FanPost

Mariners, Safeco, and Home-Road Splits

As one result of this season so far, we Mariners fans have heard a lot of cries to move in the fences, both from fans and from certain media members. This stems from the Mariners' sheer offensive ineptitude within the unfriendly confines of Safeco Field, in particular one Mr. Justin Smoak's propensity for warning-track shots.

We all know this, of course. The question is, what exactly is spawning the Mariners' incredible impotence at home? How can they have the AL's third-best road wRC+ and yet be so godawful in Seattle? I decided to do some digging on Fangraphs and see what might be happening.

In particular, this FanPost was inspired by two comments on Jeff's recent (delightfully URL'd) article about Michael Saunders, one by Boxburger and the other by ArrwCtchr. The comments:

The problem is the M's are 3rd in the AL in road wRC+

Yet their road BB% is the 2nd worst in the league. Pitchers should be afraid of a top 3 offense like the M’s have, and walk them a ton, but our aggressiveness doesn’t let us.

And they have the 2nd highest BB% at home

I think all of us would rather they hit like they do on the road while taking less walks than whatever it is they do at home.

Well, walks are good, right? How can the Mariners walk so much at home and still be so godawful there? In short, what the fuck is actually going on?

THEORY ONE: APPROACH CHANGES

It's not just walking that the Mariners are doing more of at home--it's striking out. The Mariners do indeed have baseball's second highest home BB%: 10%, outpaced only by the Rays (10.7%). But they're a whole percentage point ahead of second in home strikeouts, K'ing at a 22.5% clip. Oddly, neither of these figures are matched on the road, where the Mariners K at a 19.7% rate, good for eleventh-lowest in baseball, and walk at a 6.9% rate, also well below average. In short:

Mariners at home: 10.0% BB, 22.5% K, 67.5% balls in play

Mariners on the road: 6.9% BB, 19.7% K, 73.4% balls in play

On the road, the Seattle squad puts down the seventh-most balls in play of any team in baseball. At home, the Mariners drop down the LEAST balls in play of any team in the major leagues, and it isn't very close. Now, Safeco inflates walks, and Safeco inflates Ks... but not by that much. If the Mariners' true talent level is their road performance, we would expect them to walk much less AND strike out much less at home. If their true talent level is their home performance, we would expect them to strike out much more AND walk much more on the road.

It appears as though the Mariners are indeed undergoing a fundamental change in approach when they leave Safeco Field. So what exactly is this change in approach? Less balls in play suggest more taken pitches, which suggest much less "aggressiveness". It's not just one or two batters that look different, either. Almost every Mariner walks less and strikes out less on the road. The only Mariners who have a higher BB% AND a higher K% on the road than in Safeco are Casper Wells, Justin Smoak, and Miguel Olivo, and in the cases of Smoak and Olivo it's very very close. This appears to be a team-wide thing, which makes me wonder if Safeco is at least in part intimidating the Mariners' young hitters into not being as aggressive as they normally would.

Unfortunately, Fangraphs doesn't carry Home/Away splits for the PitchFX plate discipline numbers, so it's impossible for me to verify this hypothesis at the moment unless I can get more data. Essentially what I'd expect to see is a much lower swing percentage at home, particularly on strikes, coupled with a lower contact rate.

Now there are a couple of things that could cause this problem and thus a couple of solutions. First, it could just be sample size screwery, and maybe the Mariners really aren't changing their approach at home. This is the answer implied if there is no significant difference in their swing rates at home versus on the road. In this case there's really nothing the Mariners can do but sit it out and hope that the statistics normalize themselves. On the other hand, the Mariners might be telling their hitters to be more patient in Safeco, since swinging for the fences is less likely to work there than it is in, say, Arlington. If they're doing that, they need to stop, now. Alternately, instead of telling their hitters to adapt to the park, they could adapt the park to their hitters. If the problem is indeed psychological, moving in the fences would likely to a great deal to help fix it. Honestly, aside from elevation making pitches bend more, it's hard to see why the Mariners would put so many fewer balls in play at home than on the road. I think there must be some coaching shenanigans going on here.

