Claude, build me a horse racing betting system

Sealegs wrote:

Fascinating.

Perhaps choose a particular racing centre & look for patterns within the training ranks. Pair this with your current model and there's a higher likelihood of success, given that you've now integrated the human intent factor into your system prediction!!!

 

But where to get that data? Someone told me in the early days: "You won't get an advantage unless you know something that no-one else knows. It's unlikely that you do."
 

Ii believe that formgrids provides a wealth of information. Trainers are often creatures of habit and when looking at a large enough sample you might find certain patterns that sets them apart. I don't want to get into the details!!!

The factors that affect the outcome of any race are many and not all are knowable, yet there is a sufficient amount that is knowable and could give you an edge, it's only a matter of interpretation.These various factors and the percentages that they each generally contribute towards a final outcome is something that will need to be arrived at or discovered through trial and error.

You aim is to combine as many knowable factors, through large enough historic sample sizes  to produce a prediction...which even then is little more than an educated guess