FEBRUARY 11, 2009
Trainer stats don't always tell the whole storyby Steve Zacks
Good information is essential in picking winners. The better the information the more useful it becomes. This applies to trainer information as much as any other kind. When a horse does not run to his speed figure, was it the figure that was incorrect, did the good-figure race drain the horse's reserves or leave him with a problem, or did the user apply the figure incorrectly? These questions are valid and the same ones can be used to evaluate trainer data as well.
Instead of looking at trainer data purely in terms of win percent and dollar return, evaluate them from the perspective of trainer intent and place that intent within the broader context of the trainer's overall approach. Take the time to evaluate the performance of the horse after the race; not in terms of the win or in-the-money statistic, but rather did the horse run the kind of race suggested by the statistic -- good or bad.
Good trainer data will sometimes point you to a particular horse. It can tell you things about the horse, the trainer and his opinion of that horse. Suppose a key trainer has reproduced one of his productive patterns with a horse running today. The preparation, his placement in terms of class, distance and surface, and choice of rider all signal positive trainer intent. A possible wager might exist. Following pattern identification one needs to evaluate that horse's prospects; can he beat this field, or how much does he have to improve? The answers will determine if in fact a bet should be made and the reasonable odds.
This qualifying process may tell you that the horse will have to run a career best race to beat several opponents. There may be unknowns such as a change in distance, surface or equipment and a switch from bug to the hot stable go-to rider. Your database tells you that the trainer has pulled off similar upsets; you anticipate the improved race. The morning line is 6-1; you decide to make your bet if he is that, 6-1, or higher.
The horse opens up at 5-1, and then drifts upward. He goes off at 10-1. The bet is made. Your knowledge of the trainer which you have compiled over several years his says he has done part. You have identified a winning, profitable pattern for today, followed a reasonable method to qualify the horse. Now things rest in the hands of the goddess of racing!
It would be nice if your horse won. What about the information's value if he does not? Only a subjective analysis of the race and the horse's performance will give you a true understanding of the outcome. It could be that the horse won with a perfect stalking trip and really did not need to show improvement, or he won because all of the favorites were unsound. Or he ran a great race and was beaten by better horses exiting stakes races. Your post-race examination should test more than the validity of simple win percent or dollar-return statistics. The race might have been an artistic success from the trainer's perspective, with a poor result for the bet. With the proper analysis, these outcomes could be blessings in disguise for future bets and for gaining a better understanding of the trainer.
Any of the following might occur during the race; they still allow you to conclude that the horse actually did run well, as expected.
-- a trainer gets a dull layoff horse to show good speed after employing a favored and productive pattern of preparation only to see him get caught up in an unexpected and torrid speed duel after which he is allowed to fade with no pressure. The trainer pattern produced the form reversal, but the race match-up came up wrong!
-- a horse placed in an impossible spot in terms of class, distance, or surface runs a much-improved race; trainers of young horses often experiment early in stakes races, at new distances or over new surfaces; the horse ran well but not well enough. The trainer's efforts worked but the placement was wrong today? Careful assessment can be productive as these runners frequently are good bets as overlays when returned to a more suitable spot next time.
-- a horse runs a strong race and gets a bad ride; just be sure it was the bad ride and not your evaluation of the horse or race.
-- a horse runs a good race and encounters bad racing luck. Poor starts, an errant horse, an accident, carried wide are all regular in-race occurrences. These become particularly important for later if you notice the rider wrapping up on the horse.
-- a highly competitive race with a big field often wastes good efforts for several that get forced into unavoidably tough trips because of the way the race came up. The eventual finish position may be meaningless, especially if the horse was not overused throughout.
-- a trainer runs an entry or soft-entry; only one can win. Is one prepping or having a training race today? What does this do to a trainer's statistics?
-- and my personal favorite, NOT: the jockey exits the inside path to go wide and loses a photo to the horse that followed through on the vacated inside path to beat him a nose.
In all of these instances, the trainer may have accomplished what he set out to do: get the horse to run a good race. Often enough, human error or external circumstances negate the finish results and thus the statistics! Only thorough subjective evaluation of the horse and the race will determine how well the horse actually ran; this may be the best and most productive way to evaluate your database or any published statistics. Not all positive outcomes are necessarily reflected in better win or in-the-money statistics. If the horse produced an improved performance then the trainer pattern worked. This reinforces the underlying factor and does not negate it. This subjective and analytical approach gives a better understanding of the trainer's command of his horse and his approach than statistics do.
