The superfecta takes the trifecta one step further, demanding that you come up with the fourth-place finisher as well as the top three. The bet made a brief appearance in American racing during the 1960's and 1970's but fell from favor, first because of a lack of popularity due to its cost and difficulty, and then because it was tainted by race-fixing accusations surrounding the bet at New York harness tracks in the 1970's.

In a typical eight-horse harness race, there are 1,680 (8 x 7 x 6 x 5) possible superfecta combinations. Some race-fixers figured out they could lower their costs and ensure their likelihood of cashing by bribing drivers to hold back two or three horses from finishing fourth or better. Now the crooks only had to box the remaining five (5 x 4 x 3 x 2 = 120) or six (6 x 5 x 4 x 3 = 360) trotters or pacers, and they were guaranteed a profit since the holdbacks included a favorite or two.

The cost issue of the superfecta is in the process of being addressed radically with the recent introduction of 10-cent minimum bets, which are discussed separately below. The importance of this current experiment cannot be overstated, because it actually makes superfectas more affordable than trifectas, has serious implications for the taxation and withholding issues addressed in Chapter 5, and may transform the entire exotic-betting landscape if these lower minimums are extended to other exotic wagers. First, though, let's consider the attractions and drawbacks of the bet at any minimum level.

When deciding whether to extend from a trifecta to a superfecta play, a bettor faces many of the same issues he does when considering whether to step up from an exacta to a trifecta. The difficulty and number of possible outcomes increases by a factor of however many additional horses there are in the field to fill the final slot. You might, for example, switch from a three-horse exacta box to a 3 x 3 x Something trifecta part-wheel if you have a clever underneath idea, or if you're playing against a favorite or two and think you can get them out of the money.

Those decisions become blurry when choosing between a trifecta and superfecta, primarily because we are now talking about the fourth-place finisher in a race, a distant placing less likely to be pinpointed by your handicapping. Also, it's one thing to try to bet against a favorite to win or get him out of the exacta, but significantly more ambitious to decree that he can't even finish fourth.

The whimsy or desire for action that might lead you to try to turn an exacta into a trifecta becomes awfully expensive with an additional slot to fill in the superfecta. A 3 x 3 x 9 trifecta part-wheel costs only $42 (3 x 2 x 7) for $1, but a 3 x 3 x 9 x 9 superfecta is six times as much at $252 (3 x 2 x 7 x 6). Even if you're willing to spend that much, it is not going to pay six times as much as the trifecta unless something really interesting happens for fourth place. Unless you have a good reason to think that's especially likely, you're back to gambling and hitting the "all" button.

One popular way to play the superfecta is to consider it as a sort of combination of a win selection and a trifecta - keying one horse to win only and then betting on the effective trifecta combinations underneath to provide the 2-3-4 finishers. Keying one horse over three others, for example, costs just $6 at a $1 minimum, but you only have six among thousands of possible combinations. The cost of keying one horse over several others is exactly the same as that for a trifecta box among the same number of "underneath" horses. Keying one horse over four others is $24, over five horses is $60, over six horses is $120, etc.

A variation of this play that makes a superfecta at a $1 minimum more manageable is a trio of double-key part-wheels, where either of your two key horses must win and the other must finish 2-3-4. You could hook up two keys with four other runners for $72 this way:

1,2/1,2/3,4,5,6/3,4,5,6 = 24 bets (2 x 1 x 4 x 3)

1,2/3,4,5,6/1,2/3,4,5,6 = 24 bets (2 x 4 x 1 x 3)

1,2/3,4,5,6/3,4,5,6/1,2 = 24 bets (2 x 4 x 3 x 1)

Adding a fifth underneath horse to this play would increase the cost to $120, and a sixth would make it $180. There will be plenty of frustrating occasions when your two keys both hit the board but fail to win, running 2-3, 2-4, or 3-4, but adding those three permutations would double the cost of the play.

The attraction of the superfecta is the possibility of a very large payout that may be a massive overlay. The cost and difficulty of the superfecta mean that smaller players will be boxing a limited number of choices and fearfully including two or three favorites, while the sheer number of combinations means that some perfectly reasonable outcomes will be overlooked and underbet.

The superfecta is the first bet we have looked at where "pool scooping" becomes a realistic possibility. Pool scooping occurs when there is only a single ticket sold on the winning combination, meaning that one person receives a payoff that may be many times higher than the pure odds would suggest it should be simply because that particular combination was lost amid the vast number of possibilities.

