Petco and Safeco, Three Years In
To be perfectly honest with you, I don’t think about park factors very much anymore. Obviously, they matter as much as ever, but you just encounter them less since so many advanced numbers automatically fold them in. They go somewhat unseen, but they’re important, and I was recently reminded that three years ago, Petco Park and Safeco Field debuted new dimensions. There are other factors that affect how stadiums play, like weather patterns and nearby construction, but what’s most important tends to be the shape of a given field itself. So now that we have a good amount of data, let’s see how Petco and Safeco have played more recently.
To be straight, what follows isn’t very rigorous. I didn’t make adjustments or regressions, and almost anyone would tell you that the ideal involves more than three years of information. There are ways to do this more precisely. But, three years are three years, and it shouldn’t be hard to observe any significant changes. Off we go!
This is what happened to Petco:
There have been more recent alterations, but they were much more minor in scope. The big changes happened between 2012 and 2013. Note that there’s been some area construction in downtown San Diego that has presumably had an effect on wind. That probably matters. It probably matters less than the fences. Now, this is what happened to Safeco:
The focus in San Diego was making right and right-center more hitter-friendly. The focus in Seattle was making…almost everything more hitter-friendly, but with a concentration on the left-center power alley. Used to be, Petco was murder on powerful lefties, and Safeco was murder on powerful righties. So changes were made to try to make the ballparks more fair. Nothing wrong with the idea.
Let’s look at results, beginning with Petco Park. I looked at three years of data before the changes, and the three years since. This is a table with a bunch of percentages. What those show is the performance at home divided by the performance on the road. For example, take that first 92% next to BA — that means the batting average in Padres games in San Diego was 92% of the batting average in Padres games elsewhere. (Which means Petco reduced batting average, which is one of those pitcher-friendly indicators.) The last column shows the change between the earlier columns, expressed in percentage points. Any perceived differences are just due to rounding. So 0% isn’t 0% — it’s just less than 0.5%.
| Statistic | 2010-12 H vs. R | 2013-15 H vs. R | Change |
| BA | 92% | 93% | 1% |
| OBP | 95% | 96% | 0% |
| SLG | 90% | 93% | 3% |
| ISO | 86% | 93% | 7% |
| BABIP | 96% | 95% | 0% |
| 2B/3B% | 97% | 91% | -6% |
| HR% | 79% | 98% | 19% |
| BB% | 105% | 102% | -3% |
| K% | 108% | 107% | -1% |
| R/PA | 86% | 87% | 1% |
Like with any set of baseball statistics, there’s noise in here. Still, you can make some observations. The first one, and the strongest one: San Diego has, indeed, become more homer-friendly. The last three years, there have been almost as many homers in Petco as there have been in non-Petco games. In that sense, it’s been mission accomplished. It gets even more interesting if you look in other places, though. Doubles and triples have been somewhat reduced, in terms of frequency per batted ball. Stands to reason some of what used to be doubles and triples are now homers. There’s maybe been a small effect on walks and strikeouts. And if you look at the last line, actual run-scoring has hardly budged. Petco, for three years, has remained strongly pitcher-friendly overall.
Now, because there’s a reduced effect on OBP and slugging percentage, it’s probable that Petco is more hitter-friendly than before. The last line could just be capturing some noise. Even still, the effect of the changes on homers seems far greater than the effect on runs. Based on three years, Petco remains a pitcher’s environment.
It’s worth noting that Petco was less pitcher-friendly in 2015 than it was the two years before. Because of some minor alterations, that can’t be ignored, and so it could be a better sign of the “true” Petco Park. Still, pitchers should like it there. Even despite the increase in dingers.
Moving on, we’ve got the same table, this time covering Safeco Field:
| Statistic | 2010-12 H vs. R | 2013-15 H vs. R | Change |
| BA | 91% | 95% | 4% |
| OBP | 94% | 95% | 1% |
| SLG | 86% | 94% | 8% |
| ISO | 77% | 92% | 15% |
| BABIP | 94% | 97% | 3% |
| 2B/3B% | 82% | 91% | 9% |
| HR% | 77% | 97% | 20% |
| BB% | 105% | 95% | -10% |
| K% | 109% | 107% | -1% |
| R/PA | 80% | 92% | 12% |
Just like with Petco, you see a big jump in homers, to something just about average. But unlike in Petco, there’s been a corresponding big jump in runs, with a substantial increase in doubles and triples. It would seem fewer fly balls have just been dying on the track. BABIP is up a little bit, but of some interest, it seems like there’s been a strongly reduced effect on walks. I don’t always know how to explain those things, if they’re real.
