the wake from the lead boat. It follows that having the inside track at the start of the race is a very substantial advantage, and likewise the outside track is a major disadvantage. When the #1 boat gets away at the first turn to win, that is called an in-nige, or ‘escape from the inside.’

The other patterns are the makuri (wind-round; the winning boat overtakes from the outside), the sashi (stab; outside boat slips inside lead boat at the turn, which can happen if the lead boat fails to decelerate sufficiently and overshoots the marker buoy); the makuri-kaeshi (wind-and-return; the eventual winner goes around the outside of following boats, then slips inside the lead boat; very difficult); the nuki (lead boat is overtaken on the final circuit); and the megumare (blessing; one or more boats are disqualified or disabled and a lower-placed boat gets the win).

Fig. 7 Mantani Akira and Katō Shunji

(Photos from JMRA homepage)

Since the nige is the most common race pattern, far more races are won by the #1 boat than any other, and a win for #6 is rare. In fact, the #1 boat is roughly ten times more likely to win than the #6 boat.16 I should mention that the boats do not necessarily line up with #1 on the inside going through the numbers to #6 on the outside. The boats come out of their pits (pitto-auto), then engage in “waiting activities” (taiki kōdō), which includes doing a practice circuit. During this time there may be some jockeying for position resulting in one of the boats grabbing a better position (mae-tsuke). It is a ticklish matter, for these restless little boats have no brakes. Hence an over-zealous attempt to grab the inside track may result in one’s boat drifting too close to the start line, preventing one from generating enough speed for the flying start. This error is called a fuka-in (‘deep in’).

The industry has never quite decided what it thinks about jockeying for position before the start. There was an attempt in 1996 to ban it, and start all races with the line-up printed on the card. But this was unpopular with punters who thought it made the race more boring. Nowadays enforcing the lane numbers on the card (a style of racing called shin’nyū kotei kyōsō or fixed lane racing) is restricted to a handful of novelty races… at some courses only the ninth of the twelve races on the daily card is run that way.

Fig. 8 Main Winning Techniques

Fig. 9 Main Winning Techniques

Nonetheless, grabbing a better lane before the off seems to be relatively rare, and most races start off with the boats in their numerical order (waku-nari). Boat racing also has various unwritten rules – inexperienced newcomers are supposed to stick to the outside lanes, for instance. This is supposed to stop them getting in the way of the more experienced racers, but effectively it creates a pecking order. At one course – Edogawa in Tokyo – tidal river conditions make jockeying for position especially difficult and dangerous, and it is customary (though not obligatory) for all races to be run with boats in their assigned lanes. By contrast, at Toda in Saitama prefecture, some racers feel that the relative narrowness of the course makes it harder to win from the inside lane and will actually seek to get into the second lane even when drawn in the first lane.

Fig. 10 Taiki kōdō 

Maneuvering by boats before the official start.

The taiki kōdō phase is extremely ambiguous. Although it means “waiting activities,” and although racers are forbidden from going at full speed, they are still trying to get a better position – rather like the cyclists in keirin, although in the case of cycling the slow-motion battle for position occurs after the official start and so is unambiguously part of the race. According to my knowledgeable friend Ichikawa Kazuyuki (see below), about 80% of all races start with the boats in the published order, with the percentage being higher in eastern Japan than in western Japan – reflecting the pushier character of Kansai folk, he says.

One factor deterring racers from overly aggressive lane-grabbing is the fact that the committee picking the starting positions will generally try to be fair, and if one is given an outer lane on the first day, one may well get an inner lane on the second. If one aggressively tries to grab someone else’s lane, he may end up being given the same treatment on later days.

On the last two days of a meet there will be semi-finals and finals, and then the starting order will be decided on form, with the racers with better records getting the inside lanes. So the likelihood of high-number boats winning will be even lower on the last two days, for reasons of form as well as lane positioning.

Typically, boats #1 to #3 start on the inside, lining up closer to the start line than boats #4 to #6, which start further back. This is because the inside boats need to make a slow start (surō) to avoid overshooting the marker buoy at the turn, while the outer boats need to start with a dash (dasshu) as their wider course means they have a greater distance to cover. The #4 lane, as the “innermost of the outside boats” is favored by some racers, and this position is sometimes called “the corner” (kado). My data in note 16 shows boat #4 doing marginally better than boats #2 and #3 inside it.

Anyway, the point is that any bet that includes boat #6 is likely to be a romantic bet on a natural loser to somehow become a winner. The exception is if the man on the outside is an autoya (‘out specialist’) such as Awa Katsuya or Ogawa Kōji, who specialize in making a very fast start from the outside lane and cutting in front of the field at the bend for a makuri. For 2013, Ogawa is graded A2 and Awa B1 (out of A1, A2, B1, B2), but despite his humble rank, Awa in particular is enormously popular because of his radical approach to the sport. He deliberately tilts his engine by three degrees (chiruto 3-do), the maximum angle permitted at eight powerboat stadia (the other 16 only allow two degrees). This puts his propeller (perā) closer to the water’s surface, giving him extra speed, but also making cornering more hazardous and increasing the risk of capsizing (Suzuki 2008: 74: 80). He is nicknamed ‘Awakatsu’ or ‘Misutā Chiruto-san’ and has had delicacies named after him, such as the ‘Chiruto-sandoitchi’ (Tilt sandwich) and the ‘Chiruto-sandon’ (Tilt rice bowl), which are sold by vendors at stadia where he is competing.

Figs. 11, 12: The Outsider – Awa Katsuya.

Photos from fan-club homepage.

