THE BOOK--Playing The Percentages In Baseball

Do you think a pitcher like Peavy, who has both strikeout and groundball stuff, might have the ability to play to his park and suppress the park value a bit?

From the article on ESPN:

“That’s enough to give his ERA a bit of a bump but not enough to prevent him from remaining an elite pitcher.”

So how does a park effect have ANY bearing on whether a pitcher is an elite pitcher or not (other than if a park impacts a player differently than the average player, which is not the subject of the article)?  It doesn’t!  And that is a pet peeve of mine when people equate park effects with a pitcher’s value.  Repeat this 10 times:  A park effect has no bearing on a player’s value, a park effect has no....

A park effect, as measured by a factor, is a method of matching performance to talent. Peavy will likely perform relatively worse pitching for the White Sox in the AL in their homer happy home park than he did for the Padres of the NL in Petco Park. How much so?

Even if the league and park modifies his performance, does that have any effect on his value? I’m agnostic at this moment, so I hope I’ve phrased the question in a way that we can learn something here.

But that is why I did park neutral projections, to value everyone on an equal playing field. I view park-specific projections as a distortion, although they can be valuable to predict outcomes in specific contexts.

Performance is a relative thing as well, so I am going to dispute the notion that park or league effects have any influence on performance, but that is semantics and is not important.

What I am railing against is when commentators suggest that when a player changes parks that his value changes, which is simply and obviously not true, yet that is spoken of all the time, including in this article.

Certainly, whether Peavy plays in Petco, The Cell, or Coors Field in pre-humidor days, has virtually no bearing on whether he is an “elite pitcher” or not.  I take great issue with that sentence being in this article. A player’s talent and value has almost nothing to do with the context in which he plays, other than how that context might disparately affect one player compared to another player or the average player, which is something we know little about anyway.  The only time a player’s context non-neutral performance should be even discussed is in a fantasy context or salary negotiations (where park and other contexts are not fully understood or considered by players, agents, GM’s, fans, and arbitrators)…

2/3/4

I just saw a clip of Buster Olney saying that the Twins had the best lineup in the league.  This is pretty obviously not true, but what was even worse, is that he used Target Field as part of the evidence.

#1 - We don’t know that Target field will be a hitter’s park, and I’ve heard it projects to be more of a pitcher’s park.  Either way, dimensions aren’t everything, and we’ll have to see how it plays before we determine anything.

#2 - Unless Target field is especially friendly to left-handed power hitters, this doesn’t matter!  The lineup didn’t get better by moving into a new park! 

Ridiculous.

http://espn.go.com/video/clip?id=5040759

MGL,
while park effect mostly won’t change player’s value, in some cases they will a lot. Take guys someone like Castilla, who posted much better relative numbers in Colorado than everywhere else. That stadium suited him better than other players, so ho got more boost and gave team more value than expected.
Just quick example: tahe two good pitchers, one flyballer, with slight HR problem (more HR/fly than average), the other extreme groundballer, great at preventing HRs. If you put flyballer in old Coors, and groundballer in Petco, flyballer will post very bad numbers, much worse than his career numbers adjusted for Coors park factor, and groundballer will have very small improvement, much smaller than average pitcher coming to Petco. On the other hand, if you put flyballer in Petco, and groundballer in Coors, there is a chance that they would post ace-like adjusted numbers. Their value to the team depends greatly on their stadium and the effect it has on their tendencies. So, for players who are disproportionately helped/hurt by their park, park can really change their value.

Davor, yes, I believe I added that qualifier and caveat twice in my two posts in this thread.  And clearly that was not the point of the article referred to.

The author said,“That’s enough to give his ERA a bit of a bump but not enough to prevent him from remaining an elite pitcher.”

It naturally follows from that statement that if Peavy were moving into a REALLY batter friendly ballpark that the author would have to conclude, “That’s enough to give his ERA a giant bump, which is enough to prevent him from remaining an elite pitcher.”

Which is nonsense of course.  And Carson should know better.

Olney’s comments are a typical example of what I was talking about.  “So-and-so team will be moving to a hitter friendly ballpark and therefore they will have one of the best offenses in the league.”

Wouldn’t a guy like Olney, who is reasonably intelligent, realize that everyone else who plays against the Twins in that (supposed) hitter friendly ballpark will have the same boost?

BTW, it is doubtful that an outdoor stadium in Minneapolis will be a hitters park, unless it is really small, because of the average temperature.  I believe it will be one of the coldest parks in baseball.

As Carson shows us, Peavy has allowed a slightly lower than league average HR/FB on the road and well below at home.  So Carson appears to be assuming that Peavy has no discernible HR/FB skill based off of his projections.

This reminds me a lot of the Matt Holliday situation.  Many people projected his career road numbers as his true talent level given the huge park effects of Coors - and they were wrong, and Carson is wrong to do the same for Peavy for a couple of reasons. 

1) Both of their numbers at home were better than their road numbers + park effects, meaning that they likely did play much better at home (which we would expect due to home field advantage).  So chalking up Peavy’s home .069 HR per FB solely to park effects is a mistake I think. 

2) Their road numbers are artificially deflated because they don’t play any road games at their extreme home park.  Given Peavy’s distribution of road games, which don’t include Petco, I expect that a league average pitcher would have a higher than average HR/FB ratio.  How much higher, I do not know and maybe someone with Retrosheet skills can help me out!