One of the things I truly love about horseracing is that all the bull one kind of learns in school about statistics, risk and prediction actually has a practical theater where it applies.
I particularly like looking at statistics books, not that I understand a tenth of what's in them, but because they challenge me to think about The Game differently. [By the way, you would be amazed at how few trials
scientists perform when they study certain clinical topics--no different than any one of us picking a pony based on some knuckleheaded insight like "the last 5 Pletcher move up horses have won with 15 days or more rest.". Scientists--to hear them tell it anyway--look at hundreds of variables, while we stumble through with 8 or 10, but the truly craftiest of them appear to me to be handicapping their work about the way we do.]
Anyway, here's something from a book preface that helped me; it's supposedly one of the harder questions in analysing data and making predictions:
Quote:
"What predictors are most likely to help predict in this prediction problem?" [That's the way they talk sometimes.]
I like to think about statements like this as if they were
koans--mull over them, see how they would have impact on a handicapping analysis.
I hope it will open some handicapping doors for you and not be just more bull. This has helped me immensely in this Belmont meet where I've always felt a bit lost and at sea after Aqueduct with its post position trends. I think I've found the (well,
a) Belmont key now--it helped me find
Hint today and two or three others.
Good luck, all.