A bigger task for the gambler than picking the winner is picking the second and third place finishers. Theoretically, one might bet on a single boat to win (tanshō), but this is not usually an option since it does not appear on the standard betting slip (tōhyō kādo; see figure 13 below). I will discuss this paradox a little later. The bets easily available are:

  1. The 1st and 2nd place finishers in correct order (ni-ren-tan; a forecast in British English, an exacta or perfecta in American)
  2. The same bet but with the two boats in either order (ni-ren-fuku; a dual or reverse forecast in Britain or a quinella in the US). Equivalent to buying two ni-ren-tan bets.
  3. The 1st, 2nd and 3rd place finishers in correct order (san-ren-tan; a tricast or trifecta).
  4. The 1st, 2nd and 3rd place finishers in any order (san-ren-fuku; a combination tricast or box trifecta). Equivalent to buying six tricasts,
  5. The nagashi or ‘wheel’; this entails a bet for first or first and second, followed by betting the field for the next finishing place. For example one might bet on the first and second place boats, and select all four of the possibilities for third place, effectively buying four tricasts.
  6. The bokkusu or ‘box’ means betting on all the winning combinations possible from three or more boats. A three-boat box is the same as the san-ren-fuku tricast), i.e. equivalent to buying 3x2 = 6 tricasts. A four-boat box means that three of the four boats selected must finish in the top three, so this is equivalent to buying 4x6 = 24 tricasts

Fig. 13 Tōhyō kādo (ordinary bets)

Photo by author. The standard machine-readable card used to make a bet at Yokohama Boatopia. The names of Japan’s 24 boat-race stadia are written at the top, and the number of the race (1-12) on the left. Then comes the four type of bet: ni-ren-tan, san-ren-tan, ni-ren-fuku, san-ren-fuku (first two or first three finishers, in specified order or any order).

One consequence of all this is that the key to successful betting on powerboats often lies in predicting who is going to trail home second and third after a successful nige, a significant difference from kinds of race betting where the main challenge is to predict the winner.

In deciding who to back, punters are invited to take into account not only the form of each contestant, but also the quality of the boats and engines they use. These are distributed randomly before a race meet, and although they are supposed to be identical, that does not stop the racing press from including data on the winning percentage for each boat and each engine as well as for each contestant. There is an interesting contrast here with American powerboat racing, where contestants will make great efforts to decorate their boats and stamp their personality on them as well as competing to make them more powerful and maneuverable. Japanese racers are all given virtually identical boats at the start of a meet and are not allowed to customize them in any way, say for relatively small modifications like the angle of the engine, discussed above.

Fig. 14 Tōhyō kādo (wheels and boxes).

Photo by author. The nagashi (wheel) selects one contestant to come first (and/or second, or third), and covers all the contestants in the other positions.

The bokkusu (box) covers all the combinations of a number of contestants (usually three) to fill the first two or three positions. Note (bottom center) that it includes the option of covering all six contestants (30 outcomes for first two places, 120 outcomes for first three places), so that it is possible to bet on every single result, guaranteeing a win (but also guaranteeing 29 or 119 losses).

However, powerboat racers do need to be good mechanics. This is low-budget racing, so they do not have teams of technical assistants. Instead they have to maintain their own boat and engine during a meet, and each will have a suitcase full of their most favored propellers – the propeller and spark plugs being the only part of the boat supplied by the racer rather than the management. Discussion about propellers is a staple of powerboat gambling conversation, and if any racer changes his propeller prior to a race, the fact is duly announced over the public address system.

The Heiwajima Tickets

Very well: let us take a look at boat-race betting strategies. My admittedly crude sample consists of 103 losing betting tickets picked up off the floor at the Heiwajima boat race course on March 21, 2010. I have reproduced four of them in figure 15.

Fig. 15 Heiwajima betting slips.

Photo by author. 

Top left: 4ni-ren-fukubets @ ¥100. Total stake ¥400

Top right: 5san-ren-tanbets @ ¥200. Total stake ¥1,000

Bottom left: 1-chaku nagashi (1st place wheel). 4 san-ren-tan @ ¥400. Total stake ¥400

Bottom right: Bokkusu Box. 6 san-ren-tan @ ¥500. Total stake ¥3,000

I picked these tickets up near the end of the day’s racing, just after the 11th of 12 races. Most of them were tickets for the 11th race, some for the 10th, and just a few for earlier races (the cleaning ladies come round at regular intervals to sweep the floor).17 I present the data here for total amount staked (table 1), and number of outcomes betted on (table 2). The minimum stake is 100 yen and bets must be made in units of ¥100.

Table 1: Amount staked in Heiwajima sample

Range of amount staked

No. of tickets

¥100 to ¥500


¥600 to ¥1,000


¥1,100 to ¥1,500


¥1,600 to ¥2,000


¥2,100 to ¥2,500




Over ¥3,000




* Two bets of ¥5,000; one each of ¥6,000, ¥8,000 and ¥20,000.

Most of them are fairly small bets; the mean amount staked was about ¥1,000 or roughly $10. Partly this reflects the place where I found the tickets: like other race stadiums in Japan, Heiwajima has a separate enclosure with reserved seats where most of the wealthier punters will be found,18 but these tickets were collected in the open part of the stadium, which only costs ¥50 to access. Note, however, that even ¥1,000, staked on each of the 12 races at the meet, would be about $120 at recent exchange rates.

Table 2: Type of bet in Heiwajima sample

Type of bet


No. of tickets

ni-ren-tan [forecast]



ni-ren-fuku [reverse forecast]



san-ren-tan [tricast]



san-ren-fuku [box]#



nagashi [wheel]***



san-ren-tan bokkusu [box]#






* General mathematical odds, not taking account of form etc.

** One ticket had a mixture of ni-ren-fuku and san-ren-fuku bets.

*** 29 wheels on 1st place, 1 on 2nd place, 1 on 3rd place.

# The san-ren-fuku and san-ren-tan bokkusu are virtually identical. They are both three-contestant box bets, i.e. you bet on all six possible combinations of three contestants to fill the first three places. But with a san-ren-fuku bet you can place just ¥100, effectively betting ¥16.67 yen on each of six outcomes; whereas the san-ren-bokkusu requires a minimum stake of ¥100 on each outcome, so ¥600. The pools for the two types of bet are different, so the odds will also differ.

