Will There Be Enough for Martin/Dyson to Do?
So, with the trade of Mallex Smith, the M’s are back down to 2 or 3 CFs, with Mitch Haniger – a competent CF himself – playing RF. There’s nothing wrong with defense first OFs, and in general, I subscribe to the idea that a run saved is essentially equal to a run scored. As M’s fans, we’ve seen an object lesson that putting two all-world defenders next to each other in the OF doesn’t reduce either one’s productivity: Mike Cameron in CF and Ichiro! in RF remains the single best OF duo I’ve ever seen, and the M’s pitching staff reaped the benefit.
The idea that elite defenders don’t necessarily take plays away from nearby defenders is a longstanding one, and it’s critical to Dave Cameron’s view of the Dyson/Karns trade. Quoth my erstwhile boss, “Now, though, there’s reason to think the 2017 Mariners might have the best outfield defense in baseball, or at least be in the conversation. Dyson, of course, was part of the Royals great wall of defense, and now he’s going to be roaming the fairly vast left field in Safeco.” He points out that the Royals OF racked up 135 runs above average by UZR and 121 by DRS in the past three years, while the M’s were solidly in negative territory. It’s not as simple as flipping the latter numbers for the former, so, really: how many runs are out there for the M’s to gain?
This is, sadly, a really difficult nut to crack. There are a number of different ways to look at this, and fortunately, a number of data sources to pore over. All of them have strengths and weaknesses, and ultimately, there’s no silver bullet here; there’s no equation that’ll spit out the runs the new-look M’s OF *will* save in the future. Futures are slippery like that. But we can try and examine the various factors that influence fly balls and figure out how the 2017 M’s might differ from their OFs of the past few years.
What might influence the number of balls an OF sees? There’s the total number of balls in play their pitchers give up; of those, the percentage that are hit in the air; how hard/where they’re hit; the physical size of the OF they’re attempting to cover; and a bunch of atmospheric/weather-related factors that are sadly going to be well outside the scope of this. Just from this partial list, you can see a number of reasons why the M’s of 2017 might be in a slightly different space with regard to OF chances than the Cameron/Ichiro!/Winn group of 2002 or so. The strikeout rate league-wide continues to grow, meaning there are balls in play, and the M’s have helpfully cut down their LF area which, as we’ve already seen, has had the effect of reducing the numbers of doubles and triples that batters hit. Still, a reduction in the number of chances still leaves plenty of room for the M’s to improve, right? The M’s trailed the Royals in OF UZR last year by 55 runs.
Let’s start by looking at total OF chances, as seen on Baseball-reference.com. The 2001 M’s saw 1,228 chances, 2nd most in the league. The American league average back then was 1,125, and the AL’s K% was 16.5%. Last year, the M’s saw 1,061 chances, a bit more than the league average of 1,045 – in related news, the AL K% was up to 20.9%. Taking a 3-year rolling average, the M’s have lost over 200 chances per year from 2003 to 2016. Obviously, the rise of strikeouts is a big contributing factor, as we can see from the fact that the league average itself has dropped by 62 chances per year. But the other part of this is Safeco’s dimension. From 2003-2012, the M’s averaged 1,188 chances/year and their average rank in the AL was 4.67. Since then, the M’s averaged 1,042 chances and have never topped 1,100; their average rank in the (now larger) AL is just 11th. Again, the confounding factor here is that the M’s have been almost intentionally bad defensively since the walls were brought in, so it’s tough to pull these various threads apart. However, it’s probably not a coincidence that the number of chances drops precipitously right when the dimensions changed, and that it’s pushed Seattle from a park that saw many *more* chances than the average park to one that’s seen fewer than average.
