State of the art artificial intelligence algorithms have been developed that allow computers to use machine learning to decipher and analyze the factors such as prior races, ancestry, pedigree, and speed figures to come up an astounding resolution to an ultra complex problem.

By Litt Gunn

The Sport of Kings——Center Stage.

Suddenly it’s— that time again. The time to consult the past performances, apply the “mental gymnasium” and pick a winner. Why most people will be happy to just land in the money. The steeds on the track: yes they halt, check and cut short; flail, flounder and fall short; are transitory and short lived; many are fleeting in temperament and short lived; some are handled by the barn that is short-staffed; and others are hot-tempered, irascible, fiery, testy, touchy and short-tempered. There are bad actors and breakdowns. But then there are ones who have the “Look of Eagles.” Those horses gaze into the distance with an air of knowingness. Find those horses, apply the handicapping theories and patterns and discover the winner.

So what’s the problem?

Well here’s the problem: It is a very, very long distance. It’s and mile and a quarter. Much longer than the standard distances run by horses in North America. No Horse in the entire field has ever competed at this distance. This is a dirt race. For horses suited for dirt this is about as long as it gets.

Horses just do not handle’s this distance comfortably. That means that all the horses have to push themselves and find the reserves to finish this race effectively. As handicapper I have to know which are the ones which will succeed.

In America it’s speed not stamina which gets the breeder’s attention. The tilt is to the horse who can shine at a mile or less. We have created abnormality. At the top end they are all an anomaly. They are running on eggshells. Even though some horses could be somewhat suitable to handling distance but there is no absolute certainty a three year in North America can be brilliant or even lively at this distance.

Even if the horse can handle the distance none of the horses (the 20 in the starting gate) have ever done it before. There is a big difference between being able to do it and actually doing it. If a horse has never hit the wall before and fought through it then they don’t know what it feels like, and they are likely not to react to it positively.

Horses that get tired do unpredictable things. What will the effect of a tired horse or horses have on the race. It is harder for a tired horse to run in a straight line down the stretch. They can slow down so much it look like they are moving backwards compared to the rest of the field. Which of course means that they get in the way of other horses trying to make a move. So if the horse you want to win is a closer who is looking to move through the field late this is a problem. The ones in front and mid-pack inevitably create disruption at a distance.

So what does a handicapper have to throw at the problem?

Beyer figures, Tomlinson ratings, post positions, track bias, prep race schedule, workouts, and jockey-trainer combinations, positive and negative jockey or trainer switches, quality of the past trips.

Compare numbers, be they raw times, speed figures, finish positions, or beaten lengths. Calculate “speed points” to determine the front-runners.

“Pattern recognition, the most important skill in using the Beyer Speed Figures effectively.

Apply bounce theory—determine those due to regress out of contention and those who have already regressed and are about to cycle up to a figure closer to his career best. (no fillies in this race this year)

Determine the effect of the inefficient betting pools. Derby day as you know attracts a huge crowd at the track and large betting pools so the potential for an inefficient market comes into play. (Many people betting into the pool are not handicappers and instead are betting on such non-factors as colors, lucky numbers or birthdays, or the names of the horses or jockeys alone.)

Use the trainer angles, namely: claiming (first off a claim and class drops), layoffs, surface changes to and from the main track, equipment changes, 2 medication changes, juvenile starters, and go-to jockeys. Positive jockey-trainer combinations from across North America with winning percentage and ROI.

Then there is the Dosage Index and Dual Qualifiers.

Dosage figures give the dope on a contender’s distance potential based on stallions from the first 4 generations of that horse’s pedigree. It’s the geneticallybased theory that takes into account not only how well certain ancestors of the horse did, but also how far back in the horse’s line they are. Points are awarded for speed and stamina of top-notch horses that often appear in contenders’ pedigrees. (these superstallions are called “chefs de race.”) Any horse with a dosage of 4.00 or less is likely to be able to make the derby distance.

Dual qualifiers are those horses that not only have a dosage of 4.00 or less but also are within 10 pounds of the high weight in the Experimental Free Handicap rating. It is based only on races from the contenders’ 2-year-old season. (The Experimental Free Handicap has been published annually by The Jockey Club since 1935, is the weight-based assessment of the previous year’s leading 2-year-olds, with the weights compiled for a hypothetical race at 1 1/16 miles on dirt.)

So Now What?

In light of all of the above I, and any other rational human being, realize it is beyond possible for a handicapper to mesh it all together. But you are in luck. The new world of Artificial Intelligence is at hand and with great skill and determination I have develop an algorithm to solve the problem.

Swarm Intelligence

As you might be aware, last year the cutting-edge artificial intelligence startup, Unanimous A.I., created an algorithm based on Swarm Intelligence. They defied the odds by using Swarm Intelligence to predict the superfecta outcome— which is the top four horses in exact order—of last year’s Kentucky Derby. (The top four choices in the betting finished 1,2,3,4 in last year.)