These results are very suggestive. There were no win or place bets. After those, the most conservative bet (best chance of winning, usually with correspondingly lower odds) is the reverse forecast (ni-ren-fuku), which picks the first two finishers from six contestants in any order, giving general mathematical odds of 1/15. But only a tiny minority of punters selected this kind of bet. The forecast (ni-ren-tan), also with relatively low odds, was also very unpopular. Far and away the most popular type of bet was the simple tricast (san-ren-tan), which is also by far the toughest one to make, with odds of 120 to 1. Including the boxes and wheels (which are multiple tricast bets), tricasts made up 96 of 103 bets.

This would appear to suggest that these punters in the cheap part of the Heiwajima stadium favored a high risk / high return approach to gambling.19 It seems likely that the tricast is the dominant type of bet in Japanese race gambling more generally, judging from the large number of books dedicated to the specific theme of how to make a winning approach to tricast betting. On July 27, 2010 I counted 67 books with ‘san-ren-tan’ in the title on sale at Amazon Japan against 39 with ‘san-ren-fuku’ and none with ‘ni-ren-tan’ or ‘ni-ren-fuku’ in the title. The humble forecast bet apparently lacks the mystique to have books written about it.

This propensity for high risk/high return betting may also explain one little mystery I noted earlier about powerboat betting: the almost complete absence of win and place bets. The official publications of the BRPA, and their various web sites, clearly state that win (tanshō) and place (fukushō) bets are available, yet the sample betting slips displayed in these same materials do not have any box to check for these two types of bet. My research indicates that (1) it is impossible to make a win or place bet at an off-course betting center; (2) most boat race stadia have a single, hard-to-find location that has win and place betting slips. The betting slips are not on display and you have to ask for them specially. The reason usually cited is that there is almost no demand for this kind of bet, because of the low odds. Yet this same kind of bet is the most common one found in the UK.

However, this impression of a risk-loving class of punters is immediately tempered by a couple of other items in the data. First, these punters are not risking much money on each race. The relatively small amount of money usually staked means that taking on the long odds of the tricast is the only way to make a significant return – or to put it bluntly, the tricast’s popularity may reflect the simple truth that punters with small budgets cannot hope to make serious money on short-odds bets. This could explain the virtual non-existence of the low-odds win and place bets. Though very rare in boat and bicycle racing, these bets are popular enough in Japan in horse-racing. Firstly the larger number of contestants makes for longer odds on single-horse bets; secondly, there are more wealthy punters, who might like to invest ¥10,000 to win ¥20,000. But people who can only afford to stake ¥100 need much longer odds to make it attractive.

Second, the popularity of the wheel and box bets must give us pause for thought. These exotic bets are really just multiple tricasts on the same race. For example, a wheel bet on boat #1 to come first, boat #2 second, and any of the remaining four boats to come third is really four tricasts: 1-2-3, 1-2-4, 1-2-5 and 1-2-6. So these gamblers are essentially hedging their bets – reducing the odds against their succeeding, but also reducing the amount they stand to win because they are dividing their stake money between different outcomes. That is not the strategy of a risk-loving gambler.

At this point I should note that many of the bets classified as tricasts in table 2 were also hedged, in that the punter had wagered on more than one tricast on the same ticket. This brings me to the important consideration of multiple-outcome betting. Table 3 shows the number of outcomes selected in each ticket in the Heiwajima sample.

Table 3: Number of outcomes forecast in Heiwajima sample

No. of outcomes forecast

No. of tickets





















* The reason for this high number is that the popular tricast box (first three finishers in any order) amounts to betting on any of six outcomes: A-B-C, A-C-B, B-A-C, B-C-A, C-A-B, C-B-A.

** One ticket with a complex selection of ni-ren-fuku and san-ren-fuku bets.

The meaning of table 3 is that 98 out of 103 punters were hedging their bets – offsetting their addiction to high-risk tricasts by buying many of them on the same race.

The Heiwajima punters spurn the low-risk win/place bets totally and few of them use the relatively low-risk forecast. Instead they overwhelmingly prefer the higher risk tricast. They make a bold statement – they are high-risk, devil-may-care gamblers. But then they look down into the abyss – at the strong likelihood that they will simply lose their money – and they pull back from the brink. Having chosen one unlikely but possible outcome, the thought occurs to them that a different trio might cross the line in front – and they bet on that outcome too. Many of them end up betting on six or more tricasts – which is actually a more conservative approach than choosing a single forecast bet (odds of 1/20 for six tricasts, against 1/30 for one forecast).

These gamblers come across as very serious – one intuitively imagines that people who bet casually, for fun, would not use such complex bet-hedging strategies. They also appear psychologically compromised, caught between the two stools of wanting to win big and wanting to win frequently. Their uneasiness reminds me of the restless little boats themselves, with no brakes, drifting ambiguously between lanes.


What sort of outcomes will result from this kind of betting? To test the relative degree of success of various strategies empirically, I created a simple model with eight hypothetical punters.

Punter 1 always bets ¥1,000 on the 1-2 forecast (one of the most likely forecasts to come up).

Punter 2 always bets ¥200 each on the forecast wheel of 1-2, 1-3, 1-4, 1-5 and 1-6, for a total stake of ¥1,000 (betting on the #1 boat to win like punter 1, but equally favoring each of the other 5 boats to come 2nd).

Punter 3 always bets ¥250 each on the tricast wheel of 1-2-3, 1-2-4, 1-2-5, and 1-2-6 – again for a total stake of ¥1,000 (betting on the #1 and #2 boats like punter 1, but equally favoring each of the other 4 boats to come 3rd).

Punter 4 always bets ¥1,000 on a single tricast, 1-2-3 (following punter 1 in backing #1 and #2 to finish first and second, but further backing just boat #3 to come 3rd).

