|
BGonline.org Forums
statistical significance
Posted By: Bob Koca In Response To: statistical significance (Timothy Chow)
Date: Tuesday, 26 June 2012, at 10:15 p.m.
"Looking retrospectively at results that have already been obtained and claiming "statistical significance" is always a scientifically suspect procedure. "
Yes. Relevant to this particualar case is that if you were to have all the Giants play let's say 20 7 point matches the player with the lowest PR is probably a player who outperformed his true level. A similar phenomen is that the top player in an ELO system is usually overrated.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.