skyking162

While almost everybody will admit that Coors is still a great place to hit, many people refuse to discount Matt Holliday’s performance this season even though he plays half his games there. The protests includes statements like these:

1. I don’t believe in “what-if” scenarios. I judge players on what actually happened.
2. Holliday also has to play road games in the extreme pitchers parks of the NL West: PETCO Park, AT&T Park, and Dodger Stadium.
3. Why don’t you penalize the Phillies’ hitters, too? Citizens’ Bank Park is also a hitters’ haven. And Houston. And Cincinnati.

Some of those are better than others. Let’s start from the bottom and work up, because I need to set up a few things before making my main point.

#3 - We absolutely should penalize other hitters who play in hitters’ parks (and reward hitters who play in pitchers’ parks). However, it’s important to have an accurate concept of which parks actually are hitters’ parks, and to what degree. Here’s a list of NL park factors averaged over the past three seasons. 100 is league average and a park factor of 110 means that 10% more runs are scored at that park. For AL park factors, click here.

Park	Factor
COL	120
CIN	113
ARI	111
PHI	109
CHC	109
PIT	100
MIL	99
LAD	99
SFG	99
ATL	98
STL	97
HOU	97
FLA	95
NYM	92
WAS	90
SDP	81

Not surprisingly, Colorado is the best park for hitters. Philadelphia helps hitters, too, but only half as much as Colorado. Notice that Houston’s park factor is only 97 — Minute Maid Park increases homeruns, but actually reduces overall run scoring.

#2 - Are the other NL West parks actually hell for hitters? I see park factors of 111 (ARI), 99 (LAD), 99 (SFG), and 81 (SDP), for an average of 95. Los Angeles and San Francisco aren’t really pitchers’ parks these days, and Arizona is a pretty big hitters’ park. Of course, San Diego is as extreme as Colorado, but benefits pitchers.

We can create an overall adjustment factor for Rockies’ players that weights each park factor by how many games they played in each stadium. With the unbalanced schedule, Colorado plays 81 games at home, 9 or 10 games away at each NL West opponent, 3 or 4 games away against the rest of the NL teams, and 9 total away games against AL opponents. I used the Rockies’ schedule from Baseball-Reference.com and calculated the exact weighted Colorado adjustment for the 2007 season. It’s 109.6. That means Colorado’s games included a bump in scoring of 9.6% because of the parks they played in. (Rockies games actually had only 4.5% more scoring than average, meaning Colorado and their opponents had some combination of below-average offense and above-average pitching.)

#1 - Some people, actually I’d say most people, think that park factors are only used in what-if scenarios. For example, what if Wade Boggs didn’t get to hit doubles off the Green Monster or what if Babe Ruth didn’t have the short porch in right field? But that’s not the only thing park factors are used for.

In fact, for an MVP discussion, it doesn’t matter at all what a player “would have done” in a different environment. We want to know what Matt Holliday did in his actual environment. I think we can all agree on that.

What was Matt Holliday’s environment? One where runs were scored 9.6% more often than average. What does that mean? It means runs are less valuable. There are 162 games in a season (well, 163 sometimes). If more runs are scored, that means it take more runs, on average, to win each game. Contributing 100 runs in the Rockies environment helps win fewer games than contributing 100 runs in the Padres environment.

By neutralizing Holliday’s batting average from .340 to .318, we’re not saying that’s what he would have batted if he played for a different team. We’re just saying that batting .340 on the Rockies is as productive as batting .318 for another team. Creating 151 runs on the Rockies is as productive as creating 135 runs for another team.

I hope my point is clear. The “penalty” given to Matt Holliday and other Rockies hitters is not a “what-if” adjustment. It’s an adjustment necessary due to the run-scoring environment created by playing home games at Coors and road games at other divisional parks, other NL parks, and a few AL parks. Matt Holliday is a great hitter, but his raw stats appear more valuable than they actually are.

Popularity: 5% [?]