Punter 5 is like punter 4 but has a soft spot for the auto-ya and always bets ¥1,000 on the 1-2-6 tricast.

Punter 6 always bets ¥1,000 on the 6-5 forecast, the same type of bet as punter #1 but backing one of the forecasts least likely to come up and therefore likely to have long odds.

Punter 7 always bets ¥250 each on the tricast wheel of 6-5-4, 6-5-3, 6-5-2, 6-5-1 – betting on the #6 and #5 boats like punter 7, but equally favoring each of the other 4 boats to come 3rd.

Punter 8 is a radical who always bets ¥1,000 on arguably the most unlikely tricast, 6-5-4.

I then applied this model to all the boat races held in Japan in the week of July 17 to 23, 2010, during which there were 85 meetings, each with 12 races, for a total of 1,020 races.20 Hence each punter staked a total of ¥1,020,000. If everybody were betting in the most rational way possible, then all eight of these strategies ought to produce roughly similar results over the long term,21 since the small number of wins on unlikely outcomes would be offset by much better odds than on the more likely outcomes. The one thing no-one can escape is the 25% betting tax, so that the outcome of totally random gambling ought to be a loss of 25% (¥255,000) for all punters.

Results were as follows:

Punter 1: 118 wins; net loss of ¥65,800 (6.5% of stake)

Punter 2: 395 wins; net loss of ¥269,630 (26.4% of stake)

Punter 3: 117 wins;22 net loss of ¥96,175 (9.4% of stake)

Punter 4: 32 wins; net loss of ¥204,300 (20.0% of stake)

Punter 5: 23 wins; net loss of ¥81,800 (8.0% of stake)

Punter 6: 12 wins; net loss of ¥497,100 (48.7% of stake)

Punter 7: 12 wins; net loss of ¥523,625 (51.3% of stake)

Punter 8: 2 wins; net loss of ¥974,800 (95.6% of stake)

These results are a mixture of the obvious and the surprising. Least surprisingly, everyone has lost money. Punters 2 and 4 have lost roughly the amount one might expect, given the 25% government betting tax factored into the odds. Punter 8 is betting at such long odds that 1,020 races is not really enough to test his luck – he has lost nearly all his money, but theoretically could have won it all back with one lucky win on the 1,021st race. What is more interesting is the surprisingly good results for punters 1, 3 and 5. What they have in common is that they all back the #1 boat followed by the #2 boat with all their stake money. Punter 2, only betting 20% of his stake money on 1-2, has done much worse. The 1-2 result is probably the most common outcome of the lot – the simplest kind of nige, in which the boat in the inside lane wins and the boat just outside it comes second. Indeed, this result came up 118 times in 1,020 races or 11.6% of the time, compared with 3.3% (1/30) for the average forecast and 1.2% (12/1,020) for the least likely forecast, 6-5. If Japanese gamblers took full account of the strength of the 1-2 forecast, they would back it so heavily that the odds would be reduced to the point where the lowness of the odds offset the likelihood of the outcome. Had they done so, punters 1, 3 and 5 would have lost roughly 25% of their stake money. The fact that they actually lost just 6.5%, 8.0% and 9.4% respectively indicates that the 1-2 combination is significantly under-backed. This is a consistent pattern over the whole 1,020 race sample.

If it is the case that Japanese gamblers under-back likely outcomes, one would expect that conversely they would over-back unlikely outcomes. I tested that hypothesis with punters 6, 7 and 8, who consistently backed the long-odds outcome of 6-5. The heavy losses sustained by all three appear to confirm the hypothesis. The longer odds available on outsiders are not as long as they should be, given the extreme unlikelihood of the outcome, because too many people are aiming for them, which brings down the odds in the pari-mutuel system. This pattern may well reflect the numerological instinct of many Japanese gamblers to aim for a deme (pron. deh-meh, the kanji being emerge+eye) – the idea that a number which has not come up for a long time is more likely to come up next time.23 Most people know that this is a fallacy even rolling dice or tossing coins, but in the case of powerboat racing, the conditions that produce a string of wins for low or high numbers are likely to produce more of the same result, making deme thinking even more problematic. With the deme approach, if the #6 boat consistently loses, people will gradually become more, not less likely to back it… producing the disastrous results shown by punters 6, 7 and 8.

Ironically, Higaki, in his data-free meditation on the meaning of gambling, flirts with the opposite idea – that outsiders may be under-backed and favorites over-backed, with the result that punters and tipsters who specialize in picking outsiders may do relatively well over the long term (Higaki 2008: 55-56). Unless gambling psychology is radically different between boat and horse racing, Higaki’s argument, in a book that bewilderingly mixes sophisticated philosophical theory with sometimes rather naïve thoughts on betting, is unconvincing.

My own tentative conclusions are that (1) Japanese boat-race gamblers tend to prefer complicated betting strategies to simple ones, being far more likely to bet on multiple tricasts than a single forecast; and (2) They tend to under-back favorites and by corollary to over-back outsiders. The former habit is harmless – the comparison between punters 1 and 3 shows that (as one would expect) over the long term the returns on betting a single forecast (e.g. 1-2) will be roughly the same as betting the four tricasts that carry the same condition of success (in this case, 1-2-3, 1-2-4, 1-2-5 and 1-2-6). Likewise, punters 6 and 7 have had almost identical outcomes from betting a single forecast and the corresponding four tricasts respectively. But the latter habit means that many punters, by exaggerating the frequency of wins by outsiders, will lose more than 25% of stake over the long term, enabling conservative favorite-backers to do significantly better – though not, alas, sufficiently better to overcome the massive handicap of the 25% betting tax. The net result is that on any betting day the vast majority of punters will go home losers, and over a number of repetitions, only a very exceptionally brilliant gambler – or one with inside knowledge – will be able to come out ahead.

