As some of you know, I’ve been creating a list of 6-max rules by running some statistical analysis on my database. I had to learn queries, and I thank all the guys in the SQL forum for helping me learn.

I have done the following:
1) I used Excession basic 10-max rules as a starting point. You can find his latest rules at http://www.andymcnish.btinternet.co.uk/newauto.htm
2) I adjusted my database for outliers because they were skewing the data. A few of the hyperaggressive types (say AF of over 20 or even infinity) were making the average AF much higher than the median.
3) I declared that anything greater than 1 standard deviation or in the upper 1/3 or bottom 1/3 of the population qualified as extreme. For example, a tight player must be more than 1SD from the average or in the bottom third of all VP$IP.
4) I used 50 hands as a minimum threshold to qualify for the autorate rules. The problem here is you need a minimum threshold that is large enough to be representative of that group’s style, yet not so large that a losing style has already busted out. By 100 hands, many poor styles (like Extra-loose) have already busted out making the data look different.
5) I then finetuned the database rules so that there is a relatively even distribution between the major categories (e.g., rock, fish, etc.). This is extremely time consuming.

I hope this helps everyone. I have this all in Excel and I’m happy to forward it to anyone who wants the full data. Thanks to Excession and Aperfect10 for all their help.