Moving the Fences: A Follow-Up
On Monday, I looked at a few ballparks that, in the past, had moved in their fences in an attempt to make the park more hitter-friendly. The Royals did it in the mid-90s, the Tigers did it between 2002-2003, the Padres did it between 2005-2006, and the Mets did it between 2011-2012. This very offseason, the Padres are doing it again, and the Mariners are doing it as well. Based on data, the home-run factors for each ballpark went up after the adjustments, but the run-scoring factors didn’t, which I found to be of interest.
This post, as indicated, is a follow-up, made necessary for two reasons. For one, I overlooked the White Sox, who, between 2000-2001, moved in the fences in what’s now known as U.S. Cellular Field. I don’t know how that escaped my attention before. For two, if we’re going to look at places that moved in the fences, we should also look at places that moved out the fences, because we can learn from those examples too. Between 2003-2004, the Royals moved the fences in Kauffman back to where they were long before, prior to the adjustments in the 90s. And, between 2005-2006, the Phillies made adjustments at Citizens Bank Park, in response to the feeling the place was a bandbox. Once again, we examine the data, in order.
U.S. Cellular Field
U.S. Cellular has undergone a lot of renovations, and down the two foul lines, the fences were moved in before 2001, making room for the bullpens and the Bullpen Sports Bar. At least, that’s what Wikipedia is telling me. It’s not that anybody thought U.S. Cellular was too pitcher-friendly, I don’t think. Over three years, between 1998-2000, the home-run rate at U.S. Cellular was 107% the home-run rate in White Sox games on the road. Over the same three years, the run-scoring rate at U.S. Cellular was 101% the run-scoring rate in White Sox games on the road. Over the next three years, between 2001-2003, the home-run rate at U.S. Cellular was 126% the home-run rate in White Sox games on the road. The run-scoring rate at U.S. Cellular was 103% the run-scoring rate in White Sox games on the road. Homers went way up; runs went up a very tiny bit. We know U.S. Cellular today as something of a bandbox.
Kauffman Stadium
And so we re-visit the home of the Royals, who between 2003-2004 changed their minds from the decade before. Now we’re talking about fences being moved back, and not in. Over three years, between 2001-2003, the home-run rate at Kauffman was 114% the home-run rate in Royals games on the road. Over the same three years, the run-scoring rate at Kauffman was 128% the run-scoring rate in Royals games on the road. Over the next three years, between 2004-2006, the home-run rate at Kauffman was 80% the home-run rate in Royals games on the road. The run-scoring rate at Kauffman was 101% the run-scoring rate in Royals games on the road. Homers went way down, and runs also went way down.
But this is odd. As we noted yesterday, after the Royals first moved the fences in, the park played neutrally, if somewhat pitcher-friendly. Over the course of some years, the park started to play hitter-friendly. There were no dimension adjustments, so I’m left thinking about the sample sizes we’re dealing with. Kauffman probably wasn’t as neutral as it seemed immediately following the first change, and it probably wasn’t as extreme as it seemed prior to the second change.
Citizens Bank Park
It took just two years for the Phillies to decide their new stadium was too kind to hitters, and too rude to pitchers. So back went some of the fences, as the Phillies tried to even things out. Over two years, between 2004-2005, the home-run rate at Citizens Bank was 119% the home-run rate in Phillies games on the road. Over the same two years, the run-scoring rate at Citizens Bank was 109% the run-scoring rate in Phillies games on the road. Over the next three years, between 2006-2008, the home-run rate at Citizens Bank was actually 123% the home-run rate in Phillies games on the road. The run-scoring rate at Citizens Bank was 104% the run-scoring rate in Phillies games on the road. While, over the years observed, homers didn’t go down, runs went down, which was basically the idea. Citizens Bank developed a certain reputation early on, but these days it’s unobjectionable.
Now we have a better understanding of all the data, and we also have this helpful comment from Monday from MGL. When fences have gotten moved in, homers have become easier to hit, but there hasn’t been much impact on overall run scoring, probably owing to reduced singles, doubles, and/or triples. When fences have gotten moved out, there seems to have been an impact on overall run scoring, as runs have gone down, relative to runs on the road. Citizens Bank is a little tricky to figure out, based on our window above, but the FanGraphs park factors suggest the homer factor is now lower than it was. So that’s more intuitive.