Comparative Study: UK Greyhound Betting

Despite the many obvious differences, UK greyhound racing resembles Japanese powerboat racing in one key aspect: there are always six contestants,24 unlike the highly varied fields in horse-racing. In some senses it also resembles powerboat racing in sociological terms, being patronized mainly by working-class people, without the upper-crust element found in horse racing. I have therefore chosen to briefly look at greyhound racing in the United Kingdom, making several research trips to the Oxford Greyhound Stadium at Sandy Lane, Blackbird Leys, and to off-course betting shops.

The first thing that strikes one is the far greater range of betting options available to the British punter. It is perfectly possible to bet on just the winner, or to bet on a dog to place (finish first or second). I have outlined above how these simple, low-odds bets are barely even on the menu in Japan. At the other end of the scale, British bookmakers offer bets of such complexity that they make Japan’s boxes and wheels look like child’s play. These include doubles, trebles, accumulators, and accumulator derivatives such as the Yankee (four selections, 11 ways of winning), Canadian (five selections, 26 ways of winning), Heinz (six selections, 57 ways of winning, like Heinz’s 57 varieties), Super-Heinz (seven selections, 120 ways of winning) and Goliath (eight selections, 247 ways of winning). None of these bets are on offer in Japan. What they have in common is that they involve placing bets on different races. Thus a double or treble means predicting the winners of two or three races, while a Yankee is a four-race accumulator which also allows smaller pay-outs if two or three of the selections win.25

At Sandy Lane, most of the bets are win or place bets, with a few forecasts and tricasts. I have no way of telling how many punters purchase multiple bets on the same race, because you have to buy a separate ticket for each forecast, whereas in race-betting of any kind in Japan you can do it with a single ticket. The British system is not designed to encourage the punter to gamble on numerous different outcomes. It would raise an eyebrow if somebody bought eight different tickets for the same race at Sandy Lane, because most people buy just one. At the Japanese boat-races, however, you can bet on many different outcomes with a single ticket. whereas kyōtei tickets are designed to encourage this strategy.

Another crucial difference is that the UK has legal bookmakers, who set their own odds. As discussed above, this pits punter against bookmaker, and entails the possibility that the bookmaker may make a deficit on any given race if it errs in setting odds or if a lot of money is bet on a long-odds outsider. That can never happen in tote betting, since a lot of money being bet on a long-odds outsider would automatically shorten the odds. In the UK, the tote is available as an alternative to the bookmakers; Japan has only the tote. At Sandy Lane there are usually four bookmakers competing for business. They offer only the simple win bet, have a minimum stake of £5, and compete for business by offering slightly different odds on some of the dogs.

I conclude, therefore, that Japanese boat-race gambling is a case of what my old mentor Maurice Bloch might call “arthritic communication.”26 If the act of betting is a form of communication between the individual and the collective, then it is a relatively limited conversation in Japan. The punter is effectively excluded from the most conservative strategies (which my data suggests are also likely to be the most successful strategies) because, at the off-course betting centers at least, he is not allowed to just pick the winner. That ticket is not on sale. It is on sale at the courses themselves, but as mentioned above, outlets are few and it is actually quite difficult to buy the ticket. The punter is also excluded from very high-risk/high-return wagers on multiple races. Those tickets are not on sale either. That leaves him with the medium-risk ni-ren-tan or the high-risk san-ren-tan. He usually picks the latter because he has only a small amount of money to stake and needs long odds if he is to win any significant amount; but he then tries to hedge his bets by buying multiple tricasts. Despite the appearance of conservatism or sophistication generated by the use of multiple bets, the fact remains that he still tends to bet too much on long shots, resulting in substantially worse results than if he stuck to backing favorites. Systemic framework meets individual character and produces a pattern of betting that conceals elementary errors behind the complexity of the strategy.

Finally, a word about atmosphere. Sandy Lane is a small-time dog track, with no more than 500 or so people attending most meets. However, when the dogs come home, there is a certain excitement in the air, and people will sometimes shout for joy if their selection has won. The atmosphere at Heiwajima, and even more so at the Yokohama Boatopia, is quieter. There is a modest increase in noise as the boats approach the line, rapidly fading away as they cross it. I see two factors in the subdued atmosphere. First, the tendency to bet on long-shot tricasts means that at the end of each race the great majority of those watching will have lost their money27 – a greater majority than at Sandy Lane, where the large number of win and place bets means that a relatively large number of people will be winners. Second, there will sometimes be a few people who have won a substantial amount of money by successfully backing an unlikely outcome. Those people will hide their excitement for fear that thieves or hard-up friends will notice their joy, draw the obvious conclusion, and attempt to steal or cadge the winnings from them. Higaki confirms this, adding that “losers are the stars of the place” (the racetrack) (2008: 53). The difference between someone betting £5 on a greyhound to win at 3 to 1, and someone betting ¥500 on a combination of tricasts that all depend on the race being won by a rank outsider, is that the latter is more cloaked in seriousness and feels more like science and less like fun. One is reminded of Roland Barthe’s famous description of a pachinko parlor: “the parlor is a hive or a factory - the players seem to be working on an assembly line. The imperious meaning of the scene is that of a deliberate, absorbing labor” (Barthes 1982: 27).

This naturally leads one to wonder about the relative likelihood of Japanese punters to get into trouble and become compulsive gamblers. It is a problem little discussed in Japan and I cannot find any official data on the subject. Hahakigi (2004: 57-58) estimates that there are two million compulsive gamblers in Japan, but his grounds for coming up with this figure are highly speculative and unconvincing – they mainly consist of taking bits and pieces of data from other industrialized countries and assuming that Japan is roughly the same. Inami (2007) gives a figure of 1.6 million “latent patients” – but this time the grounds are not even discussed in the book. The Japanese version of Gamblers Anonymous was only launched as recently as 1989, but now has 114 active groups in 44 prefectures.28 Yokohama’s One Day Port (Wandēpōto), founded in 2000, is a rare example of a residential facility devoted specifically to gambling problems. It can only handle about thirty patients at a time and they come from all over Japan. Given the enormous scale of gambling in Japan, one strongly suspects that the issue is not being adequately tackled by the welfare state.

