THE BOOK--Playing The Percentages In Baseball

Then there’s that enormous home ballpark of his, which saw 64 balls stay in the park that would have flown out of Shea Stadium

http://www.nytimes.com/imagepages/2009/10/02/sports/02mets.HR.html

And for everyone who likes to point out that the 1-year park factor for HR at Citi Field is 1.057, let ME point out that the 1-year park factor for HR at Citizens Bank Park was 1.005 for 2009.  So, while you’re snarking that for a tough HR park, Citi sure has a high HRPF, please attach your impassioned argument for how it’s easier to hit homers in Queens than it is in Philly - I’m dying to hear it.

That was a preemptive shhh… I’ve got a whole can of shhh here with your name on it…

I think this also illuminates a problem with linear weight’s based, I think.  Santana’s homeruns will count for 13 in the formula, just like everyone else; however, based on the fact that he walks so few hitters and strikes out so many (meaning he will allow fewer hits), the homeruns should be less detrimental than for the average player, right?

Nick, which is why BaseRuns is better for computing a component ERA or ERA-like number from a pitcher’s individual stats (actual or projections or what have you).  BaseRuns essentially adjusts the value of each component based on the magnitude of the other components.

Then again, it doesn’t make all that much difference.  A lwts-based ERA or ERC is going to be pretty close to a BaseRuns based one, I think, in almost all cases, especially when you are using projections (which tend to be more regressed towards a mean) for all of the component stats.

Ok, thanks MGL.  I think Colin Wyers developed a Base Runs model of FIP awhile ago, I wasn’t sure what that was at the time though; I’ll have to dig that up.