I'm curious how much is known about the coding of L/R splits here for both position players and pitchers., especially in the cases where a player showed an extreme platoon split over one particular season. Possibilities I can think of are:
1) The extreme season platoon split is coded into player performance for that season here;
2) The extreme split is ignored, but a generic or average league wide platoon split is used;
3) The players career platoon splits are used for each season, and the single season extreme split is ignored;
4) The extreme season platoon split is considered but regressed towards either the league platoon splits or the players career splits.
If 1) is used, the extremes would seem to be exploitable by the more experienced owners here, but I haven't seen much evidence of that happening. But I could be wrong.
If 4) is used and there is some mathematical regression going on, I'd be curious as to the details of that technique, to the extent they are publicly known.
Thanks much for any insights!