CERA does, indeed, have it's limitations. Can you list for us what those limitations are ? Can you comprehend why they aren't applicable in this case ? Quite simply, there are far too many variables to neutralize or even out over such a short sample size. The opposing team's health and performance. Where the games are played. What the weather is like when playing. The health of your own team. And when you have your starter missing a month, plus other injuries forcing player changes, it becomes even harder to account for the changes... especially with such a ridiculously simple formula. Even if you want to give Koyie Hill 100% of the credit for the better pitching performances (and there's no reason to give him any at all, as every single attempt to quantify or prove that certain catchers can call a better game has come up inconclusive at best or proven outright wrong at worst), Alfonso Soriano hit .337/.412/.558 during the month where Soto was out and Hill got all the starts. Now does it really seem fair to you to assign the brunt of the positive impact on our team's offensive output, run differential, and record to Koyie Hill's .194/.262/.258 line in that time?