Quote:
|
Originally Posted by Capper Al
Delta,
I guess whatever the capper's goal is, either to make money or pick winners, it really doesn't matter. Whatever formula or method they use to achieve their goal, have they found themselves modifying their ways because of changes in racing?
This does lead to an interesting question for those who purchased software. How does your software company keep up with these changes in racing when they happen if the company does keep up them at all?
|
I am not well aware about commercial handicapping software, as I am excursively using my own. I still believe that this kind of software should be “factor agnostic”, meaning it should not hard-code them but should only rely on raw data as input and been able to mutate its agents in an on going basis.
Of course this sounds easier that it really is, as the required design and implementation challenges are of significant difficulty and complexity. An easier solution might be a “hybrid” approach, where part of the platform relies on hand crafted scenarios that the user can express in a DSL periodically, as a means of assisting the algorithm to discover new patterns, utilizing the handicapper's intuition and experience to some extend.
