The Changing Battle Between Batters and Pitchers
This is a strange post without real conclusions that was sparked by an unrelated quote from M’s manager Scott Servais. I’ve grabbed a whole bunch of data from Statcast, Pitch Fx and Fangraphs, and it tells the story of how much – and how *fast* – the game has changed over the past 10 years, and the last 3-4 years in particular. I think that’s interesting enough for a post in this remarkably slow hot stove season; I mean, what else are we going to do?
On January 4th, David Laurila had one of his characterically-great quote round-up posts, this one focused on how teams valued speed in today’s game. It featured a quote from Servais, noting:
“Where the power is come from in the game — obviously we’re going to focus on it. You see guys… I call it ‘hunting their pitch.’ They’re shrinking the strike zone and trying to drive the ball out of the ballpark, trying to get the ball in the air more consistently. I think that’s the biggest change. Hitters are more aware, mechanically, of their swings — what they need to do to get the ball in the air. And not chasing pitches, but sitting on their pitches and trying to pull the ball in the air. The result has been more home runs.”
This quote, as I’m sure you’ll agree, is NOT about speed at all, but about home runs and what’s causing the HR spike in the game. Servais points to “hunting” pitches, and I wanted to see if this was true or not. So, how should we go about verifying this? How would “hunting pitches” and then really swinging hard at a narrower band of the strike zone (“shrinking the strike zone”) show up in pitch data? To me, I think, given the quote, that we’d see fewer swings overall and more called strikes. Do we? No, we do not.
What we DO see is that total pitches are up notably, by over 10,000 compared to 10 years ago, and by *22,000* compared to 2015. The problem is that these “extra” pitches aren’t a bunch of called pitches – they’re increased swings. Batters are swinging the bat more frequently these days, rising from about 45% of all pitches in 2008-2010 to about 47% from 2015-2017. With increased HRs and a rising league-wide on-base percentage, the league is seeing both more plate appearances than we did during the little batting ice age of 2010-2014, but also a steady, unbroken increase in pitches per plate appearance. Even though there were more plate appearances in 2008 compared to 2017, 2017 saw more pitches as pitches per plate appearance grew dramatically between 2014-2017. Again, batters offer at more of these pitches, as both foul balls and whiffs are up, while called strikes are down. Called balls dropped around 2009 and stayed low through 2015, and then recovered slightly in 2016-17; this is consistent with everything we’ve heard about the growth of the strike zone over this time period.
All of these swings haven’t related in more balls in play, of course. Strikeout rate has risen in each year between 2008-2017, and the league’s HR rate has followed suit. Walk rate declined markedly at the start of the 2010, and while its grown in each of the last three years, it’s still below 2008-2009 levels. While the trend isn’t quite as clear, fastball usage league-wide is lower than it was in previous years, dropping in each of the last four years, and dropping three entire percentage points from 2009-2017. This would seem to play a role in the increase in whiffs and perhaps foul balls.
The story here, such as it is, has much less to do with Servais’ conjecture and much more to do with league-wide trends that would seem to require more pitchers (to throw more pitches) and that recognize that batters are swinging more. The M’s have endeavored to control the zone in recent years, not so much by increasing whiff rates/K rates (though they’ve done that), but by limiting bases on balls. The M’s walk rate lags the league average, while their K rate is more-or-less at parity with the league. The M’s began the 2008-2017 with a remarkably bad pitching staff, leading to bloated walk rates. Jack Z’s new-look M’s quickly took care of that, but they did so by dropping K rates as well. It’s stabilized more recently, but the M’s again fell behind the league average K rate for the first time in quite a while in 2017. Still, what we’ve seen, especially over the past 5 years (the years of Safeco’s new dimensions) is that walk rates have been consistently low, while K rates have been more or less average. It’s been easier to get to an average or better K-BB% or K/BB ratio or whatever than it has been for many teams, and this despite the M’s not spending a ton of money on big-name free agent pitchers (though of course this time period covers peak Felix). What HASN’T been so easy is keeping the ball in the ballpark.