It is altogether possible to tap into the intelligence of groups. Such method has worked in a large number of fields from forecasting movie box-office to predicting the price of bitcoin. (The inspiration often comes from nature, especially biological systems. Natural systems of Swarm Intelligence include ant colonies, bird flocking, animal herding, and bacterial growth.) Swarm Intelligence follows simple rules, and there is no centralized control structure dictating behavior. The random interactions lead to the emergence of “intelligent” understanding.

Swarm Technology has been used in financial trading with mixed success. This type of intelligence relies on individuals that are relative homogenous. At the Kentucky Derby the diversity is huge with most not necessarily thinking together on the complexities. How many will be required to think together to solve horse racing riddles? Thousands? Million? We don’t know yet. We know the goal— pooling of each individual’s knowledge and insights into a single elevated intelligence. We are probably a good distance away. Even though the favorites have been outsized in their success in the last few years it is more than unlikely that picking the winner by crowd thought, even including pros in the handicapping business, will pick out the winners.

Fabulous New Algorithm

For your benefit, I have applied a skillful application of numerous factors, in the development of a unique algorithm that the computer then gunn-mcuanalyzes and processes through self-learning. This state of the art artificial intelligence algorithm that has been developed allows the computer to use machine learning to decipher and analyze the factors such as prior races, ancestry, pedigree, and speed figures to come up an astounding resolution to this ultra complex problem.

Once the computer is fed the Big Data from thousands of previous three year old races an understanding is reached and a judgment is made as to the winner of the Kentucky Derby. How could Jane ever have believed that saving old racing forms would not one day be a treasure trove?

Contenders History Telling a Tale

With machine learning, the process of selection is not revealed. How the decision is made is forever a mystery. The machine meshes the torrid stew, analyses the factors, and predicts the future.

Machine learning operates like a neural network with the ability to not only assimilate and build upon incoming data, but to create a self-learning program —to adapt to the massive amount of information fed in. Its predictive learning.

To your great advantage the algorithm has been developed and the big data applied.

The Mix in the Secret Sauce

Data applied include key facts on:

Results of the Major Preps: Spiral Stakes, Sunland Derby, Florida Derby, Louisiana Derby, San Anita Derby, Blue Grass Stakes, Wood Memorial and the Arkansas Derby.

And Contenders:

Gormley, the Santa Anita Derby winner has superior numbers as far as speed figures, wins, starts, and prep race qualifications. However he finished 7th in the Breeders’ Cup Juvenile. Front Runner win qualified him as a “Dual Qualifier.”

Classic Empire: 2-year-old champ. Defeat in the Florida Derby. Foot abscess. Third in Holy Bull. First in the Arkansas Derby.

Always Dreaming: Trainer—Todd Pletcher. 5 length score in the Florida Derby. Maiden win by 11 lengths at Tampa Bay Down. Son of Bodemeister.

Irish War Cry: A no show in the Fountain of Youth. Won the Holy Bull and the Wood at Aqueduct—with impressive Beyers 102 each.

McCraken: Won first four starts at two. Swelling in an ankle. Third in the Blue Grass Stakes at Keenland. No 100+ figures this year.

Gunnevera: Venezuela connections. Purchased for $16k and has made more than $1.1 million. Deep closing style. Won the Fountain of Youth by 5 5 lengths at Gulfstream. Florida Derby broke from far outside—third. Delta Jackpot win earned him “Dual Qualifier” status.

Practical Joke: Hasn’t raced since October. Multiple Grade 1 winner at two. Runner up in the Fountain of Youth and in the Blue Grass. Has speed to be mid-pack. Never been passed in the stretch. Has not run a 100+ figure this year.

Battle of Midway: Early speed. On the pace, caught but hung on doggedly to be second in the Santa Anita Derby. Lacked starts as a juvenile.

Irap: It took 8 races to break maiden. Won the Blue Grass Stake at 31-1. Not a “Dual Qualifier.”

Thunder Snow: Dubai horse. A win would be a first for Dubai. Won the UAE Derby with a speed figure of 101. Six starts as a juvenile with a win in the Criterium-G1 where he earned an estimated speed figure of 105.

Hence: Sunland Derby final prep like former Kentucky Derby winner Mind that Bird. Speed figure of 103 in the Sunland. No stakes as a Juvenile.

Final Reminder and Selection

No matter how good the system; the analysis; the horse or the handicapper it is always advantageous to remember advise some would regard as sage.

  • “A racetrack is a place where windows clean people.” (Danny Thomas)
  • “Horse sense is what keeps horses from betting on people.” (Raymond Nash)

Predicted 2017 Kentucky Derby Winner: Irish War Cry

At the Finish Line

In the rain, however, it was Always Dreaming #5 that won the 143rd Run at Churchill Downs in Louisville, Kentucky, for trainer Todd Pletcher and jockey John Velasquez.




Leave a comment