M’s Pitchers Stopped Giving Up HRs All the Time: Approach Change, Schedule, or Luck?
If you’ve read this blog for any length of time, you’re familiar with me rambling on at length about home runs and their increasing importance for run scoring. Since the recent HR explosion began (now confirmed to be largely due to a change in the ball), the M’s in particular have suffered as they’ve given up a disporportionate number of dingers, which has dwarfed some real improvement in other areas. That is, K:BB ratio is wonderful, but it’s less wonderful if any improvement in controlling the zone is accompanied by a flurry of home runs, and unfortunately, that’s where the M’s found themselves.
In April, things looked much the same: the M’s had the second-highest HR/9 behind Cincinnati. Their record wasn’t too bad, but the dingers and some dinger-related blowouts meant that their run differential wasn’t great. Their offense hit enough to overcome a sub-par rotation, at least when James Paxton wasn’t pitching. How has May looked? Really, really different. The M’s allowed *the fewest HRs* in all of baseball, and that’s enabled them to post the highest fWAR in the game, just ahead of the historically-great Astros (technically, the Astros were better on a rate basis, but still…Holy $&%^). The M’s had a weakness, and not only has that issue stopped being a problem, it’s a massive, out-of-nowhere strength. What the hell happened here?
There are, of course, a number of possibilities, and I’ve listed a few in the title. I don’t have a definitive answer, so don’t go searching for one. When dealing with month splits, you’re not going to have any kind of clear-cut, verifiable conclusions: there’s not enough data. That said, we can walk through a couple of possibilities and see what we see.
1: The M’s Changed Their Approach
To look into this, I got each team’s pitch mix as measured by Pitch Info from Fangraphs. I looked at 2017, as well. There are a number of ways to see who’s throwing a lot of fastballs, as there are multiple fastball types – four-seamers, sinkers, and then cutters/splitters. For a lot of analysis I do, I typically limit things to four-seam and sinkers (and, as applicable, the old category of two-seamers). But here, it might be valuable to look at cutters as well. While some cutters are definitely more breaking ball-like, others really are just versions of a fastball, as someone like Nick Vincent (or Kenley Jansen) shows. So, we can do it both ways.
What you see is that there are a few teams that are especially reliant on one type of fastball or another, and then there are teams that have really moved away from fastballs and have shifted towards breaking balls and change-ups. The Pirates are a good example of the former, while the Yankees attracted a lot of attention as the exemplar of the latter approach. In 2017, the M’s were really middle-of-the-road: they ranked 15th in four-seam+sinker usage, and 10th if you include cutters. They used a four-seam or sinker about 56% of the time. THat’s dramatically more than the Yankees’ 41%, but *everyone* was dramatically more than that. It was well shy of the Pirates’ 63% usage, which was driven by the the highest rate of sinkers in the game. For four-seamers, the Rockies led the way, something I mentioned when the M’s faced them last year – they thumbed their nose at the HR explosion by throwing a ton of four-seam fastballs *in Coors* and posted surprisingly low HRs-allowed numbers. The M’s bet big on their OF defense last year, and thus prized pitchers who threw rising four-seamers (and would’ve thrown even more if Drew Smyly had been healthy), so they ranked 6th in four-seamers, and quite low in sinkers. They were middle of the pack in cutters, but there are quite a bit fewer of those.
In 2018, their pitch mix has changed markedly. Their four-seamer usage has plunged from 6th down to 27th, and while their sinker rate has increased, it hasn’t offset the decline in four-seamers. Thus, if we measure *just* four-seamers and sinkers, the M’s go from a middle-of-the-pack team to the 25th-ranked team for FB usage. But a fairly large chunk of those missing/eschewed four-seamers have become cutters (thanks Mike Leake and Marco Gonzales), so now the philosophical discussion about what bucket to put cutters in becomes pretty relevant. If you put them in the fastball bin, the M’s overall FB usage rate hasn’t changed *at all*. It was at 62.6% in 2017, and it’s at 62.9% now. If you put them in the breaking ball bin, the story’s pretty clearly one of the M’s ditching FBs in favor of bendy pitches – even if that story’s complicated by the fact that cutters are perhaps the least bendy of pitches. It’s also relevant to us in that the pitch classified as the cutter is one of James Paxton’s favored pitches. HE uses it as a breaking ball, whereas Leake/Gonzales use it more as a fastball. In any event, it seems relevant that the M’s rotation is full of pitchers who throw one, even if they exemplify the fact that “cutter” is something of a catch-all term.
Still, isn’t there some kind of impact from turning four-seamers into cutters? Yes, there certainly should be. The league has put up a .320 wOBA on cutters, compared to a .339 mark against four-seamers and .340 against sinkers. They typically have some platoon split issues, and that’s true here, but there’s some evidence they’re harder to drive. The average exit velocity off of a cutter is a full 2 MPH lower than against sinkers, and a touch more than that against four-seamers. A lower percentage of balls in play against cutters were classified as “barrels” by MLB than against four-seamers (though here it’s worth noting that sinkers went as barrels even less often). So, platoon splits be damned, there’s some evidence that a shift like this should result in lower HRs. And that’s what we’ve seen.
