The track odds of a horse is the most predictive handicapping factor I know. Its downside is that it has, by definition, no wager value � the stronger the track odds factor, the lower the odds. But nobody can beat the 33% win percentage of the public�s top choice betting every single race, every single day. The public, as a whole, from a win percentage perspective (not an ROI perspective) is the best handicapper around.

So why, since the exacta pool typically has more than twice as much money in it as the win pool, are the track odds always quoted based on each horse�s percentage of the win pool? Tradition, probably. But the bulk of betting action nowadays is in the exotic pools, and it can be misleading to judge a horse only by its win odds. The exotic odds, being larger and obfuscated, are probably more predictive than the win odds. And aside from the predictiveness of money bet, if you believe in the phenomenon of smart money (at least occasionally, as I do), it�s logical that the money would go where it�s mostly hidden from the crowd: in the exotic pools.

There are some slick ways of tracking the flow of money in at least some of the exotic pools. One that bears looking at is the E-Horse Arbitrage program.  I haven�t used the program, but I know people who have, and have been told that it�s a nice piece of software. It looks for "inefficiencies" in the "horse racing market," using the stock market concept that Ziemba and Hausch translated to the racetrack in the 1980s. (For the academically inclined, see their book Efficiencies of Racetrack Betting for [lots] more information.) In everyday terms applied to the racetrack, an inefficiency just means that a horse is being bet less in a certain pool than it is in the win pool, thus its payoff is higher than it should be in that pool � sometimes, theoretically, high enough to generate a profit.

As far as I can tell, the inefficiencies in the place and show pools that Ziemba and Hausch once saw are long gone, especially with tools like the E-Horse Arbitrage program around. Plus the fact that, nowadays, a large portion of the pool has still to come in after a race has started makes searching for pool inefficiencies a shaky way to make a profit. Inefficiencies in the exotic pools � that I�m not sure about. It looks like the inefficiencies there are larger, but the smaller hit rates might make that a moot point if you go on a two-month losing streak using advantaged-but-still-losing exotic combinations.

If I were to use the E-Horse Arbitrage program, it would be to look for horses that are overbet, not underbet, in the exotic pools. I would use it as a confirming, not a primary, handicapping factor. If there was a horse I liked and it was getting bet more than expected in the exacta pool, that would make me like it even more (as a win bet, not an exacta bet).

This is a technique that can be applied in a quick-and-dirty way at the racetrack. All you need is a monitor showing the projected exacta payoffs, your track program, a pencil, and the ability to subtract and divide. Here�s what you do:

1. Look at the exacta payoffs monitor and, in the track program next to each horse except the favorite, write down the exacta payoffs for each of the non-favorites to win on top of the favorite to place.

2. If these are $1 exacta payoffs, subtract 1 from each number to get the odds for this combination. If these are $2 exacta payoffs, subtract 2 from each number and then divide it by 2 to get the odds for this combination. Then write that number down for each horse in the track program.

3. Write down the current win odds of each horse (except the favorite) in the track program. (If it�s 7/5, for instance, convert that to its odds-to-one, which is 1.4. You get this by dividing 7 by 5.)

4. Divide each horse�s exacta combination odds (the second number) by its win odds (the third number). (A calculator is allowed at this point.)

The ratios that you come up with for each horse can be enlightening. Horses don�t get bet equally in the exacta and the win pools. The lower the ratio, the more action a horse is getting in the exacta pool relative to the win pool. Again, because of their size and their concealment of each horse�s odds, I consider exacta pools more predictive than win pools.

Look at the lowest three ratios. These horses are getting the most action in the exacta pool relative to their win odds. Be careful if any of these horses are big longshots � the exacta payoffs are less meaningful for lightly-bet horses. But if any of these horses are going off at 9/1 or less and their ratio is considerably lower than the others�, it�s a factor to note.

Also, it helps to look at the raw exacta payoffs you wrote down in Step 1. The lowest payoff should belong to the second favorite. Does it? The second-lowest payoff should belong to the third favorite, and so on. If you see a change in rankings here, it can indicate betting action in the exacta pool.