Safeco Field in 2016: It’s Weirder Than I Thought

marc w · July 14, 2016 at 6:12 pm · Filed Under Mariners 

My post a few days ago was perhaps short on data about Safeco Field itself and long on speculation. There’s nothing really wrong with that, I don’t think (but I *would* think that, wouldn’t I?), but without any real conclusions beyond “Safeco’s 2016 spike in HRs could be the result of any number of things,” it was pretty unsatisfying. The premise rested on a series of hypotheses that it’s probably time to make plain. First, that the M’s (and perhaps their opponents too) are throwing more pitches up in the strike zone, leading to more fly balls. Second, that some pitchers aren’t really suited to this, and therefore their “high strikes” are hit like Chris Young’s in 2016, not like Chris Young’s in every other year. Third, that the M’s believed Safeco field to suppress the value of fly balls whether it actually does so anymore or not. I decided to look into these assumptions using the incredible’s statcast/pitch fx search tool. And now I’m not sure that any of my assumptions were right. Safeco Field really DOES seem to be making each fly ball more dangerous, but it’s got nothing to do with how the M’s themselves are pitching. Where that leaves us, I’m not entirely sure.

I decided to look at how often pitchers in Safeco Field (M’s and their opponents) have thrown pitches in the top of the strikezone and higher, and what happened when batters put those pitches into play. Safeco’s dimensions changed between 2012 and 2013, so I thought it’d be good to look from 2012-2016; we get one year of “old Safeco” and then some trend data on “new Safeco.” I also did the same thing for the league as a whole, so we have something to compare the Safeco results to. This table shows the slugging percentage on contact of high pitches by season. Two data points kind of jump out of the table, I think:

2012 0.420 0.507
2013 0.516 0.491
2014 0.499 0.468
2015 0.497 0.505
2016 0.612 0.514

(HP SLGCON= The slugging percentage on contact of all elevated pitches in Safeco, LG SLGCON=The slugging percentage on contact of all elevated pitches in MLB)

In 2012, the league *slugged* .420 in Safeco on high pitches, compared to a league average of .507. To make a crude “+” metric comparing Safeco to the average, 2012 comes in at 82.8. But look at 2016! Out of nowhere, batters are slugging *.612* on those same pitches this year. Whereas Safeco ranked dead last in MLB in 2012, it’s now 2nd (barely) behind Coors Field. The 2016 SLUGCON+ figure is 119, or about as far from average now as Safeco played in 2012. That’s…that seems insane to me, even as all of us know that a hell of a lot of HRs have been hit in Seattle this year.

So is this because pitchers and coaches *remember* old Safeco field and blithely assume that the marine layer will absolve their pitching sins? Does it cause pitchers to throw more high pitches (which are hit in the air more often) because they assume, perhaps wrongly, that fly balls are harmless in Seattle? Well, no, that’s not what’s happening. This next table compares the percentage of all pitches that are elevated, by year, in Safeco and MLB as a whole. To see if the M’s are pitching differently than their opponents, the last column includes only Mariners hurlers at Safeco.

Year HP% LG HP% M’s HP%
2012 34.24% 33.97% 35.49%
2013 31.26% 34.11% 28.99%
2014 31.66% 32.92% 30.79%
2015 28.62% 31.10% 27.75%
2016 29.12% 30.66% 29.56%

(HP%= Percentage of elevated pitches in Safeco, LG HP%= Percentage of elevated pitches in MLB, and M’s HP%= Percentage of elevated pitches in Safeco thrown by Mariners pitchers)
Outside of 2012, Safeco’s seen *fewer* high pitches than the league average. Even in 2016, with Nick Vincent, Nate Karns and the like, the M’s are solidly below average in the fraction of pitches they elevate. That’s interesting, given the numbers we looked at in the last post – the M’s giving up more fly ball contact at home, more strikeouts, fewer walks.

So, league-wide, batters are doing a bit more damage on high pitches, and they’re doing a TON more damage in Seattle specifically, so perhaps the league is right to avoid throwing elevated pitches. I decided to run the same analysis but with low pitches – the bottom 1/3 of the zone and below. Here, Safeco Field doesn’t look all that remarkable. It’s got a lower SLGCON than the league average, but 2016 looks just about the same as 2013. 2012 was again the low mark, but Safeco’s SLGCON on low pitches has risen along with the rest of the league.

2012 0.405 0.48
2013 0.478 0.463
2014 0.421 0.468
2015 0.477 0.483
2016 0.48 0.517

(LP SLGCON= The slugging percentage on contact of all low pitches in Safeco, LG SLGCON= The slugging percentage on contact of all low pitches in MLB)

And as you’d imagine after seeing that pitchers are throwing fewer high strikes now, they’re targeting the bottom of the zone more often:

Year LP% LG LP% M’s LP%
2012 49.67% 50.35% 48.16
2013 51.55% 49.98% 53.52
2014 51.80% 51.09% 53.05
2015 56.05% 53.38% 57.06
2016 54.78% 53.71% 54.66

(LP%= Percentage of low pitches in Safeco, LG LP%= Percentage of low pitches in MLB, and M’s LP%= Percentage of low pitches in Safeco thrown by Mariners pitchers)

It makes sense that pitchers are throwing the ball lower than ever, because that’s precisely where the strike zone’s been growing. These new strikes are harder to elevate, and thus it’s kind of a pitching sweetspot – you’re rewarded if they don’t swing, and not punished too much if they put it in play. So good on the M’s for throwing at the knees! Well, not so fast. Look at that final league-wide SLGCON figure. For the first time in this little 5-year look, batters have a higher SLGCON on LOW pitches. That’s clearly not true at Safeco, but it’s true for the league as a whole.