My number-one conclusion is that I wish we had more examples of ballparks changing their dimensions. My number-two conclusion is that park factors are complicated, and effects of changes aren’t always easy to predict. You might be able to predict effects on specific things — changing the fences does seem to change home-run rates — but, overall, there are a lot of moving parts, a lot of things to put together. As far as the Mariners and Padres are concerned, 2013 should bring more frequent home runs than 2012 and seasons before. But in terms of the ease of run scoring, Safeco and Petco might change surprisingly little. There’s reason to believe it’ll be easier to score down the road, but the magnitude of the increased ease could be quite modest.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
I hate to get all touchy feely but homer reliance doesn’t seem to translate to more runs or more championships. It will be interesting to see how the Prado/Upton trade works out. Arizona seems to believe they can score more runs and especially, more important “close game” runs by cutting down on strike outs and home run reliance. While sacrificing Upton’s ceiling is risky, I think they are on a winning course.
So did they change the fences at Chase Field? Or did you intend to post this comment to some other article where it was actually on-topic?
Ouch.
Did you have anything constructive to add? Or did you just come here to displace your anger issues?
I believe the connection lies in the idea that a higher home-run rate doesn’t necessarily equate to higher run-scoring rate.
And how does this relate to the article?
I remember once in Grade School, I accidentally went in to the wrong classroom — oh gods, for a fourth grader, that was humiliating.
Oh, please tell us more about this interesting story!
A couple things:
(1) I don’t think run creation is even on teams’ minds when they move the fences in. I think they move the fences in ONLY to create more homeruns which in turn creates more fans in the stands (and on TV). After all, chicks dig the long ball.
(2) It would be nice to know the exact year in which the Royals moved the fences in during the mid-90s. It could be that the closer fences coincided with the increasing use of PEDs thus yielding the extreme homerun/run home/road splits that is cited in the article. I always think of the offseason prior to the 1996 season as the year when PEDs became prevalent. I have no real data here, other than Brady Anderson.
It was 1995 when the Royals changed from turf to grass, lowered the fences from 12 to 9 feet, and shortened the power alleys from 385 to 375 and CF from 410 to 400.
Another example of a team changing ballpark dimensions is the Orioles back in 2001 – they moved home plate back by seven feet (rather than changing the fences), only to put it back where it was after one season. Probably not a large enough sample to make any inferences, but worth noting.
And to add yet another facet to the sample, couldn’t you also use Coors Field as an example? They didn’t change fence distances in or out but the storing of baseballs in a humidor accomplishes the same effect as moving the fences out. And seeing as you’re not using specific distances in your discussion (feet in, feet out) I believe this could be part III of what now will be your now ongoing Moving the Fences Series. It could be like a written soap opera, As the World Turns or Days of Our Lives……..The Young and the Fenceless!
Kind of off-topic, but I actually would love to see baseball played without fences. The only home runs would be inside-the-park home runs. It’d be really interesting to see how outfielders and hitters reacted to this situation. I imagine it’d reward fast players over big sluggers, but I’m not entirely sure.
On a practical level, a 500-foot or so fence would achieve the same effect.
You were born 100 years too late.
Yeah, if nothing else, bring back the Polo Grounds. Stadiums today mostly aren’t very much fun.
Thing is, I’d like to see this with modern players, who are better atheletes and more skilled baseball players than 100 years ago.
And I, too, like quirky parks. Fenway may be a bandbox, but it’s pretty interesting. Houston’s stadium, at least when the roof is open, is pretty cool. Love Tal’s hill.
Of course, I also like Miami’s home run feature.
I’m a bit sad about San Diego and Seattle moving their fences in – I like variety. I enjoyed RFK quite a bit despite it supressing home runs for a lot of hitters, though not enough to stop Soriano’s 40/40 season.
Jeff, you can’t look at parks individually. You will necessarily see/get lots of random fluctuation. It would be like if you took a die (as in dice) and replaced the “1” with a “6” and then had 100 people roll the die anywhere from 5 to 20 times each. And you reported that some of them came up with fewer than 1 in 6 sixes and some of them came up with more than 1 out of 6 sixes. You don’t conclude, “Well the results are inconclusive since some people had an increase in the frequency in the number of sixes and some people had a decrease.” (Obviously it is very likely that a lot more had an increase.)
You combine the results and have each person roll the die as many times as possible and then your try and draw some conclusions.
So to with this park “study.” You combine the results and look at as many years as possible (not just 2 or 3 years after the change).
I’ll bet my hat that if you did that, you would get a statistically significant result in the direction of a positive correlation between fences out and decreased run scoring, and vice versa.
Also, some of the changes you looked at were weak (like PHI in 06) and some were mixed (like the WS in 01 – some parts were actually moved OUT). Others were strong (like KC and the Mets last year).
