Using a formula to predict a players offensive performance in a single game prior to game start. While this will obviously have huge margin for error, the idea is if you used this formula a 100 times it might balance out.
(Let's assume wOBA, for the sake of argument, is our best offensive metric).
All stats are based on ZIPS(RoS) projections:
wOBA * ([Park Factor] / 100) * ([ERA of Oppositiong Pitcher] / 4) * ([Avg ERA of All Opposing Relievers] / 4)
I'm using ERA, since we're not evaluating the pitcher solely we're evaluating the pitcher performance as part of the team I think it makes more sense.
Assumptions:
-League Average ERA is 4
-Park Factor is from Previous Year, to get full data
What do you think?