THEORY TWO: BATTED BALLS

The Mariners have an interesting discrepancy in batted balls at home versus on the road that might be hurting their performance. Now, you should take this one with a grain of salt or two, because batted ball profiles are rather heavily dependent on scorers (as I learned in the comments of my previous FanPost about BABIP and LD%). However, the differences here are actually quite interesting and seem to bypass scorer error at least a little bit, seeing as it's hard to confuse a ground ball for a flyball.

The Mariners hit a very similar number of line drives wherever they are; 20.9% on the road and 20.8% at home. However, the rest of their batted ball profile varies widely depending on where they're playing. The Mariners have a 44.2% GB% on the road as opposed to a 41.4% figure at home. Unfortunately, the missing ground balls are all shunted into fly balls and infield flies. The Ms hit .8% more infield flies at home than on the road (where their infield fly numbers are already bad) and 2.9% more fly balls. Also, their HR/FB practically halves when they're in Safeco, shooting down from a slightly high 12% to an abysmal 6.3%.

Now, this batted-ball profile matters because Safeco is an absolutely awful place to be hitting fly balls. Batters have, in the past, complained of a "pillow effect" in Safeco with the roof open. Supposedly, cool air blows in off the ocean from LF to RF, cushioning flyballs as they land. This boosts center fielder's defensive numbers in Safeco, making very good defensive center fielders look like legendary defensive center fielders. (See: Mike Cameron, Franklin Gutierrez.) Apparently the cushioning effect stops to a large extent when the roof is closed.

The practical upshot of all of this is that when the Mariners do put balls in play in Safeco (which is rare, as discussed in Part 1) many of them are fly balls, which die. Their struggles with .SLG at home are thus explained: fly balls which might go for home runs in other parks are swallowed by Safeco's dimensions and the humid air, and fly balls that would normally be doubles get "floated" on the crosswind and caught by outfielders. This explains the downwards spike in ISO for the Mariners at home; they have a .160 mark on the road and a .106 mark in Safeco. Obviously, without extra-base hits you don't score, and the Mariners haven't had a whole lot of extra-base hits in Safeco. The Mariners do historically have a higher road ISO than their home ISO since Safeco was built. However, the gap has never before been quite this big--I have little doubt that there's some luck involved here.

It's also no great surprise, then, that the Mariners have baseball's second-lowest BABIP on fly balls, ahead of only the Blue Jays. That's not entirely luck; at least in part, it's Safeco. The Mariners have a .097 fly ball BABIP because Safeco kills all of the fly balls, particularly the ones hit to LF. Last year, the Mariners had the sixth-lowest FB-BABIP in baseball at .126. Since 2000, their first full season in Safeco, the Mariners have the eighth-lowest fly ball BABIP in baseball.

Fly balls, in Safeco, are bad. Especially for right-handed hitters! To prove this, here's a look at some individual hitters:

Miguel Olivo FB-BABIP: .025

Justin Smoak FB-BABIP: .067

Jesus Montero FB-BABIP: .077

Alex Liddi FB-BABIP: .083

Chone Figgins FB-BABIP: .086

What do these hitters all have in common? They are all (at least sometimes) right handed! They also are five of the seven lowest FB-BABIPs on the team, the other two being Ichiro and Michael Saunders. In fact, the only right-handed Mariner who has hit more than twenty fly balls, more than 10% of which have dropped, is Brendan Ryan. Guti and Casper both have good looking FB-BABIPs, but Guti just came back on this road trip and Casper has only hit 18 fly balls.

Anyways, what makes all this worse is that the right-handed Mariners are almost exclusively fly-ball hitters. Smoak, Liddi, Olivo, Wells, Ryan, and Figgins are six of the Mariners' eight fly-ball-hittingest hitters. It's also no coincidence that Kyle Seager, the Mariners' most extreme fly ball hitter, is faring so poorly in Safeco, though he's at least left handed (and getting BABIP screwed with only .216 in the Safe). Right-handed flyball hitters do not survive in Safeco unless they are legends--hell, even Beltre couldn't do it. A Mariners roster where all the righties are fly ballers is a recipe for disaster.