Still, even just limiting things to four-seamers/sinkers, the M’s haven’t had the most dramatic change in approach – that title probably goes to the Rockies, who’ve cut FA+SI usage by over 11 percentage points. Like the M’s, they’re throwing more cutters to fill the gap. The Cardinals and Tigers are neck and neck as teams who’ve shifted away from fastballs AND cutters, while the Twins, Cubs, and especially Braves have moved the other direction. If you think of cutters more as breaking balls, the M’s have closed the bendy gap with the Yankees, just without the wave of sliders: the M’s throw a decent number of curves as well as cutters and change-ups.
But to assess if this is responsible for the M’s change in HRs-allowed, we need to see if they’ve changed their approach within 2018, right? Again, small numbers, but there’s some evidence that they have. In April, they threw 51% FA+SIs, while in May that’s dropped a bit to just under 50. That’s obviously not enough to account for the sharp drop in HRs, so the bigger story is lower HRs per fastball. Still, it’s notable that the M’s are throwing more curves than ever before.
Verdict: It’s probably not the change in FB mix or pitch mix overall, but this bears watching. The M’s thought they could throw a bunch of four-seamers in Safeco, got burned, and now aren’t daring teams to elevate four-seamers anymore. I like that.
2: The M’s played teams that can’t hit. Er, can’t hit HRs. Or no, just can’t hit.
One of the logical consequences of a league with superteams AND an unbalanced schedule is that you should see a number of runs where teams give up runs by the bushel and others where they post higher fWAR than the Astros over the course of a month. Is this what’s going on?
Again, it’s going to be impossible to answer, but let’s try something simple. The M’s have faced several teams in May: the A’s (6 times), the Angels (3X), the Tigers (7X), the Twins (4X) the Jays (3X) and Rangers (3X). Those teams have homered varying amounts, but if we simply divide HRs per game, we get a rough and dirty expected number of dingers. If the A’s homered their customary amount (1.22 per game), they’d be expected to hit 7.33 against the M’s in 6 games. They actually hit 4. Doing this for the entire schedule, we get an expected number of HRs allowed of 28.35 versus an actual number of just 18. Did luck play a role? Sure. Did the Tigers/Twins being two of the worst power-hitting teams help? Sure. But even with that, you’d have expected the M’s to give up a few more.
The problem with this theory is that the M’s haven’t ONLY played the Twins/Tigers, as much as I’d like that. The A’s/Angels really are good power hitting teams, and the M’s held those teams in check. Things will change in June as the M’s face a much tougher slate of teams, but it surprised ME to see how many games the M’s had against some teams that are decent at hitting dingers during this brilliant run of pitching form.
Verdict: They played some bad hitting teams, but they played some pretty good ones, too. Not buying this.
3: Maybe this is just luck, right?
By statcast’s wOBA-xwOBA measure – actual results versus predicted results using launch angle and exit speed – the M’s have been pretty fortunate in May, with a .317 actual wOBA allowed versus an expected .372. The M’s defense is absolutely a part of this story, as their BABIP-allowed plunged this month. Given what we’ve seen in recent games, is it possible that Safeco’s playing more as a pitcher’s part again? Er, no – the M’s have been “luckier” this month in road games. Paxton’s no-no in Toronto’s a perfect example where some very hard hit balls were reeled in by M’s defenders in defiance of your nerdy “hit percentage” values.
Still, even if you think xwOBA measures luck (as opposed to something else), the real story is that the M’s actual and expected wOBA-allowed dropped *so bloody much* compared to April. The M’s were nearly as “lucky” by the same measure that month, it’s just that both numbers were shifted upwards.
Verdict: Yes and no. Yes, the M’s have been fortunate this month, but they seem to be pitching better, too. Their K-BB% is better, driven by a decline in walks.
Things will get tougher, but for now it’s worth noting that the team’s rotation has stepped up at just the right time, staving off a decline in offense. They’ve beaten the teams they should beat, sure, but they’ve fared better against some relatively tough teams as well. I still don’t know how, as even the shift away from four-seamers isn’t enough to explain this. Leaguewide, the HR rate has grown with the temperature from April to May, but it’s tumbled for the M’s. Long may this continue; I’d love to think that the M’s really figured something out, even if I’m still not sure what that thing is.
Comments
One Response to “M’s Pitchers Stopped Giving Up HRs All the Time: Approach Change, Schedule, or Luck?”
Leave a Reply
You must be logged in to post a comment.
Felix and Rzepczynski are the two pitchers on the staff showing no signs of changing for the better. At what point do you shake things up?
If it were my call, and Rzepczynski couldn’t be moved, I would DFA him and move Felix to the bullpen.