Game 132, Yankees at Mariners
Matt Wisler/Tommy Milone vs. JA Happ, 7:10pm
Ahhh, yesterday’s win felt great. It’s nice that even in a lost season, the team’s capable of making you pump your fist or high five strangers. Walking around downtown Seattle yesterday, it’s hard to overstate just how visible the Canadian invasion of the city and T-Mobile is. People in Jays uniforms and shirts were *everywhere* and they’re heading home with a series loss. This weird annual tradition must be great for a lot of people, from Seattle area hoteliers to the Jays themselves. It has to be odd for the M’s players, and I could see it having some very small impact on their play. But for the ultra-competitive Marco Gonzales, I think it may have backfired. Gonzo looked motivated and pitched one of his best games in an M’s uniform yesterday.
One of the recurring themes on the old blog this year has been what to make of Marco Gonzales. This is where we typically dig in to some statistics or trends to help put his performance in context, but this year, that’s just adding to the confusion. What do I mean? By FIP, which looks only at strikeouts/walks/home runs, Marco’s had a great season – he’s essentially followed up on last year’s breakout by surpassing last year’s WAR total, standing at 3.5 fWAR today. Baseball Reference’s WAR formula starts with RA/9, and attempts to correct for park and team defense. Those last two adjustments are pretty important, and they pull Marco up to 1.9 WAR, or essentially right at league average. Baseball Prospectus’ DRA formula is based on linear weights, or counting up all of the hits/walks/outs etc. a pitcher’s given up, and then adjusting and regressing for everything from park to umpire to batter to weather. When they do all of THAT, Marco comes out at a DRA of 6.05, good for a WAR(P) of *negative* 0.4. Three big WAR formulations, and we’ve got a replacement-level scrub, a league-average hurler, and an all-star.
We can just forget all about this WAR stuff, right? What about plain-old ERA? By that measure, Marco’s at 4.17. But if we count unearned runs and measure his RA/9, it’s nearly 5, at 4.94. However you plan to tackle a question like, “How good has Marco Gonzales been this year?” you’re confronted with some challenges like what to include and what to exclude. FIP largely excludes balls in play, under the rightish but not 100% accurate idea that pitchers don’t have a lot of control over how often balls in play are converted into outs. RA/9 WAR like BBREF’s includes everything, from unearned runs to bad sequencing luck. There’s no real right or wrong answer, and in 99% of the cases, they tend to arrive at the same basic answer. But not so with Marco.
This isn’t the first time this has happened, and it won’t be the last. A few years ago, the great Sam Miller had an article at ESPN talking about how the three systems were hopelessly split on Arizona’s Robbie Ray, with DRA loving him, RA/9 hating him, with FIP somewhere in the middle.* But even in that case, everyone agreed he’d had SOME positive value. It was just an argument over how much. Here, we can’t even agree on that. What’s going on? While Marco doesn’t have a great K:BB ratio, he’s been stingy with home runs, and since they’re weighted so heavily in FIP, he comes out looking quite good. Essentially, Marco gives up a heck of a lot of base hits, but hasn’t allowed a ton of runs – and those runs that HAVE scored on him have been charged as unearned more frequently than normal. All of those base hits are getting added up in DRA, and they push his total-average-against or wOBA-against higher. That accounts for a bunch of his poor DRA, but certainly not all of it.
Marco’s TAv-against is .262, ranking 56th out of 79 qualified pitchers (#79 is Yusei Kikuchi). It’s that context stuff that’s dragging him down; DRA expects, given the kind of batters he’s seen and the places he’s pitched that just about anyone could stumble into a TAv-against of .262. Is that realistic? I don’t really know, but I looked at 20 pitchers (including Marco) with similar TAv figures, ranging from .255 to .267. As a whole, the group has an ERA of 4.40, a FIP of 4.33, and a DRA of 4.83. Because we’re only looking at *qualified* pitchers, there’s a pretty big survivorship bias, as all of these pitchers have been successful enough to get a lot of starts (it’s a low bar for some teams, but it’s a bar). Hence, I’m not too shocked to see a lower ERA, just like I’m not surprised to see a low FIP. Beyond that, it’s a wildly disparate group, from strikeout maven Matthew Boyd of Detroit hanging out with strikeout-allergic Brett Anderson. Robbie Ray’s in the group, as is Homer Bailey and Dylan Bundy. Their RA/9s range from 4.29 to 5.92, and their DRAs range from 3.4 all the way up to 6.8. I’d love to know more about what’s dragging Marco and Ivan Nova down (and pushing Joe Musgrove up), but the average RA/9 in this group is 4.84, essentially dead on the average DRA.
Today, the M’s face another pitcher that DRA hates, ex-Mariner JA Happ. Happ’s creeping up on negative 2 WAR by BP’s formulation, the result of a DRA over 7. For the record, FIP isn’t terribly impressed either, giving him a 5.69, good for 0.7 fWAR. The problem here is home runs, as Happ’s allowing well over 2 per 9. His 31 HRs allowed tie him for 4th-most in the game along with Yusei Kikuchi, and just three back of the leader, Mike Leake (it’s been a bad year for M’s pitchers). The bulk of the damage has come against his four-seam fastball. In his career, he’s given up a .413 SLG%-against on the pitch, but this year, it’s way up to .627, thanks to 23 HRs-allowed on that one single pitch. The pitch has had strongly positive pitch-type run values for several years, but they’ve all tipped negative this year. His velocity’s down, I suppose, compared to his great years in Toronto, but by less than 1 MPH. The pitch is a tad straighter, but there’s no obvious red flag in movement or velo. His park plays a role, I’m sure, and then there’s the baseball, but it’s noticeable that he wasn’t destroyed in 2017, the last time the league played with a new, different, batter-friendly ball. I’m a bit surprised the Yankees – a club that eschewed four-seamers in favor of breaking balls in recent years – hasn’t had him rein in his four-seam usage. As you’d expect, righties are doing the most damage because righties see more four-seamers. He throws more sinkers to lefties, and thus has extreme platoon batted-ball splits – lefties hit the ball on the ground (but if they get it in the air, it’s going for a dinger), while righties hit a lot of fly balls. This just isn’t the kind of year where you can give up a ton of fly balls and succeed.
1: Lopes, LF
2: Crawford, SS
3: Nola, 1B
4: Seager, 3B
5: Murphy, DH
6: Narvaez, C
7: Fraley, CF
8: Moore, 2B
9: Broxton, RF
SP: Wisler/Milone
Tacoma’s home schedule is over. They dropped their final game 2-0 to Las Vegas on a 2R-HR by ex-A’s OF Nick Martini. I say “ex-As” because he was DFA’d today to make room for Seth Brown, who was called up. Wonder if Jerry Dipoto’s already asking about acquiring him…
Tacoma’s in Fresno tonight, but the big prospect interest in the system is Justin Dunn starting for Arkansas against Corpus Christi.
* The debate’s kind of moot now, as DRA was reformulated subsequent to the article, and rescinded all of the praise it had heaped upon Mr. Ray, changing his 2.95 DRA to 4.83.
Comments
Leave a Reply
You must be logged in to post a comment.