This year the moved by SEA and the Padres appear to be strong moves. I would suggest that the true run factors for both parks are going to be increased significantly. OF COURSE what we will actually see in 2013 is going to be subject to random fluctuation so OF COURSE there is no guarantee that we will see an increase in the run factors for those teams – but who cares about that? We only care about what these changes will do to the TRUE park factors.
And I think it is misleading to suggest that we don’t know to any level of certainty what these changes will do to the TRUE PF’s. We do. There is a very high chance that it will increase those park run factors.
It is one thing to make a conclusory statement, such as “the evidence does not suggest an answer to any high degree of certainty,” (as you clearly state in this case), when you do a well-constructed, rigorous study. It is another to come up with the same conclusion when you do a “study” that is not really a study at all and suffers from all kinds of maladies, namely severe sample size issues. OF COURSE your results will not lead you to any significant conclusions. Even if you had come up with a strong connection between moving fences and park run factors, those results STILL would not mean much because you are not really doing a scientifically valid study…
I can pretty much predict what’s going to happen next year at Safeco. Everybody knows that Safeco has been a pitcher’s park, and will likely to remain so with the new fence alignment. But last year was Safeco to the extreme.
In one game at Texas, the Mariners scored 21 runs. There’s no park factors that can explain why the usually inept offense of the Mariners could score 21 runs in a game, no matter where they’re playing. There were back-to-back games in Arizona where the two teams combined for 45 runs. While better hitter’s parks make these things more likely, they are still way out of the norm, and sometimes these things happen for no other reason than things just went wonderfully well for the hitters that night.
Back in Seattle, there were two no-hitters and one perfect game thrown. While a pitcher’s park makes it easier for these things to happen, it only makes it very slightly easier. These are still extreme events, and you wouldn’t predict even one of these things happening in any one year even in the best pitching park you could imagine baseball ever having.
You don’t have to have too many of these extreme events in one season to skew a park factor in one direction or another — just as you said, one-year park factors can be subject to a lot of variation.
Moving the fences in will likely result in a boost of offense for the ballpark. I can’t really see it making a huge difference, but I’m sure it will have a positive effect on run scoring. But just as likely is that the difference in runs scored home/away is going to revert to more normal levels. You put those two together, and the difference between runs home/away is likely to decrease a lot.
And here’s where my real prediction is: Everybody is going to point to the increased scoring in Safeco, and the change in park factor, and ascribe it to moving the fences in. There’s going to be a lot of talk about how the new dimensions have dramatically improved offense at Safeco. But you know what? The majority of that change will probably be due to simple regression to Safeco’s mean, rather than the change in the dimensions.(well, not really the mean, because the mean will most likely change next year, but you know what I, uh, mean.)
Whoops! Two perfect games and one no-hitter. Why does it always take till after I post something to remember (not so) little facts like that?
Hi Jeff. Yes it is hard to isolate the effect of fences being moved due to many variables, but did u try to account for them? I would like for you to try to account for the change in average opponent uzr home and away, before and after move. Also the away park factors and how all these impacted the run scoring etc. I don’t think it is good enough to say it’s hard and give up.
I am not in position to devote myself to one study for like one or two or three weeks. Consider this an opportunity for somebody else.
Yes, I’ve noticed you have taken the role of pumping out content, which you do a great job of.
I just want to note that I think this throwing out of observations and hoping other people will find them interesting enough to pick up on is a really good idea.
Also, can someone point me in the direction of a list of the dimensions of each ballpark, including foul territory?
All dimensions for all MLB parks for all seasons (2012 about to be posted)
http://www.seamheads.com/ballparks/index.php
Dan, there are several good web sites that last the dimensions including periodic changes of all parks past and present. The one I use a lot is Clem’s. google it.
No one gives foul territory dimensions AFAK other than distance to home plate to the backstop. However websites like Clem’s give you a to-scale diagram of each park. You can infer the foul territory from that.
The size of foul territory has a large impact on run scoring. Look at dodger stadium over the last 10 years or so. The have decreased foul territory a number of times. It used to be an extreme pitchers park. Now it is only a mild one.
Also you can use a program (there are several good freeware or shareware ones) to compute the exact amount of foul territory in each park from the diagrams. If you do a search on my old blog, you might find an old list of foul territory areas in square feet.
MGL, thank you for the detailed response. Much appreciated.
There are two other factors in play with Safeco. 1. The fence height has been standardized to 8 feet. Due to the scoreboard down the left field line being moved that space in no longer 12 high. 2. The ambient air temperature is lower and the humidity is significantly higher than parks in warmer or higher elevation parks. Cold moist air at sea level produces more drag on a batted ball than in the warmer and higher parks, enough to make a real difference. The internal air flow in Safeco also flows from left field toward home plate most often due to external wind direction and building configuration.