Posted by tlowster on 11/15/2020 6:36:00 PM (view original):
I think you can do it throughout different Worlds. The intent is to find players with a lot of at bats so you can weed out small sample size, then reverse engineer the players ratings to their stats. Since all players in every World are playing under the same simulation, I think your base formula can be from players in different Worlds.
However, I wouldn't make it that hard on yourself. If you participate in a World that as been around at least 35 seasons, I would start with players in the Hall of Fame and place them in your sample first. This will give you a good barometer of the players that are elite. Then maybe review some guys that have mediocre to above-above ratings like guys in the last three seasons of the all star player awards page (try to find an older player). Finally, stay on the same Player Awards page and take a peak at the gold glove winners for 2b, CF and SS from the last three seasons and try to find an older player. Typically guys that win a gold glove at those positions are not very good hitters unless they are a truly elite player. This should get all three types of players in your sample and you can use trial and error to determine your own formula. I have mine and it works pretty well, but I still struggle a bit with predicting slugging. It is difficult to isolate variability over a season with the variability of speed and power for a hitter.
That makes sense...however, rather than starting from scratch I have a formula that I use and I want to test out the current level of accuracy. Would you mind if I sent you a sitemail to walk you through how I'm considering testing it?