I believe the Metrodome only serves 3.2 beer during Twins games, and it's nice to see Bill Veeck's anti-gravity machine still works. Yes, I'm emailing Tom Skilling about home runs in Chicago.

How’s that for a surprise? The Cell makes more outfield flies into home runs than Coors Field.

How is this surprising at all? It's common knowledge that US Cellular is an extreme home run park. Coors Field is designed to reduce the number of altitude-aided home runs, hence the huge outfield which plays a large part in the high run environment there.

Now what about walks? I’m fairly certain that sabermetric conventional wisdom is that the effect of parks on walk totals is minimal, if not non-existent. I found, however, that this was not really the case.

Anybody that has ever spent 2 minutes looking at detailed park factors like the ones in the Bill James Handbook already knew that.

What was really interesting for me was the 'walk factor' for Safeco. We've heard for several years that Safeco boosts strikeouts, but I'd never heard anything about walks. This is even more amusing taken in the context of the M's team construction. A team that neither walks nor stikes out much and a pitching staff that, until the addition of Bedard, wasn't striking out a lot of people either.

I didn't realize Camden Yards supressed strikeouts. With the move to Safeco and a full season, Bedard could strikeout 300.

Maybe I'm just not that smart, and I definitely didn't RTFA, but I don't really understand the K and BB park factor thing. Can anyone help me out?

A player is 28 percent more likely to hit an infield fly in Milwaukee than he is in San Francisco. Why is that? My guess is that it has to do with foul territory. Since infield flies are only recorded when the ball is put into play, parks with a lot of foul territory are more likely to see foul pop-ups stay in and get caught, whereas ballparks with little foul territory will see a lot of pop-ups land in the stands and go unrecorded.

If this were true, then why isn't Oak in the top 5?

Has anybody ever studied how much a home team player's splits affect park factors? One player can easily have 10 percent of a team's home PA.

When Thome was in Cle he had huge numbers at home compared to his road numbers. Jacobs Field was then a hitter's park.

Hafner has huge road numbers compared to his home numbers. After Thome left and Hafner became the big hitter in the lineup, Jacobs Field became a pitcher's park.

Has anybody ever studied how much a home team player's splits affect park factors? One player can easily have 10 percent of a team's home PA.

I suppose you could perform this same exercise but only use the visiting teams stats?? Over the course of 3-5 years you'd have less of a sample size problem. Perhaps they already account for it? You could email Gassko and ask him.

Maybe I'm just not that smart, and I definitely didn't RTFA, but I don't really understand the K and BB park factor thing. Can anyone help me out?

Uhmm...I'm not smart enough to explain it. There's some explanation in the article, but I'm guessing it's the same as figuring out the HR park factor. Can you dig that one?

Uhmm...I'm not smart enough to explain it. There's some explanation in the article, but I'm guessing it's the same as figuring out the HR park factor. Can you dig that one?

Conceptually I'm all over it. I understand that in this or that park, more fly balls turn into homers, or more batters strike out. In the case of Ks and BBs, what I don't get is why. Intuitively it just doesn't make much sense to me that a park could suppress walks, for example.

Conceptually I'm all over it. I understand that in this or that park, more fly balls turn into homers, or more batters strike out. In the case of Ks and BBs, what I don't get is why. Intuitively it just doesn't make much sense to me that a park could suppress walks, for example.

The hitting background can affect how easy/hard it is for batters to pick up the ball out of the pitchers hand and recognize the pitch.

I sent an email to Gassko asking him a) his theory on how a ballpark can suppress or boost walks and b) if they factor in (the example I used) Ichiro! inflating his team's (and therefore his park's) groundball numbers when they do park factors. So, like ronh wrote one player doesn't knock the whole system out of whack.

b) if they factor in (the example I used) Ichiro! inflating his team's (and therefore his park's) groundball numbers when they do park factors.

Are his GB rates different at home vs on the road? If not, then they wouldn't affect the park numbers.

I would guess it has something to do with the batter's eye, shadows, and sightlines.

I don't know if the datasets are large enough, but separate day/night factors might shed some light on whether the shadows have an effect.

I don't know if Ichiro's rates were different but it's not really the point. A lot of guys have very different rates for [insert your favorite stat here] home and away. ronh, I didn't want to steal your Thome/Hafner example, but it's a better one.

On walks:

The spread here isn’t quite as great as it is for strikeouts, but it is certainly meaningful.

Is it? Looking at the raw numbers from ESPN's splits for the Brewers and it certainly doesn't look like it to me. Looks to be within the range of what we'd expect from random chance. Maybe if you extended the time line over several years you could get a non-chance result, but either way doesn't sound like a massive effect to me.

I did this with walks a few years back and found the same thing. I'm sure parks can affect walks, it just doesn't seem like the overall effect is very big.

Looks to be within the range of what we'd expect from random chance.

Wouldn't regression take care of the variance due to random chance?

Voros- What is your take on strikeouts then? It seems the spread is further between the most to the least than with with regards to walks but can a park really effect strikeouts and not walks?

When I checked strikeouts they seemed to be a large enough difference to be outside of chance. This was several years ago and the article isn't up anymore (I'm thinking of trying to find a bunch of stuff on the wayback machine and reposting it) though.

Walks were the only ones that looked close to random (and I believe maybe there was a park like the old Cardnials stadium where maybe it was an effect). I just like to note that when you see differences like 4 and 5 percent, you'd get those sort of things by dividing games up by odd and even dates instead of home and away. To me when you start getting up in the neighborhood of 20%, that's a park effect.

It's interesting that he notes that there seems to be a year to year correlation for the walk effects though. I wonder how the mathematics are working. My guess is that it's probably one or two parks where there's actual effects going on, with the rest being pretty much all noise. That might be enough to give him that level of correlation.

I suppose you could perform this same exercise but only use the visiting teams stats??

Then you have the same problem except in reverse, if any of the team's pitchers has a significant home/road split. I do think this would improve accuracy with a big enough sample size, because hitters are more likely to have persistent home/road splits than pitchers. (Or at least, I'd be surprised if that isn't the case.) But this is just reducing the problem, not making it go away.

It does seem like a bad idea to throw out half the data. On the time scale Gassko is using (2003-2007), any reduction of bias might be offset by an increase in the margin of error. I guess the correct thing to do is to simultaneously estimate players' home/road splits and park factors through some sort of maximum likelihood method. Sounds like a job for a real statistician.

Couldn't the amount of foul ground have an impact on Ks and BBs?

If there is less of a chance that my foul ball is caught for an out, then that means it's either a strike (with less than two strikes) or a dead ball with another opportunity to either walk or strike out as the at bat is extended.

My guess is that less foul ground = more extended at bats; more extended at-bats = more Ks and BBs.