But that’s just chances – how about a broad measure that gets at converting chances into outs, like defensive efficiency? Baseball Prospectus tracks defensive efficiency, and breaks it up between fly balls and ground balls. Here, the trend is the opposite, with teams generally getting better and better at turning fly balls into outs. In 2003 (the first year they have good data), the M’s fly ball efficiency (the percentage of fly balls they turned into outs) was .876, good for best in the AL. Last year, the M’s ranked 6th in the league…but with a .906 mark. What’s going on here? The culprit appears to be a general trend towards classifying more balls in play as line drives. “Fliners” that could be called either fly balls or line drives were once largely grouped in with fly balls, are now called line-drives. That’s going to make it really, really hard to look at these numbers over time, but we can try, and we can look at other teams within the same year. So, the percentage of balls in play has dropped by 10 percentage points over time, and as a result, those easier FBs remaining turn into outs more and more often. In fact, the M’s and Royals were nearly identical in 2016 – the Royals FB DE was .912, just one spot ahead of the M’s in the AL. Again, looking at 3 year averages, the clear trend is towards a lower percentage of fly balls, but the biggest single drop in the 3-year rolling average for Seattle is the first year that the average was made up of post-dimension change seasons. By *this* measure, which is super high-level and not attempting to account for the difficulty of chances, the M’s and Royals were more or less equals in turning FBs into outs last year. Interestingly, the M’s have been consistently *good* in this measure since 2013, despite being just atrocious by UZR/DRS. In 2014, for example, the M’s had a .919 FB DE, 3rd best in the AL. They had a -4 run UZR that year. The M’s AL rank has been fair to middling since 2013, but the M’s 2013-2016 average is .9035, percentage points better than the…Kansas City Royals, who clock in at .9028.
That’s strange, but now that we’ve got Statcast data, why not look at that? This way, we can avoid classifications (liner or fly ball or pop-up?) issues and control for things like number of chances and even the speed of the ball. The downside is that it’s only available for 2016 and 2015, and 2015 data haven’t been playing nicely this past week. We shall press on: I took a look at all balls in play fielded by an OF, and combining all fly balls, line drives and pop-ups. The Mariners dealt with 1,691 balls in play that met these criteria last year – helpfully, that’s nearly identical to the Royals’ 1,695. Across MLB, the percentage of these balls that fell in for hits (BABIP) was .432. The M’s bested that mark, posting a .418 mark, not quite as good as the Royals’ .407. Here, the Royals really are an elite club, ranking 4th in MLB. But the M’s, despite their awful UZR, come in 8th, and 4th in the AL. By SLG%, the gap is even bigger, with the M’s 9th in the league, giving up around 40 bases more than the Royals, who rank 3rd best. That’s significant, but again, it’s not gigantic. In fact, the gap in total bases is a bit less than the total gap in UZR or DRS *runs* in 2016. And again, the M’s are clearly above average compared to other MLB teams. Of course, last year was…weird. If you add in HRs and don’t focus only on balls in play, Safeco gave up a much *higher* than average SLG% on contact. But what if the combination of marine layer (however attenuated or El Nino’d it was) and smaller park will result in a *consistently* high floor for OF defense? The M’s worst defense in recent memory was clearly the 2015 group, but just looking at total chances and BA compared to the league average, they don’t seem completely terrible.
I can pull my head out of a spreadsheet every once in a while, and I actually enjoy watching a ball game. I can say that by the eye test, the 2015 M’s were terrible, and that the Royals of recent vintage look incredible; Cain/Gordon/Dyson is about as good, maybe AS good, as Cameron/Ichiro/Winn. Moreover, even if there are fewer than there used to be, there’ll still be 1,000 or so balls in play for the OFs, and the M’s should endeavor to catch them. But the more I look at it, the more I think it’s probably worth trying to quantify exactly how many balls Dyson/Martin will get to that Martin/Smith+Heredia+Gamel didn’t/couldn’t. It’s more than zero, I’ll stipulate that. But I’m also not convinced it’s a gigantic number, and given the persistently low number of chances – and the correlation between chances and UZR – that the M’s will save 40-50-60 runs with this defensive enhancement. Does that matter? I don’t know, ultimately. The M’s can still tinker with that number, as they control how FB or GB-minded their pitching staff is. With Drew Smyly in the fold, they may yield more flys than in the past, but again, they can only improve by so much: they ranked 6th in baseball last year in FB%.
Still, I think it’s interesting that some teams pretty consistently rank near the top in FB% allowed. Examples include the Angels and Rays, who rank #1 and #2 in FB% given up over the past 4 years combined. As a result, their OFs have made hundreds more plays than the M’s group over that time. Part of that is the like of Kevin Kiermaier patrolling the OF, but a large part of it is the sheer number of chances they’ve had. Given this imbalance, and given that this imbalance looks a whole lot like a coherent strategy, I understand the Rays’ desire to stick Mallex Smith next to Kiermaier. I understand somewhat *less* the Angels trade for Andrelton Simmons, given the paucity of chances he’ll get. Simmons is still great, no matter what uniform he’s wearing. And Jarrod Dyson’s (almost) equally incredible. But like Simmons in Anaheim, I wonder if *some* of that transcendent skill will be wasted in his new home park.