Data Mining Indeed. That's exactly what it is. Data mining. They are not selections. It is accurate information that can be used as part of the selections-making process.
There's nothing wrong with data mining. The error comes from blindly using the mined data without the input of a thoughtful, informed and knowledgeable handicapper.
Some of the best and most profitable handicapping information has come from data mining. Of course, a lot of garbage comes from data mining as well. The key, as I'm sure you are aware, is to separate the trends that simply describe the past from those that predict future performance as well.
After all, when mining for real gold, most of time you just get sand and dirt. But when you do strike gold, it can be very profitable.
Thanks for the thoughtful question. I actually posted the trends to get input from experts on whether they were valuable or not. I enjoy the challenge of using the SDQL to find interesting trends in past performance.
Bases
