Game 151, Blue Jays at Mariners
Hisashi Iwakuma vs. JA Happ, 7:10pm
Last night’s loss was a tough one. Between Marco Estrada’s brilliance and a stadium full of Canadians cheering him on, it didn’t feel like a critical M’s game. The M’s continue to struggle against pitchers like Estrada and Mike Fiers, which sucks, as JA Happ’s rising four-seamer could presage more harmless fly outs. To be fair to the M’s, they hit the ball somewhat hard later in the game – Cano’s bases-loaded warning track drive was the most frustrating example. The loss knocked the M’s playoff odds down to 15+%, down about 7 percentage points from yesterday. The Blue Jays not only got a big win, but saw the Orioles lose as well, so their odds crossed the 70% mark, up about 9 percentage points on the day.
Tonight they face 2015 Mariner JA Happ, who I think doesn’t get enough attention for how odd he’s been over the last 3 years or so. For several years, he was a garden variety 5th starter, who walked too many and gave up too many HRs thanks to a deadly combination of both below-average velocity and “stuff” AND below-average control and command. Things started improving, seemingly out of nowhere, in 2014, when Happ was 31. He began throwing more strikes and added 2mph on his FB. His 90-91mph rising four-seamer and sinker were suddenly 92-93+, and while he still had some HR issues, he at least offered a bit of potential. That was enough for Seattle, who traded Mike Saunders for him despite his looming free agency. At the time, I wasn’t thrilled with the trade, but thought that Happ could adapt to his new surroundings by ditching his sinker and actively courting fly balls. He didn’t really listen, and while his GB% crept up slightly, he gave up a ton of base hits and thus too many runs in Seattle. He’d be great at times, but inconsistency doomed him, and he was shipped to Pittsburgh for unheralded SP prospect Adrian Sampson.
Under Ray Searage, Happ completely abandoned the sinker, and started firing four-seamers over 70% of the time. This sounds like a bigger change than it was; he’d often thrown a ton of fastballs, going all the way back to his days in Philadelphia. And he’d typically thrown them up in the zone – batters swung at more of them there, and the location plus his rising action allowed Happ to miss more bats than he otherwise would, given his so-so velocity. But in Pittsburgh, he was flat-out dominant. His K rate spiked to over 26% and he essentially stopped walking people at all. Curiously, though, his GB% didn’t really move. No one cared, because no one could actually hit him, and he looked like a completely different pitcher.
The Jays thought that was enticing enough to extend him a 3-year, $36m contract in a move that left a lot of people shaking their heads. Thus far, Happ’s been worth it. The weird thing is that he’s been successful by abandoning the approach that worked in Pittsburgh. From the beginning of the year, he’s been throwing more sinkers. Not just more than he threw in Pittsburgh (none) or Seattle (15%), but far more than he’s ever thrown – nearly 30% of his pitches. Eno Sarris notes that Happ’s sinker has a lot more drop than his four-seam, and that’s true – but it’s also not new. That was true very late in 2014, and the gap was even bigger in Seattle than it is now.
So with this new-fangled approach, he’d have to have a different batted ball profile, right? Well, no. His GB% now is essentially unchanged from where it was last year. Even comparing just his Pittsburgh stats (zero sinkers) to this year’s Toronto numbers (lotsa sinkers) shows a delta of less than 2 percentage points. The pitch is put in play often, goes for grounders often, and he’s throwing more of it…but his batted ball profile is stuck at around 40% grounders. The key here is that the approach is doing something to his *four-seamer*. His whiff/swing ratio on his four-seam is over 26%, way higher than in the past, and one of the best in baseball. Not only that, but because batters are more likely to swing at it when it’s up (or out of) in the zone, when they DO put it in play, it’s most likely a fly ball or pop-up. The increased grounders he’s getting with the sinkers are offset by the decrease he’s inducing with his four-seam.
So, are fly balls good? When he was moving to Seattle, I thought they would be, and indeed, Happ had pretty good success on fly balls last year. Like everything else, this was magnified in his time in Pittsburgh, where an ultra-low HR/FB ratio helped keep his ERA and FIP gaudy. He’s got a very low BABIP, and again, he’s been pretty successful on fly balls this year, though it’s worth noting that his HR/FB luck ran out, and he’s giving up MORE dingers than he did in Seattle. What’s actually driving that low BABIP (and thus a big gap between ERA and FIP) is ground balls. In 2015, batters slugged .305 on GBs against him (data from statcast, so it doesn’t match Fangraphs/BBREF exactly). This year, that figure is .217. BBREF has it as .207, a figure that gives him an sOPS+ of 58 (relative to the league average of 100). Why? Statcast shows a slight change in his modal or mean launch angle – here’s 2015’s and here’s 2016’s – you’ll see a bigger grouping of GBs around -10 degrees in 2016, and those grounders should theoretically be easier to convert into outs. But what about playing in front of an infield featuring Josh Donaldson, Troy Tulowitzki, Darwin Barney, Devon Travis, etc.? Toronto’s IF defense rates very high by advanced and regular metrics, and that’s probably driving some of the improvement as well.
1: Heredia, LF
2: Gutierrez, RF
3: Cano, 2B
4: Cruz, DH
5: Seager, 3B
6: Lee, 1B
7: Martin, CF
8: Iannetta, C
9: O’Malley, SS
Big changes in the M’s minor league system, as the Jackson Generals signed a two-year affiliation agreement with the Arizona Diamondbacks. That’s pushed the M’s out of the Southern League and back to the Texas League, where they had a club from 2001-2006. Back then, they were with San Antonio, a club that featured a fair to pitcher-friendly run environment. Now, the M’s affiliate will be the Arkansas Travelers, one of the toughest parks for home runs in all of AA. Statcorner’s park factors give it an almost comical HR factor of 57 for right-handed bats. Arkansas was an Angels affiliate recently.
With Bakersfield sadly contracted, the M’s needed a high A affiliate, too. I’d thought that they may need to go to the Carolina league, who’ll grow by the two teams the Cal League shed, but instead they’ve inked a deal with the Modesto Nuts. The ex-A’s affiliate and more recently Colorado Rockies affiliate, is another park that really suppresses HR, though of course the Cal League as a whole sees more HRs than the Texas League. By Statcorner, their HR factors are 46 for LHB and 61 for RHBs. That sounds insane, but I assume the league values are inflated by High Desert, Lancaster and some of the other launching pads. It’ll certainly make a change from Sam Lynn ballpark, which featured the shortest CF distance in all of professional baseball, and thus had HR factors of 120/142 (LHB/RHB).
Speaking of the minors, it’s impossible to see this Toronto line-up and not think of the 2010 Tacoma Rainiers. Sure, I think of the 2010 Rainiers when I look at most things, but Toronto has three key offensive players from that PCL-winning team: Justin Smoak, Michael Saunders, and Ezequiel Carrera. It’s tough to predict MLB success, and so much is down to opportunity, blah blah, but I will never stop being stunned that Ezequiel Carrera is a better 2016 big leaguer than Smoak, Ackley, Matt Mangini, Mike Carp, etc. Saunders has been better, of course, and we’re not even getting into guys like Michael Pineda or Shawn Kelley, but Carrera was one of the worst players I saw that year. He hit 0 HRs in 64 games and generally looked like a slap-hitting 4th OF for a AAA team. The M’s traded him in June and I thought we’d never hear from him again. He’s not been good, mind you, but he’s been more valuable than Smoak, and with Ackley hurt and so many other prospects of 2010 now out of baseball, Carrera’s pretty far up the list of most successful alumni. Baseball is baffling.