Here's an example of how that might work. Suppose there is a 12-horse field in a race at a midsized track where the superfecta pool is only $20,000. There are only 20,000 $1 superfecta bets to cover 11,880 possible combinations, and because so many of those 20,000 are duplicative tickets emphasizing the favorites, several hundred or even thousands of the possible superfecta combinations may have only a single winning bet on them. Many others will be uncovered completely, meaning that there will be an "all" payoff - the pool will be shared among those who have the first three finishers. If you had a 1-2-3-4 and 1-2-3-5 and nobody had the correct 1-2-3-6 combination, anyone with a 1-2-3-X ticket would be considered a winner and you would have two winning 1-2-3-all superfectas.

Part of the lure of the superfecta is the chance of receiving a much higher payout than seems warranted by holding the lone winning ticket. It does not take a procession of 50-1 shots for this to happen. In the eighth race at Gulfstream on Feb. 15, 2006, for example, the winner paid $18.40, the 4.70-1 second choice finished second, the 3.40-1 favorite finished third, and a 92-1 shot finished fourth. Even with the two favorites in there, the superfecta should have paid about $16,000 for $1. Only one ticket was sold on the winning combination, however, so after the 25 percent takeout from the pool of $61,877, the lone winning $1 ticket paid $46,407.70, nearly three times what it should have.

(It can work the other way. In the 1999 Breeders' Cup Classic at Gulfstream, the first four finishers were Cat Thief at 19-1, Budroyale at 26-1, Golden Missile at 75-1, and Chester House at 63-1. That computes to a payoff of more than $6 million, but because the pool was a mere $923,876, the lone winning $1 ticket returned "only" $692,907 after takeout, though it is unlikely the winner felt too bad about receiving such technically poor value.)

One betting inefficiency that crops up again and again when looking at superfecta payoffs is that a second- rather than third- or fourth-place finish by a longshot frequently leads to a massively overlaid return. This is true with trifectas but even more so with superfectas, where the increasing costs lead many players to use longshots in third and especially fourth positions only. Of the 15 highest superfecta payoffs in the 47 Breeders' Cup races that offered the bet from 1998 through 2004, only six featured a winner at 10-1 or more, but the runner-up was in double digits 13 of those 15 times.

My own forays into the superfecta have been brief and recent. Until 2004, the bet was only offered once a day on New York racing, which accounts for nearly all my handle outside major stakes races. Also, the superfecta was offered only on the last race of the day, an event whose outcome I was often already fully invested in because it was the final leg of the pick four or pick six. At this writing, New York now offers two or three superfectas a day on an irregular basis but has yet to adopt the change that would get me far more interested in the bet: the 10-cent minimum.

**Dime supers**

Reducing the cost of a superfecta by 90 percent obviously brings it within reach of many more bettors and eliminates the bankroll issues that either price out most players altogether or limit many of them to almost hopelessly inadequate investments. At a 10-cent minimum, a six-horse superfecta box or a 4 x 4 x 8 x 8 part-wheel costs $36 instead of $360 and allows average players to make these kinds of investments for the first time. It also allows players to purchase different combinations in different increments and to emphasize certain combinations more heavily than others, which is impractical at a $1 minimum. Someone who makes a six-horse super box for $36 could also go back and key his favored horse over the other five for only an additional $6 instead of the $60 such a key would normally cost.

Only two objections have been raised. First, a few track operators voiced concerns about betting lines being clogged by novices calling out multiple combinations. This may be a plausible argument for not offering dime supers on the handful of big-event days when tracks are actually full, or perhaps offering them only through self-service machines and via wagering accounts. It is a better argument for tracks to install more such betting terminals.

The other people unhappy about dime supers are the big bettors, who fear that their days of pool-scooping and 1-2-3-all payoffs may be over if, at least in theory, 10 times as many combinations are being sold on each race. Such fears are shortsighted. Would these same players like a return to the days of $5 minimum exactas? Dropping that minimum by a factor of five has led to an explosion of exacta pools through greater participation by unsophisticated players. In the long run, the good players will benefit if these pools are made more accessible to everyone.

The current disparity between $1 minimum trifectas and 10-cent minimum superfectas creates some interesting choices for exotics players. Suppose there's a 13-horse race where you were planning to make a four-horse 1-2-3-4 trifecta box for $24. For that same investment, you could instead make this 10-cent superfecta part-wheel:

1,2,3,4/1,2,3,4/1,2,3,4/All = 240 bets (4 x 3 x 2 x 10) @ $0.10 = $24

The $1 superfecta would need to pay 10 times as much as the $1 trifecta for this to work to your advantage. Hitting the "all" button is of course sheer gambling, and you would only want to make this play if you thought there was an unusually good chance of the favorites finishing out of the money. But for $24, there's no law against a little reckless gambling.