In response to Unhittable Pitching
As I write this, Ryan Madson, relief pitcher for the Los Angeles Dodgers, is walking in the tying run in the fifth inning of the World Series. He came in with two outs and two men on base for Dodgers starter Hyun-Jin Ryu before walking batter Steve Pearce on five pitches. Pulling Ryu — who had to that point let in one run in four and two-thirds innings and thrown 69 pitches — so early is of a piece with managerial strategy that has seen increasing use in baseball: bullpenning.
As Nicholas Frankovich observed on the Corner a couple of days ago in a post evaluating the “strikeout epidemic” of Major League Baseball, “managers now use their pitching staffs more shrewdly than in the past.” He’s referring to this increased flexibility managers have with their bullpens, and their increasing willingness to pull starters early. The model of a starter pitching seven innings followed by, ideally, one or two relievers closing out the game is being replaced with managers’ pulling their starters earlier, using top relievers at high-leverage points early in games rather than saving them for the end, and occasionally not even using starting pitchers at all, choosing to start relievers at the beginning of games and mix and match from there.
The practice is often called “bullpenning.” Its logic is simple. In the regular season, its soundness is hard to deny. Ample statistical evidence shows us that in the course of a single game, a pitcher’s performance declines each time he faces the same hitter. He is most effective his first time through the lineup, less so his second time through, and so on. Evidence also shows that this performance decay isn’t simply a function of fatigue. Hitters get used to the opposing pitcher’s tendencies as pitchers exhaust their bag of tricks. Managers are pulling their starters early to constantly swap in new relievers and avoid such a decay in performance.
Bullpenning is a strategy at the cutting edge of baseball, and in the playoffs, cutting-edge strategies are frequently deployed. In the 2017 and 2018 playoffs, bullpenning became de rigeur. The Houston Astros gave most of their starters, save for Justin Verlander, exceptionally short leashes; the Oakland A’s started a reliever in the wild-card game against the Yankees; and the Milwaukee Brewers and Dodgers set a record for the number of pitching changes in a single playoff series in the just-concluded National League Championship Series.
I’ve been wondering, however, whether the logic for bullpenning holds up over long playoff series. As a matter of scouting, relievers in the regular season might not be under the microscope the way starters are, allowing them to surprise batters with their repertoires. But in the playoffs, opposing teams are surely scrutinizing every pitcher on the roster and reminding their hitters of those pitchers’ tendencies. More to the point, over the course of long series, hitters see the same pitchers over and over again. Wouldn’t bullpenning lead to a similar decay in performance — one that happens not over a single game but over the course of a seven-game series? Just as starters decay every time they work through the batting order, might relievers decay over several games every time they face opposing hitters? (Not to mention the issue of accumulating fatigue.)
This seems plausible to me, but I’m willing to admit there might be some motivated reasoning (I enjoy games more when the starters stay in through the late innings) or confirmation bias (I’ve been especially attuned to relievers making mistakes this playoffs, from Madsen’s mistake just now — which admittedly is outside the purview of my theory, happening as it did in Game 2 — to Jeremy Jeffress’s giving up a three-run homer in Game 7 of the NLCS) at work. There is, however, some limited statistical evidence to support my hypothesis. In a post from 2017 at the sports-analytics blog empty-lot.com, the author compiled data from the 2017 playoffs and found that while “relief pitchers outperformed starters the first or second time they were used in a series . . . starting pitchers fared much better than relievers on their fourth and fifth trips through a lineup.”
This is what I’d expect to see if my theory were true. Of course, as the author notes, his sample size is necessarily small, and plenty of other variables might confound his analysis. Still, it seems that bullpenning in the playoffs may have its downsides — and that managers should think a little harder about the strategy before pulling their top-line starters.