What really matters is how the horse performs today race when the trainer employs what is normally a successful move for him. Not all horses respond to the same routines positively. And there are a variety of circumstances where a horse may perform well without winning or even finishing in the money; these will not show up in any of the statistics. Understanding that a pattern may produce and a horse may run well without winning might cause you to evaluate statistics in a different light.
Bad things do happen to well-meant horses. Sound, well-prepared horses from top barns with good riders tend to get into less trouble; but very often, today's race is not the primary objective; with honest and consistent evaluation of the performance of the horse, one will find that the statistics mean something different. Going through this process will also reveal a better understanding of how many trainers operate.
Here are four examples of trainer situations. Somewhere in the examples and analysis there should be one or two ideas that will enhance your appreciation of trainer information, improve your game and help you cash a ticket or two.
1) On January 11, Nick Zito ran two maiden claimers in the 1ST and 4TH races at Gulfstream. In the 1ST, EON (Fusaichi Pegasus) was first time for a price tag and went off at 1-5. He had the lead entering the stretch and folded without obvious excuse. BRIS stats rated Nick at 19 percent overall; 20 percent MSW (maiden special weight) to MCL (maiden claiming); 14 percent first MCL; and his ROI was negative in all categories.
In the 4TH, PICK OFF (Fusaichi Pegasus) was returning from a layoff and dropping from a $30k MCL at Churchill to the $15k level. He went off at 4-5 and won going away after stalking the pace. The stats listed in BRIS' Ultimate PPs had Zito at 19 percent off similar layoffs; 22 percent with shippers; and 22 percent with double droppers. In all cases, he showed negative ROIs. Both horses had the highest BRIS Speed ratings in their fields. Both were double class droppers, both were running related distances, both had multiple up-close calls in their prior races suggesting some degree of current form.
2) This is typical of many situations you encounter. The trainer information may or may not be directly relevant. A first-time starter is running nine furlongs on the turf. The horse has a decent enough pedigree for the surface and has a string of regularly-spaced works. The standard trainer stats say he is 22 percent with first-time starters, 24 percent first time at two turns and 21 percent first time on the turf. He is dollar positive in single digits in all three categories in his ROI.
3) A trainer is running a recent claim back after 60-plus days and is entered in an allowance route. He ran a so-so race in a sprint for a middle claiming price in the pre-layoff race out of which he was claimed. The horse has worked regularly with a series of six-furlong works and has finished up the preparation with two works in the last 10 days. Published stats rate the trainer at 16 percent from multiple starters; he has a hit rate off the claim of 22 percent -- a sign of positive intent; he is 8 percent after a 60 day layoff -- negative intent? His other stats are around 15 percent with class rises and stretch outs. Combinations of details often belie the raw data which very often proves crucial. The simple facts appear to conflict on the surface.
4) There is a large group of trainers with statistics akin to: 10 percent win record with first-time starters and a 23 percent second-start win percent. Some are dollar positive in one or both instances, but very often the lower win percent results in a higher ROI. The range of these stats varies cyclically. Whether it is a question of quality of current racing stock or trainer intent is unclear at times. Trainers like Carla Gaines, Cliff Sise Jr., Jimmy Jerkens, Steve Asmussen and Todd Pletcher all come to mind.
Both of the Zito runners were expensive purchases which have not fulfilled expectations; while each was a possible winner, Zito probably hoped that they would win and be claimed. Of course cheap maiden claimers are not the most reliable; he lost the 1st at 1-5 and won the 4th at 4-5, and neither was claimed. Neither fit the Zito mold of a developing stakes prospect.
There was no pre-race preparation of significance, which I use as a key sign of intent. Nick can burn a lot of the public's money, so with the stats in the BRIS Ultimate PPs in the vicinity of 20 percent and strongly negative ROIs across the board, and with no positive "karma" for support, they were best passed. There was a subtle difference between the first-time maiden claiming stat at 14 percent, below his overall win percentage (19percent), and below the 20 percent stat for MSW to MCL; I could not read a lot into that. There were more favorable numbers for PICK OFF; Nick is 22 percent with both shippers and double droppers -- above his regular win percentage. He had also run an improved race when tagged for the first time at Churchill in a race he was not favored, so perhaps he might be given a chance in what looked like a soft spot.