Now, again, unfortunately, we're held back by a lack of data in that I haven't been able to find Home/Away splits that are also batted ball splits on Fangraphs. What would be expected, and what has been talked about by a whole bunch of Mariners, would be a much lower FB-BABIP at Safeco than on the road. This would explain why the Mariners' recently high home fly ball rates have been so disastrous for them as a team, and also incidentally why Jason Vargas runs such a low BABIP at home and is thus more valuable to the Mariners than to anyone else.

I'm not sure why exactly the Mariners would hit more fly balls in Safeco than elsewhere. It may just be a factor of the pitchers they've been facing; this has been a small sample size and historically speaking the Safeco Mariners have been known for fairly low fly ball rates. My bet's that they're just seeing a little random variation that's causing them to face more fly ball pitchers at home. It's thus not really fixable in terms of improvements that can be made by these particular hitters.

Now, if the question is "how to build a successful Mariners team", there is indeed something that can be done. The obvious solution is to pursue only line drive / ground ball hitters, especially when acquiring righties. The problem is that this cuts the Mariners off from a whole class of premier players. Given that most players are right handed and most power hitters are fly ball hitters, the Mariners are in a bit of a pickle. The problem here appears to be a problem caused in part by the park, i.e. the fact that Safeco is where fly balls go to die. Sadly, this isn't a problem that can be fixed just by moving the fences in. Maybe a few of those floating fly balls drift out of the park instead of into gloves, but not many. No, this is a climate problem, and it won't be solved by moving in the fences alone. So what could the Mariners do? Well, for one thing, they could just leave the damn roof closed. Close it whenever you get an excuse and Safeco is suddenly a much hitter-friendlier land, since the wind dies and the "floating" effect goes with it. A less kludge-y but admittedly far more extreme solution would be to rotate the entire park ninety degrees, which would actually eliminate that crosswind and thus the fly ball cushioning effect. That would probably be more effective than just moving in the fences, actually, but of course it's totally unfeasible. Unfortunately, at this point the Mariners may have to move the fences in, if only as a PR move to attract free agent hitters and make their own hitters feel less intimidated.

THEORY THREE: SHEER DUMB LUCK

Sorry, but any time you run a .251 BABIP at home there's got to be some luck involved. Perhaps the Mariners have run into harder pitchers, or perhaps their hits are just finding gloves at an unusually high rate. Fact of the matter is, they're due for at least some improvement at home, just based on BABIP alone. The ISO split too is strange; though Safeco has historically diminished the Mariners' ISO by .010, I find it hard to believe that it's doing .060 of damage this year--especially since last year the damage was only .006.

So which theory is it? Well, honestly, a little bit of each. At home, the Mariners are not putting many balls in play, for whatever reason (be it luck or, more likely in my opinion, a Safeco-inspired change in approach). When they do put balls in play, an unusually high number of those balls have been fly balls for whatever reason (be it a Safeco-inspired change in approach or, more likely in my opinion, luck), which just happen to be the type of ball you absolutely do not want to hit in Safeco. And when they don't hit fly ball outs, they still get unlucky, since both BABIP and ISO are screwing them over at home. Essentially, it seems to me that if the problem is "the Mariners cannot hit in Safeco", the causes of this problem are likely the three described above. The causes of those are things that I'd need a little more data to figure out, but we can come up with some likely hypotheses, such as "they're taking more pitches at home" or "the wind floats fly balls".

So what can the Mariners do? Well, honestly, not much. Safeco is seriously damaging the team's offensive output, but this year's damage has been so ridiculous that most of it just has to be luck. There's little the team can do but sit and wait. However, if indeed they are telling their hitters to use a different approach at home, they need to stop telling them that. Everyone on the team should be using their road approach. Unless they are Brendan Ryan or Casper Wells, who have been much better at home actually, and should use their home approaches.