Another Approach

Ichikawa Kazuyuki, ‘Kaz’ to his friends, is the proprietor of a small coffee shop in Yokohama. He came to boat-race gambling relatively late. He has been betting on boats since 1988, when he was 34. “I’ve gambled on horse and bicycle races since I was 15 or 16. I used to think boat racing was boring, because the patterns are limited and the result is often decided on the first bend, but I gradually got to like the simplicity of it. It’s all about the struggle to grab the best course. I still bet on the other sports occasionally, but nowadays the boats are my main thing.

“Boat racing is more interesting because racing on water introduces an element of uncertainty. With six propellers churning up the water, collisions can easily happen, and a weak contestant always has a chance of beating a strong one. That’s when you can win on a rank outsider (ōana, lit. ‘big hole’). I think: “there’s no way this guy can win” – then I go and bet on that guy. And occasionally a couple of the front-runners will bump into each other and the no-hoper will run away with it. Sometimes the shape of the race favors the outsider, and sometimes he’ll get lucky and draw a good engine.”

He used to go to Heiwajima regularly, but nowadays he says he doesn’t have the time or the money to spare. He is mostly busy running his coffee shop. But he follows the races on closed-circuit TV and places his bets via the internet.

In placing bets he takes account of three factors: first, astrology and numerology, second form, and third, local conditions. For example, he generally bets the 1-3 forecast when the day is Taian, the most auspicious of the six days in the Taoist week (rokuyō), which came to Japan from China and is still noted on many Japanese calendars. On Butsumetsu, the most inauspicious day, he will often bet the 6-1 forecast – it may be an unlucky day, but someone has to win a boat race, and a day with accidents is a good day for the outsider in lane 6. He pays no attention to the other four days in the rokuyō Taoist system. He does, however, pay attention to senjitsu – an indigenous Japanese system of lucky and unlucky days separate from both the seven-day and six-day week systems, thought to have developed during the Edo era. One of them is Sanrinbō, a day which comes roughly twice a month, thought to be a bad day to start building a house or to set out on a journey. On such a day, Kaz will consider the 6-5 and 1-5 forecasts. Or, if betting on March 21, he would consider the 3-2-1 tricast, based on the date – 3/21.

Yet another calendrical system often used in Japanese fortune-telling is based on the planets, matched with numbers and colors. Despite its name – Kyūsei – literally nine stars or nine planets – it refers only to the five planets used for naming days of the week, combined with numbers and colors, so that some planets, and the color white, appear more than once in the cycle. For example, if the day is Shiroku-Mokusei (four-green-Jupiter) he will focus on the #4 and #6 boat (which goes in green). He would be looking at forecasts or tricasts starting with #4 or #6 and with #1 coming second or third. With its inside lane, the #1 boat is always likely to be a strong challenger, but Kaz observes that you get better odds if you bet on #1 being overtaken by one or two boats than if you bet on #1 holding them off.

Sometimes Kaz will combine different calendrical systems. Thus if Butsumetsu in the Taoist six-day week coincides with Roppaku-Kinsei (six-white-Venus) in the nine-day week, he will definitely be betting on the 6-1 forecast. The #1 boat is always white, of course. Kaz claims that a lot of punters who include calendrical elements, or lucky number combinations from mah-jongg, in their betting strategy – especially old people, “because they’re old enough to know they’re not going to win.” Kaz’s friend, Mr. Ueno the dentist, disagrees. He thinks Kaz’s approach is eccentric. Ueno san is also a keen gambler and at the time of this conversation was looking forward to making a trip to Osaka especially to see a powerboat meet at Suminoe where he hoped to catch a glimpse of Nishimura Ayumi, a particularly glamorous female racer classed at A2. He used to follow the horses, and claims to have won ¥200,000 on a ¥500 bet one time, but now he favors mah-jongg and the boats. He and Kaz both think that powerboat gambling is less corrupt than other race betting, and fairer – you get your stake refunded if your contestant is disqualified, which is not always the case in horse racing. Ueno recalls an incident at the Kawasaki horse races where the second favorite fell over in its stall at the start and the bets were not refunded. Angry punters set fire to the stadium.

But Kaz’s strategy is not merely mystical. He is also knowledgeable about form and local conditions. For example, Heiwajima is the only tidal boat-race stadium in Japan. When the tide is going out, the course favors the outer boats; when coming in, it favors the inner boats. “That’s just basic knowledge.” Likewise, a headwind will favor the outside channels and a following wind will favor the inner lanes. “Basic.”

Kaz also stresses that the unit of account on which he assesses his gambling success or failure is not the individual race, nor even a day’s racing, but an entire meet. Typically a meet will last four to six days. The same pool of racers will be pitted against each other in various combinations, as preliminary heats lead to semi-finals and finally the championship race. Typically each contestant will race a couple of times a day. Form on previous days at the meet counts for far more than form at other courses. There will be more mug bets on opening day. For instance, punters on day one will look at old-timers like Mantani Akira or Katō Shunji (see above) and assume that these old-timers have no chance. That is when Kaz will back them, knowing that they are both doughty performers despite their age – and that these days they race at B-1 level, often against relatively inexperienced opposition, or in senior races for the over-fifties. They are also good mechanics, which as mentioned above is a factor in boat races. If Mantani or Katō manage to win on day one, the mugs will start noticing them and back them on later days – by which time Kaz will be looking elsewhere for his selections.

Kaz’s approach seems like a bewildering mix of arcane superstition and hard-nosed knowledge-based calculation. Kaz will argue his corner, however. “I have no money in the bank. I cannot afford to stake thousands of yen on each race. Usually I stake just one or two hundred yen, and my daily budget is ¥1,000. If I don’t get a win in the first five or six races, I quit for the day. If I bet scientifically, based on the form and lane placement of each contestant, I would end up mostly backing favorites, with short odds. What is the point of risking one hundred yen to win two hundred yen? Backing favorites only makes any sense if you can afford to stake big money. For us small-timers, the best approach is to consider the most likely outcome of the race and then bet against it. That is the only way to get decent odds. I don’t have to win every time. So long as I am betting long odds, one win in ten races is more than enough.”

