Posted by zubinsum on 2/14/2011 1:31:00 AM (view original):
I did a statistical study on this sometime ago by looking at low ip starters (presumably used in relief) and comparing them to actual relievers (or guys classified as such by the database). I found the low IP starters on average performed worse than the relivers when using # stats as a baseline. I am going to reinvestigate and post results.
Interesting, I am not finding the same pattern I saw before. In fact, low IP starters faired better:
Starters:
SIMoav - oav# = .043
SIMoav / oav# = 1.213
Relievers:
SIMoav - oav# = .047
SIMoav / oav# = 1.254
I am reversing my position... I do not believe there is a performance penalty for starters used in relief.