So, should they move in the fences? Well, yes and no. No, because closing the roof would probably work better and definitely be much cheaper, because theoretically the location of the fences shouldn't affect balance for one team or another, and because Jason Vargas. Yes, because the difficulty of acquiring right-handed free agent hitters in Safeco is hurting the team and because the distance of the fences may be scaring young hitters into changing their approaches for the worse. Then again, if they're going to move in the fences I say just go the full monty and spin the park 90 degrees. Hey, if we're going to be absurd...

Individual Players

In which I do a rundown of several players with particularly interesting Safeco/Away splits, discussing what each one's problem is and how he could stand to improve.

Dustin Ackley

On the road, Dustin Ackley is a 105 wRC+ hitter, making him a highly valuable player even considering the excellent current crop of major league second basemen. At home, his wRC+ is a miserable 58. This is despite his running a higher BABIP at home than on the road (.308 vs .305). He's striking out more at home, 25.2% vs 19.1%, and walking slightly less, 9.2% vs 10.6%. While the walk split was reversed in 2011, Ackley actually had the same strikeout split in 2011 as he's having this year: 25.1% at home, 17.1% on the road. This might be indicative of a change in approach, but again, we have no PitchFX Home/Away splits, so we can't really investigate that further. The interesting thing is that last year Ackley was helped by a .361 road BABIP, which he has failed to duplicate this year. The other interesting thing--and this, I believe, is the real root of the problem--is his ISO.

Last year, Dustin Ackley had no home/road ISO split. ISO was .145 in both places, higher in his amazing June/July and successively declining from there. This year, he has a huge, massive ISO split: .131 on the road, a little lower than last year but still OK, and a stupidly low .037 within Safeco. Can you remember the last time Dustin Ackley hit a double at Safeco Field? It was April 19th, more than two months ago. A triple? May 6th. A home run? August 30th--of last year.

Safeco suppresses offense, but it doesn't kill ISO like that, as we saw above. Now, his recent struggles haven't helped any: Ackley's ISO is a disastrous .026 in July after a May that was actually pretty solid. This is of course due to regress back towards his career norms. On the one hand, I'd like to call this luck because Ackley didn't show any home/away ISO splits last year. On the other hand, I'm hesitant to do so since his overall BABIP is right about where it should be. For some reason or other, Ackley has turned into a singles hitter in Safeco, particularly this last month. Could it be his approach? Does it have more to do with his recent struggles? Or is it just strange luck? I'm not entirely sure.

Kyle Seager

Seager has stupid, absurd, insane home/road splits. He's got an 11.8% walk rate in Safeco, better even than Ackley--to go with a 21.8% K rate, worse than Olivo. Conversely, on the road Kyle walks at just 5% and strikes out at just 16%. Oh, and that Safeco ISO gap that I mentioned earlier, the one that's killing Ackley? Seager, too, is in its clutches: .125 ISO at home versus a .250 ISO on the road. Literally double the ISO. So what's up? Is Seager literally an entirely different player on the road? Well, not exactly...

Seager's problem is that he is an extreme, extreme fly ball hitter. He's actually hitting more line drives in Safeco than he is on the road, but the extra line drives are coming from his ground balls, not his fly balls, and the fly balls are the problem. As discussed in section 2 above, Safeco is where fly balls go to die. Unsurprisingly, Seager has a ridiculous home/road BABIP split. This explains the missing XBHs, too--it seems like Safeco is floating Seager's fly balls that would go for doubles on the road and turning them into outs at home. (Home/Away batted ball splits would be very nice here.)

Seager's saving grace is that he's left handed, so being a fly ball pull hitter isn't totally disastrous in Safeco. He would die completely in Petco. However, his batted ball profile is still a very poor match for his home park. Seager, unfortunately, would be considerably more valuable to another team than he is to the Mariners. He's also OPSing 1.438 when he pulls the ball, so any manager who does not run the shift literally every single time Kyle Seager comes up to bat is a moron.