For ease, here’s a direct comparison of the last columns from each table:
| Statistic | Petco Changes | Safeco Changes |
| BA | 1% | 4% |
| OBP | 0% | 1% |
| SLG | 3% | 8% |
| ISO | 7% | 15% |
| BABIP | 0% | 3% |
| 2B/3B% | -6% | 9% |
| HR% | 19% | 20% |
| BB% | -3% | -10% |
| K% | -1% | -1% |
| R/PA | 1% | 12% |
Based on the evidence so far, the changes have made a bigger difference on batting average in Safeco. Ditto OBP — by a little bit — and ditto slugging percentage. The change in effect on isolated power has been double, and then you see a “reduced reducing” effect on BABIP. The home-run rows are basically equal, showing that the changes in large part accomplished what was intended, but the last row is telling. As noted before, there’s been a marked increase in runs at Safeco. Not so much to this point in Petco.
That one, we can keep watching, especially if Petco looks more like its 2015 self than its 2014 self. And just generally speaking, it’s clear the parks have become less pitcher-friendly, on account of the increases in OPS. But there’s a neat lesson in here about how park factors are complicated. They designed Petco to be smaller, and they increased homers while reducing doubles and triples. They designed Safeco to be smaller, and they increased homers while also increasing doubles and triples. I don’t know why that would be, but here we are. Could be it’s all just a bunch of noise, but there have been three years for this to play out.
The parks in both San Diego and Seattle remain more pitcher-friendly than hitter-friendly. That much we can say with a good amount of certainty. Yet there’s been far less complaining, because at least the increase in homers has created a perception of fairness. Pitchers overall still get to benefit, but as long as hitters aren’t flying out after hitting the ball on the screws, you’ve got a good-enough ballpark. Ultimately that’s all anyone wanted anyway.
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.
New scoreboard at Petco Park wasn’t an alteration to dimensions, but it did affect how wind flowed in and out of the ballpark. That has to be in consideration as a pretty major factor in the jump in home runs between 2014 and 2015.
Obviously, we’ll see if that remains true going forward.
Its a shame more attention is not being paid to park factors because they are badly applied and their adjustments can be worse than no adjustments at all. Not taking handedness into account in many adjustments, or a hitters/pitchers batted ball profile.
There is a tremendous need for improvement, but most everyone seems to ignore it and assume the adjustment makes sense.
I agree, but there are prob. as many over-fit park factors out there as over-generalized one. That is, the stat is incredibly noisy to start with, so if you slice the data into splits, it becomes even more unreliable for predictive purposes. In short, a whole lot more study is needed to design the optimal predictive park factors, and no one seems eager to do it.
They’re both still extreme pitchers parks, but at least Safeco is no longer absurd. I’m pretty sure the construction and weather which change the winds have a significant affect on fly balls to left in Safeco as well.
I was so disappointed in Seattle when they brought the fences in. As Adam says, Safeco used to be absurd, but that absurdity was awesome.
I counted myself a Mariners fan for 2001-2012, but after they brought in the fences I lost interest in the team.
This is the randomest reason to lose interest in a team I’ve heard in a while.
Especially an Ms fan, the fences were the breaking point
Right? I was really disappointed as well when they moved the fences in. I really liked the fact that the extreme environment set Safeco apart from a run-of-the-mill baseball park.
I also wonder how changes to the staff’s of the home teams affect park values. For example, does Petco Park’s jump in home run rate truly reflect changes to the park or the fact Ian Kennedy and James Shields had elevated home run rates? Is that the park, the pitcher, or the catcher? We study framing, how much do we look at how a catcher calls a game?
The percentages are comparing games played at Petco versus Padres games played elsewhere, so the Padres players involved from year to year shouldn’t impact this much unless Ian Kennedy pitched like 3/4 of his games in Petco or something.
I seem to remember conjecture here in San Diego that in 2015 the drought had tampered down the marine layer affect that typically plays a large part in slowing the flight of the ball.
The increase in HR for those little shifts in the fences seems high, but… could be there were that many warning-track flies that are now over the fence. I wonder what ?HR% a foot of fence motion is worth, typically.
I’m going to assume that hitters adjust their approach when they feel they have a chance to hit one out, which would exaggerate the impact of a few feet’s movement of the fences. Honestly, Seattle weather has been weird the last few years (hotter in Summer, less overcast in general), so global climate change is probably more important than fiddling with the boundaries.
Two words – Ian Kennedy.