That is perfectly true. On average, if Kaz bets forecasts, he will break even if he gets one win in every 22.5 attempts (one in 30, minus 25% tax). With tricasts, one win in 90 attempts will suffice (one in 120, minus 25%).

“If I do get a win or two, I will carry on betting and up my stake for the remaining races. If I make ¥30,000, I take it out of my betting kitty and use it to buy something nice – something for the shop, or the house. I wouldn’t use that money just for everyday shopping.”29

TG: “Do you consider yourself a good gambler?”

KI: “I get 75% of my stake money back, same as anyone else. It’s all statistically predetermined. I don’t consider myself a gambler. I just do it for fun. That’s the thing about boat racing – you can play for a hundred yen. Bets on the horses start at three hundred.”

Fig. 16 - Advertisement for night racing at Kiryū

(Gunma pref.) Blood-sucking themes are not unknown in boat-race advertising.

Kaz will clearly never get rich from boat gambling. He knows that overall he loses, since he cannot beat that 25% government edge. But there is enough satisfaction in occasionally winning that it does not matter if he loses overall. It is a small, daily, intellectual challenge. It gives him something to think about during the long hours when he has few customers. Following the fortunes of the powerboat racers is like following a very long-running soap opera – with old men and pretty girls in the cast, it adds a splash of color to everyday life. And so I would argue that, however perplexing his gambling strategy may seem, Ichikawa Kazuyuki demonstrates a rational approach to powerboat gambling.

Tom Gill is a professor at Meiji Gakuin University. He is the author of Men of Uncertainty: the Social Organization of Day Laborers in Contemporary Japan and co-editor of Globalization and Social Change in Contemporary Japan.

Recommended citation: Tom Gill, "Those Restless Little Boats: On the Uneasiness of Japanese Power-Boat Gambling," The Asia-Pacific Journal, Vol. 11, Issue 42, No. 2, October 21, 2013.


Barthes, Roland. 1982. Empire of Signs. New York: Farrar, Straus and Giroux.

Bloch, Maurice. 1989. Ritual, History and Power. London, Athlone Press.

CRS (Central Research Services Inc.; Chūō Chōsa-sha). 2006. “Gyanburu ni kan-suru yoron chōsa” kekka no gaiyō (Summary of the opinion poll on gambling). Tokyo: CRS. Accessed on-line at this location, July 27 2010.

Gill, Tom. 1998. “Transformational Magic: Some Japanese Superheroes and Monsters.” In The Worlds of Japanese Popular Culture: Gender, Shifting Boundaries and Global Cultures, ed. D. P. Martinez. Cambridge: Cambridge University Press, pp. 33-55.

----------. 2001. Men of Uncertainty: the Social Organization of Day Laborers in Contemporary Japan. Albany NY: SUNY Press.

Hahakigi, Hōsei. 2004. Gyanburu izon to tatakau (Battling against gambling addiction). Tokyo: Shinchōsha.

Higaki, Tatsuya. 2008. Tobaku: Gūzen no tetsugaku (Gambling: The philosophy of chance). Tokyo: Kawade Shobō Shinsha.

Inami, Mario. 2007. Byōteki gyanburā kyūshutsu manyuaru (The pathological gambler’s rescue manual). Tokyo: PHP.

Miyoshi, Madoka. 2009. Bakuchi to jichitai (Gambling and local authorities). Tokyo: Shūeisha.

Nagashima, Nobuhiro. 1988. Keiba no jinruigaku (Anthropology of horse-racing). Tokyo: Iwanami Shoten.

----------. 1998. “Gambling and Japanese Attitudes to It.” In Sepp Linhart and Sabine Fruhstuck eds., The Culture of Japan as Seen through Its Leisure. Albany, NY: SUNY Press.

Suzuki, Takashi. 2008. Rēsu o yomu kyōtei puro no hiketsu (Secrets of boat racing professionals who read the races). Tokyo: Sankei Shobō.

Tanioka, Ichirō. 1996. Gyanburu fībā: izonshō to gōhōka ronsō (Gamble fever: dependence and the debate on legalization). Tokyo: Chūō Kōronsha.


1 On April 1, 2010, the official name was changed from kyōtei to bōtorēsu (boat race), in a feeble attempt to improve the sport’s image.

2 These figures are handily gathered together by Keirin Zakkō (Bicycle Race Miscellanea) at this location.

3 Nihon Seisansei Honbu (Japan Productivity Centre), Rejā Hakusho 2012-nen (Leisure With Paper 2012).

4 ¥0.91 trillion in FY 2012, down 9.1% from FY2011 (Ministry of Internal Affairs and Communications).

5 The tendency for each form of gambling to be supervised by a different ministry is related to the need to share out the spoils of amakudari jobs for retired bureaucrats.

6 Kaiji shisō no fukyū; it sounds just as odd in Japanese as in English, but in practice involves publishing books and periodicals about shipping, fishing, maritime environment etc.

7 This information comes from the Nippon Foundation homepage.

8 For background on academic controversy surrounding the Nippon Foundation and one particular recent case, see ‘History on Trial: French Nippon Foundation Sues Scholar for Libel to Protect the Honor of Sasakawa Ryōichi’ by Karoline Postel-Vinay with Mark Selden.

9 In recent years, declining revenues have turned some courses into loss-makers. A number of local-authority-run horse-racing and bicycle-racing tracks have been closed in recent years. However, as I write, no powerboat stadium has been closed down since Handa stadium (Aichi pref.) in 1964. The number of sites has been steady at 24 for many years, roughly half the number of velodromes, and this limit on numbers may help explain this industry’s ability to weather economic recessions so far.