Justin Smoak

Ah, Mr. Warning Track Power himself, the source of all the fence discussions. Smoak has been getting absolutely wasted within Safeco Field--if you thought Ackley's home/road splits were bad, check out Justin's home wRC+: 38. On the road, the figure is 100. This is what it means when people say a player is more valuable to one team than to another, right? Well, no. Smoak was actually considerably better in Safeco last season, sporting a 121 wRC+ there versus a 83 wRC+ on the road. In fact, Smoak's home/road splits last year are startlingly strange. His BABIP there was thirty points below his road BABIP, he walked 5% more at home than on the road, and he literally doubled his road ISO within Safeco (.113 versus .225.). No one does that, except apparently Justin Smoak.

Actually, 2011 Home Justin Smoak is a perfectly acceptable player. 121 wRC+ with a .256 BABIP? I'll take it.

The problem is that 2011 Justin Smoak was running an ISO split that no one can sustain. Ten of his fifteen home run were in Safeco. His HR/FB in Safeco was a way high 13.5%, double his road HR/FB, which doesn't happen when your home park is Safeco.

The good news is that Smoak fixed his road woes this season, bringing his road numbers in line with his seasonal averages from last year. The bad news is... you remember all of those road numbers that Smoak was doubling at home last year? This year he's doubling his home numbers on the road, or rather halving his road numbers at home. Which is bad. Now, of course, some of this is luck. His plate discipline statistics aren't appreciably different at home versus on the road, but his BABIP is an atrocious .188 in Safeco. His ISO, too, is in the cellar--only .086 at home is awful for a first baseman. However, I don't think that all of this is caused by luck. His batted ball profile in Safeco contains significantly more fly balls, and those are coming out of his line drives and grounders. Remember how Safeco is the place where fly balls go to die? Safeco is especially the place where fly balls go to die if you have middling power, like Smoak, and this turns up in his atrocious FB-BABIP. His HR/FB numbers at home are now firmly back beneath his road HR/FB, as they should be. But his Safeco struggles are indicative of genuine problems--lack of power and being a backspin flyball switch hitter in Safeco. Smoak is exactly the kind of guy that Safeco destroys, and it's destroying him this year.

Mike Carp

I close with Mike Carp, not because he's particularly important or because I'm out of other players to talk about, but because of the sheer weirdness of his Home/Away splits--and, really, his whole darn season. There are plenty of other unusual splits that bear discussion: Brendan Ryan is an acceptable hitter at home! John Jaso walks 20% of the time at home! Alex Liddi put balls in play in fewer than half of his at bats at home! But it is the sheer weirdness of Mike Carp's lines this season that merit attention.

Mike Carp has almost identical home and away wRC+es this season despite achieving those wRC+es through wildly different means. The only other constant was his abysmal, abysmal BABIP: .158 at home, .171 on the road. On the road, he pulled up his wOBA despite a godawful batting average by hitting for power. Mike Carp had nine hits on the road, and three were home runs. Two more were doubles. His .216 ISO gave him a ridiculous slugging percentage on the road, even though he wasn't walking. At home, Carp magically transmogrified into a totally different hitter, pulling up his wOBA despite a godawful batting average by walking a whopping 22% of the time.

More weirdness! At home, Mike Carp did not hit a single infield fly ball. On the road, almost a quarter of his fly balls did not leave the infield. Mike Carp had the second-highest fly ball BABIP on the team, behind Kyle Seager, which is .190. That is higher than his actual BABIP. Mike Carp's wRC+ on ground balls was -90, because his BABIP on ground balls was .042.

I'm not going to try to explain why all of this happened to Mike Carp, because I honestly have no fucking idea. It's quite possibly the weirdest partial season line I have ever seen in my life. The reason I bring this up is to remind everyone that small samples, like the ones discussed in this post, often do freakishly weird things, and it's foolish to jump to hasty conclusions based on them. It would be foolish to disregard Mike Carp just because he ran a preposterously low BABIP. It would be foolish to disregard Justin Smoak just because he has yet to hit a double in Safeco Field. And it would be foolish to disregard the volcano you live near just because it hasn't erupted in a very long time. Who knows how long volcanic eruption rates take to stabilize? Not me! So the moral of the story is, fear volcanoes, and also Safeco Field is probably changing the approach of the Mariners' hitters while simultaneously eating all of their fly balls. So now you know.