I think for relief pitchers less than 10 IP is a better standard than 20 IP.

I'd suggest:

1) Pitchers with less than 10 IP be given a MAXIMUM general performance of INN equal to say a 4/4 pitcher no matter their actual performance with LOPS/ROPS no better than .880 either way. The sample size is just too small for resonable performance. If they have stats worse than that, then leave their stats as they are. A pitcher who only pitched 4 innings but had an ROPS of .100 can be a great guy to bring in with the bags jammed and 2 out in the 9th. That shouldn't be the case.

2) Batters with less than 200 plate appearances from either side should be limited to the number of ACTUAL AT BATS from EACH side and after that be given the general INN performance now afforded PLATOON players. In other words, if a player had 200 total at bats, 100 from each side, but from one side he had an OPS of 1.500 and from the other side batted .800, he only gets the 1.500 rating for those 100 bats...he can't be used for all 200 at bats from the more favorable side. THIS would get rid of some of the platooning abuses now being done...and it makes sense.

3) Starting pitchers with less than 20 IP should be treated as INN and their OPS should be no better than .880 from either side.

4) A player who played a defensive position in less than 10 games should not be given an actual rating for that position but instead a generic rating of no better than 2-2-2-2 (assuming the ratings are what the game uses to predict defense). If his actual stats are worse, then use his actual stats. If the game would normally have given him a better rating (say he is a 2b playing at 1b and was great Defense at 2b), then ignore his actual 1b rating and use the adjustment for playing out of position. This latter adjustment is because: "A 2B who never played a position should NEVER get a better rating than another 2b who had the same stats at 2B but ACTUALLY DID play at 1B."

4) One more suggestion though this is more a request...For RT leagues instead of having a TAB for the next year's stats which is always blank or an error message because they don't exist, why not have this year's stats plus last years stats?

Thanks for listening.