Moreover, it seems like it’s harder for batters to actually put those high pitches in play (this part’s much less counterintuitive than the SLGCON thing). The percentage of high pitches that are actually put in play is lower than it is for low pitches, and what’s more, it’s still trending down. Back in 2012, batters put 5.3% of high pitches in play in Safeco, nearly identical to the league-wide rate. Now, batters put less than *4%* of such pitches in play in Safeco, a much lower percentage than the league as a whole:

Year Con% LG Con%
2012 5.29 5.27
2013 4.73 5.22
2014 4.84 4.89
2015 3.91 4.5
2016 3.83 4.21

High pitches seem harder to hit in Safeco, but they’re now doing much more damage when they ARE hit. This same trend towards less and less contact doesn’t show up with low pitches – they’re put in play just under 9% of the time, which is a bit higher than the 8% of 2012-13. The big trend for low pitches is somewhat worrying: HRs per pitch are going up, and they’ve blown past HRs per pitch on high pitches. Batters appear to be adjusting to the new, lower zone by elevating lower pitches better than before. Given that batters put more low pitches in play, AND hit more HRs, maybe high strikes aren’t as bad as we thought.

That hasn’t helped the M’s, and it doesn’t help us understand what’s going on in Safeco Field this year, where fewer high pitches have produced lots more dingers. At this point, we’re back to speculating. Commissioner Manfred’s already come out and denied that the baseball itself is behind the trend towards more HRs, but everything we’ve seen would be consistent with the “juiced ball” hypothesis. Both in Safeco and the league as a whole, batters have a higher SLG% on contact for high and low pitches. They’re making *less* contact on high pitches and more on low pitches, which could also be the result of teams now consciously or unconsciously selecting for low-ball hitters – these players may be harmed less by the strikezone’s continued southward march. By the same token, maybe teams are now selecting for low-ball pitchers, especially those who can command breaking balls and fastballs down without relying purely on sinkers and two-seamers (as that would show up in league-wide ground ball rates) – guys who can get ahead in the count by sneaking a four-seam or slider at the bottom of the strike zone for a called strike one. Maybe the marine layer took an El Nino-themed vacation for a year, I don’t know. Whatever it is, it’s NOT the case that the M’s are targeting the upper zone; wherever they’re getting their fly balls from, it doesn’t look like it’s high strikes. Is it possible that as more and more pitchers throw more and more low strikes that they’ve essentially forgotten how (or never learned) to throw up effectively? I have no idea, but if so, it’d kind of be a sabermetric own-goal – many of us have preached the value of the ground ball and the risks of pitching up, but the facts on the ground may be changing. Just like last time, there are no clear answers here, just more questions.


5 Responses to “Safeco Field in 2016: It’s Weirder Than I Thought”

  1. mksh21 on July 14th, 2016 7:06 pm

    Dude the last two posts have been amazing, I’m shocked at the level of research and detail.. However I am way too dumb to understand half of it 🙁 But Im gonna attempt it lol.

  2. Dennisss on July 14th, 2016 8:26 pm

    I have said before that I feel like I should get college credit for reading a Tony Blengino post. This stuff is more like upper division college credit.

    In any case, after this year, maybe we can stop saying that you can’t hit home runs at Safeco, and we’ll go back to blaming the players.

  3. Notfromboise on July 14th, 2016 9:22 pm

    This was an amazing read. Thank you, marc!

    Its easy to say King’s missed half his starts and the pen’s missed Furbush or Scribner.. But there is some weird Twilight Zone-y combination of Safeco evolving and our pitching devolving that has made for a really, really puzzling season.

  4. deflated on July 15th, 2016 10:39 am

    Great post, I have some reservations on the Mariners numbers. Over that period the Mariners had two outstanding pitchers – Felix and Iwakuma – who lived at or below the bottom of the zone. Telling either of them to throw high fastballs at the expense of their change/sinker makes no sense (thought: is Felix’s absence driving the low M’s LP% in 2016?).

    My gut feel is that HP/LP changes are really only going to be reflected with ‘coachable’ pitchers – 4/5 journeyman starters, young prospects, etc.

  5. Steve Nelson on July 16th, 2016 2:49 pm

    It also suggests that Jeff Kingston got it right when he was in charge of the project to move in the fences. At the time, there considerable speculation that the changes were more of a tweak, with only minor changes in offense. As these posts indicate, as the league adjusts to the changes, the impacts are showing to be more dramatic.

Leave a Reply

You must be logged in to post a comment.