06/23/2005 11:00PM

Take a step back before using those stats


LEXINGTON, Ky. - A valuable tool was added to the Daily Racing Form past performances on June 15. Two categories of jockey-trainer combination statistics are now shown beneath the company lines in the lower right hand corner of each horse's past performances. The first category shows the number of starts, the win percentage, and the return on investment on $2 win bets for every race the jockey rode at that track for the horse's trainer since Jan. 1 of the previous year. The second category lists those same statistics for that jockey-trainer combination at races run at all tracks during that same time period.

Handicappers must decide the best way to use this information. First, they must decide which of the two groups of information they will give the most attention. The statistics geared to the track at which the horse is running on a particular day should be preferred if the sample size is large enough to be meaningful. If not, the combined data from all tracks should also be considered in the hope that it will support the general direction of the trend you see in the smaller sample. If it contradicts the local trend, however, caution is advised.

Let's consider a few scenarios. Do you like high win percentages? Some of them will be eye-catching, but a couple of things must be considered. First, how large is the data sample? One win from two races for a jockey-trainer combination produces a flashy 50-percent win rate, but the sample size is too small to be significant. A 33-percent win rate from 90 races is significant, but you must also consider the amount the winners paid. If the return on investment is $1.50, you don't want to be on board with a jockey-trainer combination that wins often, but is overbet. Combine that same win percentage and sample size with a $3 ROI, though, and you have a jockey-trainer combination that can make you money.

A return on investment of $8 or more looks great at first glance, but if it is the result of a four-race sample, it is going to be deceptive. A 10-percent win rate from 20 races with a $3 ROI might seem like a reasonable statistic to someone quickly skimming through the numbers, but it represents only two winners. If that same win percentage and ROI can be found in a group of 150 races, those horses are worth betting.

Even when you are satisfied that you have found a jockey-trainer combination that has been profitable over a good sample size in the past, and seems likely to continue to be profitable, that doesn't mean it should be bet indiscriminately. If the stable's go-to jockey has been aboard a horse in his last three or four starts but hasn't been effective, there is no reason to be optimistic next time out unless there is a significant class drop, equipment change, medication change, etc., that provides a sound reason to hope for improvement.

Another factor to consider with profitable jockey-trainer combinations is where the gains most likely came from. If you see a $2.50 ROI on a 200-race sample, don't assume you will be getting that same ROI on the 4-5 favorites you bet when that jockey and trainer hook up. The more likely scenario is that the horses whose merits were a bit less obvious were responsible for the majority of the gains in that group.

These profitable combinations are even more powerful when they occur in situations where the trainer also shows a profit in the applicable category in the trainer statistics located beneath the horse's workouts. If the trainer from the jockey-trainer combination with the $2.50 ROI on a 200-race sample also shows 25 percent wins and a $2.80 ROI on a group of 80 of his first-time starters, you have a very strong bet with his go-to jockey aboard his first-time starter.