The jump in HRs league-wide is both striking and much-discussed; I don’t think it’s a shock to anyone at this point. But what’s interesting to me at least is just how much quicker things have jumped for the M’s. The M’s allowed over 200 HRs in a season a couple of times before 2016 – once in the old Kingdome, and once in 2004 with a truly, memorably bad pitching staff. To do so in two straight years is remarkable, especially given where the club was in 2013-2014. If the M’s gave up HRs at their own 2014 HR% rate, they’d have given up 141, or nearly *100 fewer* than 237 they actually gave up. Even using the league-wide rate for 2017 – the year of the HR, remember – would produce 202 HRs when applied to the M’s actual number of batters faced, or 15% less than they actually yielded.
If you’ve followed this site for a while, you know that this is something of a theme. The question for the M’s is what to do about it. There are a lot fewer HRs than there are walks and strikeouts, and like so much, HR rate is volatile when you restrict it to one team and one year. HR rate results from a variety of factors, from pitcher quality to opponent quality to ballpark mix and even the atmospheric conditions in games…all mixed in with a big random component for chance/variance that isn’t strictly limited to one of the aforementioned factors. I understand why a team might bet on some regression in HR rate, especially when it’s outpacing league-wide HR% growth by so much. But at this point, I’d suggest weighting this factor more heavily than the M’s may have previously. Hell, maybe they do, and they just got bit by a combination of injuries, a new, springier ball and climate change. Maybe it was dumb (bad) luck.
The desire to shift more innings to relievers may be part of the team’s means of addressing this alarming spike. Relievers tend to allow fewer HRs/FB than starters, and thus whereas teams didn’t need a whole slew of set-up men, it might make sense for the M’s to employ as many as possible.
Moreover, they may want to explore just how much of the HR explosion, or more accurately, the batters’ swing changes to generate elevated contact, are related to changes in how batters react to fastballs. I touched on this in May, but looking at the data over 10 years versus 2 months is instructive. The share of HRs that have come on FBs has bounced around over this time period, from 59.2% in 2008 to 58.5% in 2017. It hasn’t dropped significantly; though it’s down from its peak in 2009 and down a bit from the average, it’s still close to 59% overall. What *has* changed is the share of *fastballs* that become HRs. In 2008, 0.71% of FBs went over the fence. After a brief increase for 2009, this ratio dropped noticeably from 2010-2014, before rising back to 0.74% in 2015. Since then, though, it’s jumped again, to 0.82% in 2016 and then 0.89% in 2017. For the Mariners, though, the picture is even *more* dramatic. After starting at the league average of 0.71% in 2008, the M’s have steadily seen more fastballs become dingers than the league average. In the HR-era, though, they’ve been even more dinger-prone, rising to 1.04% of all fastballs in 2017. The M’s have *consistently* given up more HRs per fastball than the league, and if anything, the gap is growing.
Again, this may simply be the result of injuries dramatically reducing their own fastball quality last year. But the M’s have been pretty consistent in terms of the HRs they’ve allowed on FBs: they’ve ranked in the top 10 in baseball 5 of the past 6 years, and ranked 4th last season. The M’s have talked about how they evaluate a pitcher’s fastball, and the multitude of individual metrics that make up an overall fastball grade. It’s possible that giving up a few more HRs is the price they need to pay to have solid K/BB numbers without breaking the bank for name-brand, free-agent hurlers. But this seems like it bears watching. Much of the overall data fits the narrative we’ve already created: batters are swinging at more pitches now, reacting to the growth of the strike zone. The time lag means swings are up even after the growth in the zone stopped. HRs are up due to the ball, and the growth in OBP means more plate appearances and thus more HRs even as fastball usage declined. What I can’t yet fit into some narrative or just-so story is why the M’s have been quite so homer-prone, or why their fastballs are smacked around the way they are. To compete in 2018 and beyond, the M’s need this to change. That’s not news, that’s not a meaningful conclusion. The M’s offense does not appear to be much better in terms of OBP than last year’s model, so to the extent their runs-allowed/runs-scored improves, it sort of stands to reason that the improvement will come from the pitchers. Health will play a role, of course, but figuring out this HR problem would make that job much, much easier.
Comments
7 Responses to “The Changing Battle Between Batters and Pitchers”
Leave a Reply
You must be logged in to post a comment.
How about REALLY throwing out the book and going with a pitching staff comprised of 17 relievers? The closer and two setup pitchers stay in their usual roles while the other 14 go into a rotation of one pitching the first 2 innings then take a day off with the others all filing in the middle 5 innings as match-ups suggest? Also those other 13 could also fill in to allow the setup and closing pitchers adequate rest days.
