October 3, 2012

Manufactured Runs

Mariners to Move Safeco Fences In

by Colin Wyers

the archives are now free. All Baseball Prospectus Premium and Fantasy articles more than a year old are now free as a thank you to the entire Internet for making our work possible. Not a subscriber? Get exclusive content like this delivered hot to your inbox every weekday. Click here for more information on Baseball Prospectus subscriptions or use the buttons to the right to subscribe and get instant access to the best baseball content on the web.

Subscribe for $4.95 per month Recurring subscription - cancel anytime. Subscribe for $39.95 for one year a 33% savings over the monthly price! Purchase a $39.95 gift subscription a 33% savings over the monthly price! Already a subscriber? Click here and use the blue login bar to log in.

Over time, baseball teams develop identities, and even if the players change, the identities stay the same. The Yankees (and the Red Sox, for that matter) have cultivated a slow, plodding, methodical style of three-true-outcomes baseball that scores lots of runs and takes its time doing it. The Cubs, until recently, were the team who struck out the world when pitching and wouldn’t take a walk to save their lives. The Twins were all about manufactured runs and control pitchers.

The Mariners have an identity, based largely around 1-0 losses and the brutal, crushing inevitability of death. And some of this can be attributed to the players—this is the team that has tried Adam Kennedy as a designated hitter in recent years. But as Egon Spengler put it, “It's not the girl, Peter, it's the building!” Safeco Park has been absolutely brutal on offense. This has probably not had much of an effect on wins and losses—what Safeco takes way from offense it gives back to pitching and defense. So Safeco isn’t causing the Mariners to lose, it’s just subtracting from the sum total of joy and happiness in the universe.

So the Mariners are going to alter their fences to increase offense:

The fence will be moved in from 4 to 17 feet at different points in left field and 4 feet from straight center to the right-center gap. Additionally, the 16-foot-high hand-operated scoreboard down the left-field line will be moved back and no longer be part of the fence, so the outfield wall will be 8-feet high all the way around the park.

Jeff Kingston, the Mariners assistant general manager who was part of a committee that studied the situation, said the changes won't turn Safeco into a hitter's paradise but will even the playing field to a degree.

"We still think it's going to be on the pitching side of the spectrum," Kingston said. "Our approach all along was to make it more fair and be closer to the middle. We still want it to be a pitchers' park and build around pitching and defense, but we wanted to give hitters a chance to where if they really square it up and hit it 390 or 400-plus feet that they'll be rewarded."

This is what the fence change will look like:

Except for the somewhat larger change in left-center field, essentially the fence change covers what until now has been the warning track. So something like this Miguel Cabrera flyout would, with the new changes, have a chance to become a home run:

So how do we go about establishing the potential effects of this change on offense at Safeco? The Mariners say their internal study leads them to believe moving the fences would have meant between 30 and 40 extra home runs per year each of the past few years (that’s total, so home runs scored by the Mariners and allowed by the Mariners' pitchers). If we watch television broadcasts and count deep fly balls in Mariners games in 2012, we can come up with similar totals, depending on one’s judgment of whether or not a ball would have been past the new fences (admittedly a speculative endeavor). But is there another way to establish the change in home run rates by moving these fences?

Thanks to Kevin Johnson’s invaluable parks database, we have a record of how the fences at various ballparks have moved over the years. So what I did was pull the changes in park dimension in left field, left center, center, right center and right field for years where that data was complete, and then look at changes in normalized home run rate over that time. (A methodological note—I kept all teams where there was sufficient data, not just the ones that had fences move.)

What I found was that a change in fence distance was (not shockingly) a meaningful predictor of change in a park’s home run rate, after controlling for changes at the league level; combining all five distances in a regression analysis gives us an adjusted R-squared of 0.18, which means there’s a lot that’s not explained by changes in fence distance but a fair amount that is. If we apply our model to the proposed fence changes, using 2012 home run rates as our baseline, I come up with 22 additional home runs from the proposed change.

That’s, I think, a plausible number, and in the grand scheme of thing not too far off what the Mariners found or what looking at video says. The other interesting finding is that changes in fence distance seem to be a very poor predictor of changes in doubles rates, actually turning up a negative adjusted R-squared. (I’m sure some of you are thinking that this shouldn’t be possible, but adjusted R-squared is adjusted in such a way that allows for negative values, although that’s a good sign that you should be reconsidering what you’re doing.) So in essence, each of those home runs predicted by our model would seem to otherwise be an out.

Using linear weights, that means that 22 home runs come out to 37 total runs. Divide by 162 to get runs per game (for what we’re doing, we need to prorate these over team games, not home games) and then by half to get the contribution to the offensive team only, that’s .11 runs per game. The Mariners have scored 3.77 runs per game this season, the lowest in the American League; with the additional home runs, that goes to 3.88. If we assume that instead the Mariners pick up the high-end estimate of 40 home runs, that’s .20 additional runs scored per game, or 3.97. The next lowest-scoring offense in the American League, the Cleveland Indians, scored 4.14 runs per game. The fence change, as envisioned, provides a small boost to offense, but Safeco is likely to remain one of the most pitcher-friendly parks in the league, old fences or new.

Colin Wyers is an author of Baseball Prospectus. Follow @cwyers
Click here to see Colin's other articles. You can contact Colin by clicking here