10 Yokohama’s population is about 3.7 million, about a million more than Osaka’s.

11 This often makes no difference, but it is crucial on those rare occasions when a very high-odds result occurs. For example anyone betting ¥1,000 on the 6-4-5 tricast for the 7th race on July 20 2010 at Shimonoseki boat arena would have won ¥2,305,600 at the course or a legal betting center, but only ¥100,000 at a yakuza gambling den.

12 This is unusually high by international standards. In Britain, the betting tax was 9%. The Labour government of Tony Blair abolished the tax in 2001 in a bid to stem the loss of business to tax-free offshore betting shops. Since then there has been a sharp increase in race gambling in the UK. Miyoshi (2009: 73) states that in 1955, the first year of racing at the Edogawa boat stadium, takings were ¥564 million and payouts totaled ¥549 million, for a gross profit of just ¥15 million. Punters are not treated so generously nowadays.

13 Called ‘autorace’, this form of race gambling used to include cars as well as motorbikes, but for many years has been limited to motorbikes. It is very much the poor relation of race gambling, with a far smaller turnover than the other three.

14 I sense that class factors also influence scholars of gambling. Anthropologist Nagashima Nobuhiro of Hitotsubashi University (e.g. 1988, 1998) and philosopher Higaki Tatsuya of Osaka University (e.g. 2008) seem to devote themselves to horse-racing, neglecting bicycle and powerboat racing. Higaki even states that boat and bicycles races should be classified as a “lower class of gambling” than horseracing, because the contestants are purely human, increasing the likelihood of race-fixing. Horseracing, by contrast, is a more “natural” contest, between horses and turf (2008: 46-47).

15 A divorcee, in 2010 she remarried, to fellow boat racer Yamazaki Tomoya, but she still races under the Yokonishi surname.

16 I analyzed the results of 1,392 boat races held from July 12 to 21, 2010. The winning percentage for each lane were: (1) 40.4%, (2) 15.8%, (3) 14.9%, (4) 15.6%, (5) 9.3%, (6) 4.1%.

17 One of the tickets turned out to be a hefty 21,600 yen bet on the horses at the Hanshin racetrack in Takarazuka, several hundred miles away. Like most gambling venues, Heiwajima includes an off-course betting facility for betting on meets elsewhere in the country. I had to remove this one from the sample.

18 These seats only cost ¥1,000 or ¥1,500, so are certainly not comparable to the Royal Enclosure at Ascot. But many gamblers with limited resources would rather use that money to make another bet, and besides these seats are indoors and somewhat detached from the action.

19 Up to this point, my findings seem to confirm Nobuhiro Nagashima’s view of traditional Japanese gambling, in which he contrasts light, playful gambling by the upper classes with heavy gambling by the working class, which he relates to male machismo and the doomed romanticism of the cherry blossom – “a swift fall after a brief climax.” He continues: “It formed the core ideology of yakuza culture, expressed in the idiom otoko ni naru – to become a man. This ethic of high risk/high return later dominated the gambling and racing scene and brought to it the image of bad manners” (Nagashima 1998: 351-352). Nagashima says the moral disapproval of many Japanese for gambling is encapsulated in the word shakōshin, meaning the desire to win a fortune at a single stroke (ibid. 356).

20 Results of Japanese powerboat races may conveniently be viewed at this site.

21 Whether 1,020 repetitions is enough to be considered “the long term” depends on the probability and standard deviation of each wager. It is good enough, I would argue, to offer at least suggestive evidence for all except the most long-odds strategies.

22 One would expect punters 1 and 3 to have the same number of wins, but there was one race in this period where the 1-2 forecast won, but too many boats were disqualified to give a result for the tricast, so that tricast stake money was returned.

23 For example, one powerboat blogger writes: “… I tried to get a result by betting on the deme in the second half. It was a struggle to decide what to go for, but at Toda [a powerboat stadium in Saitama pref.] you generally get a reasonable return if the in-course can get away, and since #1 hadn’t come up in the first half, I bet 1-2-6 and 1-6-2 for the rest of the races…” ("Aqua Boo" Kyotei Blog accessed July 31 2010).

24 At most tracks; five at some smaller tracks.

25 In April 2007, the Motorboat Race Law was revised to allow promoters to offer multi-race (jūshōshiki) bets. Nobody in powerboat racing has yet taken advantage of this deregulation, but we are seeing it in bicycle racing. Miyoshi reports (2009: 185-188) that since April 2008 at least two velodromes (Hiratsuka and Tachikawa) have been offering accumulator bets on bicycle races. However, the odds are so long on these bets that they are marketed as a “bicycle lottery” (chariroto). The one at Hiratsuka even involves buying a ticket with the bets randomly generated by computer, further blurring the line between race betting and lotteries. Possibly the boat promoters are more conservative because they are not in quite such dire financial straits as velodromes, many of which are in the red and facing closure.

26 A rather liberal interpretation of his paper on ‘Symbols, song, dance and features of articulation.’ In Bloch 1989.

27 Higaki writes: “When it comes to gambling, paradoxically enough, losing is the natural state of affairs” (2008: 53; my translation). No doubt that is generally true, but it is significantly more true at Heiwajima than at Sandy Lane due to the tendency of the Heiwajima punters to favor betting at long odds.

28 Gamblers Anonymous Japan Info Center

29 Here Kazuyuki shows a variant on the Japanese proverbial wisdom, abuku zeni wa mi ni tsukanai (“money acquired easily does not last long,” somewhat like “easy come, easy go”). Many Japanese gamblers think that money won by gambling should be spent that day, rather than being saved or used for everyday purposes. But they would be more likely to spend it on women, drink or luxury goods than on useful things for the home or shop.

We welcome your comments on this and all other articles. Please consider subscribing to our RSS feed, or following us via Twitter or Facebook.

Add comment
Authors: For all articles by the author, click on author's name.   Tom Gill