A Simple Way to Measure, and Appreciate, Innings Eaters

Let’s start by acknowledging that baseball is really hard. It’s hard for hitters, who over the last decade have had to keep up with rising velocities, increasing spin, and greater extension from pitchers. And it’s hard for pitchers, who adhere to stricter strike zones, and who face hitters increasingly willing to sell out for power and are more well-versed in pitcher arsenals than ever.

Given how hard baseball is, this post highlights some players who had something of a rough time of it, but persevered nonetheless: innings eaters. Now, to my knowledge there is no one set definition of an innings eater, aside from generally a pitcher who throws a lot of innings. How much “a lot” is certainly comes down to any given pitcher’s era and numerous other factors. Additionally, being an innings eater says nothing explicitly about the quality of those innings. A lot of aces are innings eaters, but it’s generally accepted that not all innings eaters are aces.

Given the ambiguity here, I have used a very simple definition of innings eater for our purposes. An innings eater is a pitcher who throws considerably more innings than the quality of performance would suggest that he would have thrown. This definition holds some water in my mind because it highlights that aces, while often being… eaters of innings, attain that workload more through skill and success than through the dogged perseverance of an innings eater.

Let’s break that down some. Generally speaking, pitchers who perform well throw more innings because their performance is rewarded with additional chances to pitch. When examining pitcher-seasons over the last decade-plus in the scatterplot below, one can see that this is broadly the case: the lower a pitcher’s season ERA, the more innings they pitch that year. Now, there are obvious exceptions and variance here. Unfortunately, good pitchers with strong ERAs succumb to injury all the time. So too may teams may try to preserve their best pitchers by limiting their innings (looking at you, Garrett Crochet, though not so this season). Still, on the whole, those with a lower ERA are generally allowed to take the ball more often, and for longer, thus allowing them to eat more innings.

It’s worth noting that while ERA is far from ideal for measuring pitcher quality, it is valuable here when considering innings eaters; we want to know who continued to pitch despite their actual outcomes versus expected results. The goal here is not to correlate to, or predict future, performance but simply to summarize prior outcomes.

Back to that aforementioned definition: an innings eater is a pitcher who throws considerably more innings than the quality of performance would suggest that he would have thrown. Using a simple linear regression model —represented by the turquoise line in the scatterplot above— pitcher innings pitched was predicted using only pitcher ERA as an explanatory variable. Those pitchers with the greater errors/residuals between predicted innings pitched and actual innings pitched based on their ERA alone (i.e. more innings pitched than their season ERAs alone would suggest), were considered innings eaters in this case. Thus, we can control for pitcher success —ERA— and consider those pitchers who carried a lot of the load despite their results.

Let’s first take a look at the leading innings eater —by this definition— in each of the last 10 unshortened seasons.

This table highlights those players who provided their teams with plenty of innings regardless of results that left quite a bit to be desired. Pitcher ERAs on this list range from 5.62 to 6.91 despite inning totals reaching into the 170s. Despite poor results, there is clearly a willingness (resignation?) to send out a guy, such as those on this list, who can soak up some innings.

Alas, Patrick Corbin may take the (dubious) crown as most consistent innings eater in recent years after having made this list 3 of the past 4 seasons. Meanwhile, Ubaldo Jiménez led the way among all pitcher-seasons by this measure with his 2017 campaign. That said, most of these leaders appear to have thrown inning totals on the lower end of the IP distribution among all pitcher-seasons considered; the histogram below depicts that distribution.

There appears to be even more evidence that innings eaters, when performances are measured in this manner, appear more often only up to a certain point in innings pitched. The table below depicts innings eater leaders when the data is grouped by the nearest 10 IP.

The greatest ERA residuals by IP bucket generally sit well north of 2 all the way up until 180+ IP. This suggested the innings eaters, as they are defined here, aren’t getting so long a leash that they pitch ~190 IP. Put another way, while teams accept needing to send innings eaters to the mound for 5 or so innings per outing, giving them the ball for 6+ innings on a consistent basis is a tall order. By the time we are looking at 220+ IP starters, the line between innings eater and ace is fully blurred.

Finally, after having cut the data on season and 10 IP buckets, the following table simply presents the leaderboard over the last 10 non-truncated seasons.

Jordan Zimmermann and the aforementioned Corbin appear on this list twice. Jiménez again tops the whole thing. On the whole, players here average just 0.4 fWAR, suggesting they are indeed playing near replacement level. That said, they clearly were not replaced, due to a lack of depth, the inertia of contract obligations, or whatever other reason. Perhaps looking through rose-colored glasses, I find some maybe un-measurable value in having a consistent innings eater fill their spot in the rotation when there is a need. No, ~3-4 earned runs over ~5 IP isn’t usually enough to leave the game in line for a win. However, it can mean leaving the game while giving your team a shot, with a somewhat preserved bullpen, and with enough shown that getting another turn through the rotation to eat some more innings is left on the table.