The Best

Money won is twice as nice as money earned; and money earned through manipulation is best of all.

The best manipulation I think occurred in 2002 when a group of enterprising young fellows, fraternity brothers at Drexel University, gained access to the Pick-6 betting data in 2002 and were able to in effect past-post their selections to the point of eventually winning the entire substantial Pick-6 pool (including the consolation payout) when a true long shot (43-1) named Valponi crossed the finish line first in the Classic, the final leg of the betting sequence.

One of the perpetrators worked for the computer people who ran the Pick-6 data transmissions and exploited a weakness in the system that allowed him to change wagers mid-Pick-6 before the entire race sequence was run. The scheme worked. However, they became the only winners of the pool, never playing a losing combination, and came to the immediate attention to those who pay attention to that sort of thing. They did prison time.

Recently, consider whoever played with a quinella pool at Gulfstream Park and made a quinella payoff be $42 when the exacta for the top two choices paid $18.

It is not known how much this Houdini might have netted for their efforts, but they pumped $18,000 into a quinella pool of $24,280 that generally tops out at $2,000.

They didn't fix the race; they fixed the payout. A whole other animal, and apparently one that is not considered illegal, but is certainly worthy of the track's attention.

A quinella is one of those obscure, forgotten bets. It involves picking the first two horses to cross the finish line in any order. It is sort of a boxed exacta, a permutation of two bets, into one bet, a combination. It generally pays half of an exacta and is available on select races.

I think the bet type might have migrated from jai alai to horse racing. I always forget about the bet, concentrating on win bets and exactas. My departed friend Fourstardave was good at exploiting them when their will-pays were going to be lopsided in relation to the exacta payouts.

It turns out the individual pumped $18,000 into a quinella pool that usually only sees $2,000 in it. They purposely picked unlikely combinations, which in effect would produce higher than expected payout for what might actually be the outcome if the two favorites ran one-two, which that did.

Pari-mutuel wagering is predicated on payouts that reflect the ratio of winning bets to that of losing bets in the pool. Usually long shots have little placed on them, so if they win, the bettors in effect get everyone's else's money that didn't use them. Tote board odds are not predicated on the statistical probability of something coming out that way; they are a function of the wagering distribution.

Gulfstream has suspended quinella wager for now. The feeling is that the perpetrators placed numerous bets with off shore accounts on the favorite combination that rather artificially was made to pay $42.

My departed buddy Foustardave was good at noticing and exploiting quinella/exacta anomalies. I pretty much completely forget about the bet's existence on selected races, instead concentrating on win and exacta bets.

What I'd love to see, even before this, is a real-time app that lets you determine the will-pays for exactas that you are interested in, rather than waiting for the tote board's scroll-cycle to reveal the possibilities. A constant readout of your own choosing.

Even better would be an app that compares the exacta will-pays with the quinella will-pays side-by-side. Say a will-pay for a 1/2 exacta compared with a 1/2 quinella. Right now they're there, but you have to align the comparison for yourself.

Generally, quinellas can be counted to pay the half the exacta. But it would be nice to know if there are other forces at work that are going to upset that expectancy.

http://www.onofframp.blogspot.com