I wish I knew someone who trained thoroughbreds. If I knew someone who trained thoroughbreds, I could probably increase his win percentage by 5 percent easily, (with all due respect to female trainers, I will use “his” for ease of use). Now, you might ask yourself, how on earth could a racing blogger help a trainer increase the win percentage? The answer lies in where horse racing is and where it needs to be heading.
If racing were compared to traditional sports, trainers would be the head coaches. Well, I’ve noticed one peculiarity in how horse racing works differently than traditional sports and it’s not a good one; horse racing conditioning is completely void of analytics!
In every other sport, the coaching is done using equal parts analytics and gut instinct-as it should be. Bill Belichick is an ardent follower of this philosophy. He studies statistical analysis of playcalling as it relates to downs and distance then seamlessly melds that knowledge with his own intuition, and the results have been the creation of the most over-acheiving team in the history of the National Football League.
So, how can trainers learn from Belichick?
Every horse is an individual, but after 150 years or so, certain patterns have emerged. There are ideal times to bring your horse back after a layoff. There are training patterns that can be used to get your horse to peak at the right time. There is even geo-spatial (location based) data that can tell you which tracks are most likely to cause injury to your horse and which tracks will cause your horse to be stronger if he trains over them.
Imagine the type of advantage a trainer would have if he used both analytics as well as gut instinct. All of a sudden, all his horses would be running in the right spots. Fewer of his horses would be injured. He could even measure his horses feed to figure out their nutritional state.
Now, I know for a fact that trainers have a good eye and can figure out all the things I measured using a pencil and a notebook, and still get it right 70% of the time. If they incorporated analytics, they would probably get it right 95% of the time; which last I checked, is even better than 70%. Going back to football, Norv Turner- a more traditional coach called a pretty good game. I would say his calls made sense 70% of the time, but Belichick almost always called the right play, and that is why the Patriots have been more successful than the Chargers even though they have the same level of talent. You simply need every advantage you can get your hands on, and analytics seems to be one of the last remaining advantages.
Like football, racing is a game of inches. If you could use analytics to make your horse one length faster, imagine how many third and fourth place finishes would turn into wins or close second place finishes. The advantages are even greater considering that most old-school trainers hate data. They seem to think that numbers are a plague upon humanity and analytics is its evil spawn.
With many trainers in charge of 30-40 horses, mistakes in training, feeding and race selecting are bound to be made. The trainer who uses software to help him keep track of feeding schedules, and analyze training practices as well as the condition book is bound to make less mistakes than his counterparts.
I am proud to be part of an organization that understands how data-driven handicapping should be. I have used Thoroughbred Analytics to give me that slight edge when handicapping, and I am a more consistent and profitable handicapper because of it.
Do I simply pick a horse based on its TA indicator? I do no such thing; but I do combine my years of racetrack experience with the TA indicator to give me an edge over the competition. In other words, I pull a “Belichick” and combine intuition with analytics to compete at a higher level than I ever thought possible. With so many handicappers picking horses based on nothing more than hunches, the handicapper that uses data is at a huge advantage, and the handicapper that combines data and intuition is incredibly tough to beat.
Maybe a middle of the pack trainer will use this article as motivation to give his intuition a solid data-driven foundation and win the training title at his track. Maybe, a young handicapper will use data to win a handicapping tournament against a bunch of grizzled veterans, one can only hope.
Beware of the Bounce
For some of the more astute handicappers, this article will not teach you anything that you already know; however, if you are an intermediate level handicapper or below then you will need to pay close attention because what I have to say will benefit you greatly.
Strategic Manuever in the Frizette, Bellamy Road in the Kentucky Derby, Big Brown in the Belmont are all famous examples of horses who bounced. The problem for handicappers is that horses bounce all the time at tracks all over the country and other than the most astute handicappers; no one picks up on this.
There are basically five rules that are to be taken into consideration when trying to figure out whether or not a horse will bounce.
The metrics and calculation criteria do not have to be deployed in order, but they absolutely must be deployed to the last amount rather than the whole amount. Failure to do so will result in a calculation error.
Has a horse made two significant jumps in its speed rating?
Though the reasons are unclear, horses can make two significant jumps in speed rating, but rarely ever a third. Yes, Secretariat did it, but he was a freak of nature and the average horse at Canterbury Park or Turf Paradise cannot imitate him. If a horse has made two significant jumps in his speed rating, I usually take his TA indicator and subtract 10%. For example, if a two jump horse has a TA indicator of 100.00 his adjusted TA indicator is 90.00
Is the horse overraced?
If the horse has run in the last twenty days, it is over raced, and 5% needs to be subtracted from its TA indicator. So, using the horse in our previous example, he started out at 100.00 then went to 90.00 and is now at 85.50.
Is this a sprint race?
If this is a sprint race and the horse is a two jump horse, then you must subtract another 5% from his TA indicator. So, using this same fictional horse, he is now at 81.225.
Is this horse a pure frontrunner?
If the horse is a pure frontrunner and he has made two speed jumps, I can assure you he must be pretty tired. Go ahead and subtract 7% from his TA indicator. So, our fictional horse is now down to a TA indicator of 75.54, and is about to fall once more.
Is the horse six years or older?
If the horse is Six years or older and he has made two jumps in speed, subtract 10% from his TA indicator. So, our poor fictional horse, who started out with a TA indicator of 100.00, is now at 67.99.
The TA indicator has shaved so much time off of my handicapping. I can look at a simple number and get an incredibly accurate picture of how good a horse really is. If I looked at nothing but this number and played cold exactas I would still make money, but putting context to the TA indicator really takes a handicapper to the next level.
Applying the “bounce metrics” that I have created allows me to account for the anomalies in racing that one has to be aware of in order to succeed. If I can take the time out to apply the “bounce metric” the TA indicator will do all of the heavy lifting and pick up on the things that I cannot accurately measure such as lengths gained and horses passed. I cannot tell you how important the lengths gained metric is, especially when evaluating turf horses stretching out, but all metrics become more powerful when you apply some context to them.
So, the next time you see a horse who has jumped twice in its speed rating, apply the “bounce metrics” to get a more accurate picture of the horse’s TA indicator and you might catch a 30-1 longshot and avoid being part of the “dumb money” that makes odds on favorites out of horses who wind up as also-rans.