The Dream of the ’70s Is Alive in Andrew Miller

Since the beginning of this year’s postseason, the present site has become littered with a collection of posts examining the somewhat novel (if also logically sound) deployment of relief pitchers during that postseason. A hasty examination of the archives reveals, for example, a post declaring the advent of the bullpen revolution; a meditation on likely bullpen usage in 2017; and then a third one about how another run might never be scored in a major-league game.

Given this trend, one might suggest that the editors of this site should change its name to BullpenGraphs. But only as a joke, presumably, is why one would do this. Because actually changing the site’s name to BullpenGraphs would represent a huge logistical nightmare — and would almost certainly hurt traffic. And therefore revenue. And therefore ruin the site entirely. Which, for someone who’s employed by that site and also possesses a mortgage, isn’t a particularly amusing joke.

In any case, mostly at the center of this enthusiasm regarding bullpen usage has been Cleveland left-hander Andrew Miller. And for good reason: not only has Miller been predictably effective, but he’s also been ubiquitous. Following last night’s appearance in Game One of the World Series, Miller has now recorded a strikeout rate of 47.1%, stranded every runner who’s been dumb enough to get on base, and conceded zero runs in 13.2 innings. So, roughly as good as possible.

Perhaps more notably, though, is that those 13.2 innings have been distributed over just seven games, meaning that Miller has averaged nearly two innings per appearances. Not only is that unusual for Miller himself (who threw just over one inning per appearance this year, even after arriving in Cleveland), but unusual for basically every reliever in the game — with the exception perhaps of those who are designated as “long men,” “swing pitchers,” and other names designed expressly to elicit giggles from teenage boys.

Amid the collective interest in Miller’s usage this postseason, however — from which the present author isn’t immune, at all — I’ve experienced the nagging sense that perhaps this thing we’re witnessing here isn’t entirely innovative. In the context of the contemporary game, it’s certainly unusual, of course — and there seems to be no doubt that Cleveland has arrived at this strategy as a product of a high-level analysis. But baseball history’s also long. Surely there have been relievers who — if only been accident — have worked in a similar capacity as Miller has during this postseason.

In an effort to identify them, I conducted a search for seasons in which a reliever both (a) recorded at least 50 innings and also (b) averaged at least 1.5 innings per appearance. I then calculated the average of his FIP- and run-based WAR figures (which appear on the site as WAR and RA9-WAR, respectively), and prorated the results to 100 innings. The top-10 seasons by that methodology are as follow. (Or, actually, top-11 seasons: Rich Gossage produced equally excellent years in 1975 and -83.)

Please note, first of all, that it’s not my intention to reveal much about the names that appear here. My interest in this post is primarily to ascertain, in somehwat objective fashion, the identities of those who’ve been used in a capacity similar to Miller (with similar effectiveness) during this posteason.

That said, it’s difficult to ignore one trend among the results here — namely that, while the sample of eligible pitchers for this study includes the 1906 campaigns of George Ferguson and Clark Griffith and Art Hoelskoetter, that the 11 best seasons by this measure all occurred within a 20-year window between 1977 and 1996. Indeed, nine of those 11 occurred in the 10 years between 1977 and 1986. One, looking for the spiritual ancestors of Andrew Miller, ought to begin his or her search right at the beginning of the Carter administration.

Chief among those ancestors would appear to be a pitcher who was actually employed by the same team currently tasked with facing Andrew Miller this World Series: right-hander Bruce Sutter. His 1977 and -79 seasons for the Cubs serve as the rough model for this new-old reliever. Of Sutter’s 62 appearances in 1977, over half of them (32) lasted longer than an inning. The effect was magnified in 1979: of Sutter’s 62 appearances in that season, 41 of them (more than two-thirds, then) lasted for four outs or more.

While Sutter won the Cy Young for that 1979 season, it’s perhaps the 1977 campaign that’s most analogous to this version of Miller we’ve seen in the postseason — not merely for the quality of performance nor for the quantity of multi-inning appearances, but also for that point of the game in which Miller has entered. It has become commonplace now for Miller to enter Cleveland’s games, as he did on Tuesday in relief of Corey Kluber, in the seventh inning. That was also commonplace for Sutter. Indeed, the seventh was the inning he entered most frequently in 1977.

At a team level, Sutter’s dominance and usage did not create instant success. The 1977 Cubs finished fourth in the NL East, at 81-81 (and with a slightly worse Pythagorean record). The 1979 Cubs were worse, recording one fewer win and placing fifth in the East. So the presence of basically the best-ever relief season wasn’t sufficient to bring an otherwise mediocre team to the championship. That said, had the 1977 or ’79 Cubs qualified for the postseason, they would have had the ideal pitcher to help them win it.