4

I am building ruby on rails (ActiveRecord) horse racing application. One of the features that I would like to include in the application is the ability to identify meaningful "Angles". For example, given the following facts:

  1. Horse A has raced 50 times and has 30 Wins
  2. Jockey B was riding Horse A for 25 of Horse As 30 wins

Now, if Horse A is racing today and Jockey B is riding, I would want the system to

  • Recognize the high win percentage Horse A has when Jockey B is riding
  • Provide / Create some sort of alert of the "Angle".

Ideally there would be some way statistically or otherwise to define thresholds so as to avoid a lot of false positives.

Is this the sort of thing a recommendation or inference engine would be used for or is the something entirely different. Also, given what I am trying to achieve can you point me to any Gems, resources, etc that might be helpful.

Thanks for your help.

mlwida
  • 9,922
  • 2
  • 45
  • 74
Mutuelinvestor
  • 427
  • 2
  • 5
  • 12

1 Answers1

4

Many have tried this over the last 30 years. Horseracing is too difficult and too nuanced to be solved by number crunching. It requires deep knowledge and skill. over the last 300 years all the angles and combinations must surely have been identified - few of them work profitably and 96% lose.

"Searching for positive returns at the track: a multinomial logic model for handicapping horse races" - Bolton & Chapman

"Computer Based Horse Race Handicapping and Wagering Systems: A Report" - Bill Benter

"Still searching for positive returns at the track" - Chapman

"Alternative methods of predicting competitive events: an application in horserace betting markets." - Lessman, Sung, Johnson

robert
  • 96
  • 1
  • robert - thanks for the reply. With the exception of the Lessman paper, I have read all of the resources you have listed. My question was aimed more directly at techniques for data mining the data for the purpose of uncovering valuable nuggets of information. I may ultimately use some of those nuggets in a model to pick the horses, but I haven't gotten that far yet. Thanks again for the input. – Mutuelinvestor May 09 '12 at 00:21