When I analyzed the results afterward, I stuck with my pre-race thoughts. Neither ran any better than anticipated, there were no surprises and there was no reason to revise my pre-race opinions. Nor did I learn anything that would cause me to put them on my horses to watch list.
The second and third examples illustrate the potential weakness of simple statistics and the value of information that is specific to today's task. Very often statistics are incomplete or conflict. If we know specifically how a trainer does with debut runners going long on the turf we have the more meaningful subset of the general statistics which may not apply. What does one do when statistics conflict as do the claim and layoff numbers detailed in the third case?
The answer lies in information which details a variety of factors which are specific to today; likely sources are a good memory, a notebook or a database. These rarely are found in public sources. Some serious players combine the details about certain winners. They normally keep records so they know what works and what does not and do not rely on win percentages.
Statistics/trainer data are a reflection of the trainer and his judgment. The PPs tell what the horse can or cannot do. If they dovetail, then you may have a strong play (e.g. a good record stretch-out trainer gets a young stoutly-bred sprinter).
Both examples show just how inconclusive some trainer statistics are. In the third example, a private database (that one takes the time to compile) might show that the trainer has recently won with this work pattern, has a number of recent misses without them, and also has five winners at 10-1 with a claimed runner, laid up for a similar period, and then worked six-furlongs at least three times before stretching him out, while stepping him up. The public might focus on the positive claim stat or the negative layoff stat -- or simply wind up confused. That database has a detailed trainer pattern which signals very strong positive intent today. Very few people would know these details. This gives him unique ammunition to take to the windows.
The statistics in example four indicate best that these are win-early trainers. The horse is prepared to run his race though may not be honed for a top effort; if the race/trip comes up right and the competition is not too salty, they will win. The first race may be a win or at least a positive learning experience. If the horse runs well without winning, then they receive additional preparation and usually run better second time.
Learn to watch for an improved work or work line; many horses "wake up" after their first experience. Be wary of horses exiting all-out efforts in their debuts if returned quickly; many fail to perform; risk takers might play against these or at least use multiple horses in the serial wagers.
Often there was a reason for that subpar effort; a layoff will indicate that. They usually return with an improved performance -- though not always a winning one. The third start, which is in closer proximity, is often the winning effort. Others that disappoint in their debuts get either a training remake or quickly drop.
To be really meaningful, early-runner statistics should detail the first-, second-, and third-time stats together. Because of published statistics there are trainers who like to fly under the radar and win a little more often with their third-time starters, a statistic seldom seen.
Generally published statistics tell only a small part of the story. The devil is so often in the details; what may first appear as conflicting statistics may contain a hidden story with the missing details added. Sometimes you can piece together multiple relevant statistics about some runners from the various entrants from the same stable on a given day. BRIS' Track Stats can help fill in the data for other categories that may be relevant today. BRIS also offers several other trainer reports if you wish to study a trainer in detail.
By spending time analyzing what happened to the horse and the quality of the race, you will learn just how well the horse ran and how the circumstances affected the actual outcome. This will then give you better insight and understanding of the way a trainer thinks, executes and his overall agenda. That is worth far more than any individual positive statistic or ROI which will be over bet by the public anyway.
Focusing on subjective post-race analysis will tell about how the horse runs in the context of the race. Just because the horse does not win does not mean that the trainer failed in what he was setting out to do. A good winning performance in a poor-quality race might mean less than a strong fifth place finish in strong race.
In the third situation, the horse could have won at big odds, the horse might have performed poorly, the horse ran really well and was simply beaten by a better horse exiting a stakes and taking an edge, or the horse ran well and was unlucky. In three outcomes, the trainer pattern performed to expectations. More careful qualifying procedures might be the lesson for you -- even with improvement a runner cannot win if overmatched. You might learn that this trainer or this particular owner is a bit of a dreamer. You may have a horse for your list and your trainer knowledge was vindicated, even if there was no cash at the end of the rainbow today.
Using the power of personal observation and analysis to evaluate the performance will give you a better understanding of the trainer, the horse, and a better appreciation of statistical data.
| More Features: |
Send this article to a friend
No comments:
Post a Comment