Relievers that can overpower for 1 to 2 innings are far more common and much cheaper than traditional starters that have been needed to go 6 or 7 innings….
Thinking waaaaay outside the box but if you look at the evolution of the game since the days of Will White pitching 680 innings in 1879, this is the inevitable direction of the future.
Just saw this Travis Sawchick article at FG that goes really well with the fastball-centric stuff at the end of this post:
https://www.fangraphs.com/blogs/gerrit-coles-crucial-pivot/
Congrats to Dave, I guess, on joining the Padres organization.
Wow, thanks for the heads up about Dave Cameron going to the Padres (formerly often sarcastically called the M’s “hated rivals” here due to their interleague match-ups). I do still read USSM religiously and only occasionally read FanGraphs so I only read his work maybe every two or three months, but Dave was and is a very good writer and analyst.
Congratulations are certainly in order for Dave! Would be cool if he would pop in here one last time for a farewell.
Getting back to the article, I think it’s worth pointing out that most of arms that had real dinger issues last season are gone (Gaviglio, Heston, Fien, Overton, Bergman, Curtis, Lawrence, Gallardo, and Scribner) or in a depth role (Miranda, Moore, Iwakuma, and Altavilla). Others, such as Povse and De Jong, don’t even appear to be in an immediate depth role.
Some of those guys were cut loose or demoted fairly early, but there’s little question injuries had us dependent on arms that never should have reached the mound. Many were minor league free agents.
The two pitchers returning to a regular role that had elevated HR/9 rates are Felix and Ramirez, though Ramirez doesn’t walk many and there’s hope for Felix improving at least somewhat.
Dingers were no doubt an issue last season, but I think there’s no real reason to expect more of the same in 2018, especially if they lean more heavily on the bullpen.
The Sawchick article on Cole suggests that a decrease in fastball usage (commensurate increase in use/quality of breaking/secondary pitches), increase in deception, and increase in endurance are things that ought to improve Cole’s success in the current league context (and it is inferred, helped Scherzer improve when he left Arizona). How do you see this playing out with our current staff? (I’d start by inviting Kuma, Moyer, and maybe RA Dickey to hold seminars at the Pitchers Summit this winter/spring).
Marc, as you noted in the article the injuries resulted in reduced quality & the penchant for preaching high fastballs compounded the problem.
Felix suffered some pretty epic bad luck & we should see some regression there. The rotating cast of piggy back style starters who pitch to contact last year was hard to watch for me personally as I love to geek out on pitchers with projectable upside. The strategy was understandable considering the extreme circumstances but the org did not have the depth to pull off such a gambit.
Miranda (37), Gallardo (24), Gaviglio (15), Moore (14), Bergman (12), Lawrence (9), Pagan (7), Albers (6), De Jong (5), Overton (4), Fien (3), Heston (3), Scribner (3), & Chooch (1) managed to give up 143 of those home runs in a mere 665.1 innings. Over half of those guys are out of the org & the only two returnees I want to see on the mound for the M’s from that list would be Miranda as a reliever if he sees a slight uptick in stuff & Moore in September if he excels in AAA.
Paxton (9), Felix (17), Leake (1), Ramirez (12), Kuma (7), Gonzales (5), Povse (1), Diaz (10), Phelps (0), Vincent (3), Scrabble (2), Pazos (7), Zych (2), & Altavilla (9) meanwhile gave up a combined 85 home runs over 699.1 innings. These fourteen pitchers should comprise the core of the 2018 pitching staff along with Nicasio & Rumbelow.
No skirting around the topic, I still firmly believe this team should add a top FA starter as there is payroll space. The prospect capital has been severely depeleted from all the wheeling & dealing so a midseason deal would be an unreasonable mortgaging off of the future. Yes, there are risks attached to such a contract but the draft pick penalty for signing an Arrieta, Cobb, or Lynn is lesser than it used to be & for Darvish there would be none at all. What was the point of banking on Gamel, Healy, & Vincent as risky bargains if the money saved was not going to be allocated for a more certain roster upgrade? Since the lineup is pretty set the obvious play would be to fill the hole near the top of the rotatation & further limit the amount of dingers yielded to the opposition that helped sink the 2017 season in the small window of time when Cruz, Cano, & Seager are at the top of their games enough to make competing for a championship a legitimate possibility.