One hundred and sixty-two innings was as good a number as any. By design—"provided that the pitcher has pitched at least as many innings in league championship games as the number of games scheduled for each club in his club’s league that season (MLB Rule 9.22(b))"—the number had a pleasing symmetry with the number of games in a season, which is one of baseball’s three or four most significant figures. And it worked. In 1962, for instance, there were 118 pitchers who started at least 10 games. A majority—58 percent—reached 162 innings. A supermajority (70.1 percent) of pitchers who started at least 15 games did; 80 percent of pitchers who started at least 20 did. If you were a starter, the minimum-innings threshold recognized you.
The league leader that year threw 314 innings, and 18 pitchers threw 250 or more. In a 20-team league, that means that a team leader, for simplicity’s sake, could be expected to top 250—that’s our “max” season. Throwing 162, then, meant throwing 65 percent of a max season. That’s not quite as high a threshold as we expect for qualifying hitters, who must reach about 70 percent of a similarly high everyday-player standard (700 PA), but it’s close. And, importantly, it excluded relievers, as only five since 1962 (Mike Marshall twice, and three others by the skin of their fingers) have reached 162 while starting 10 or fewer games.
But pitcher roles change, and the ancillary measures of pitcher performance don’t always change in accordance. Reader Bryan emailed me today to point out something odd: Ubaldo Jimenez had fallen off the pitcher leaderboards this week, not because he was missing starts but because he was getting knocked out of them too early. Granted that Ubaldo Jimenez is not in contention for the ERA title—maybe the most importantish reason for the innings threshold to exist—but, just as pitcher roles change, our preferred leaderboards (and leaderboard usage) change. It makes no sense to treat Ubaldo Jimenez—who has made 13 starts in 13 turns through the rotation—as not having thrown a full season thus far.
What I’m saying is, it’s time to adjust the minimum-innings threshold. Here’s the proof:
As noted earlier, there were 118 pitchers who made at least 10 starts in 1962*, or six per team. That includes not only the main rotation, but the semi-regulars, the swingmen, the permanent subs, the mid-season call-ups. Of those, a narrow majority qualified as “full time.” If that’s too liberal a field to draw our regulars from, we can bump it to 20 starts, and now we’ve got a solid draw of 84 pitchers, or 4.1 per team—in other words, a rotation. Of those, nearly all—80 percent—qualified as full-time.
There are now 30 teams, of course, and at six per team we’d have 179 pitchers with 11 starts or more; at five per team, replicating the concept of one rotation, we’d have exactly 150 with 15 starts or more. We could try to mimic the four-per-team standard, but that’s an outdated size of a starting rotation. All the same, for completeness sake, it takes us to 124 pitchers with 20 or more starts, same as in 1962.
How many of those pitchers qualified for the ERA title, then?
- 11+ starts: 44 percent
- 15+ starts: 52 percent
- 20+ starts: 63 percent
Buried in the math of these is something telling: Nobody who started 11-20 starts reached 162. The fewest starts by any qualified pitcher is 27—and only one pitcher, Erasmo Ramirez, managed even that. Twenty-seven starts is, in this day and age, a lot of starts! Only two and a half per team managed it. Similarly, only two and a half pitchers per team reached leaderboard existence last year, compared to three and a half pitchers per team in 1962—when there were actually 20 percent fewer rotation spots!
To reach something comparable to 80 percent qualified rate, we now have to bump the games-started search up to 24. A pitcher must take the ball 20 percent more often just to count as full-time.
That’s problem one. But problem two is that pitchers take the ball less often now. The modern rotation expects a pitcher to throw less often, and to throw fewer innings when he does. If the point of this is to hold pitchers to some standard of durability and effort* that applied to Don Drysdale a half-century ago, mission accomplished.
If it’s to capture in a leaderboard or statistical chase all the pitchers who were, essentially, regular pitchers in their clubs’ rotations, it’s a failure in need of adjustment.
*It’s also at least partially a myth that pitchers like Drysdale were tougher or exerted more effort than pitchers today; pitchers today throw fewer starts and fewer innings per start, but they throw more pitches in each inning, because evolving strategies (on offense and defense) require them to; and they throw with more effort on each individual pitch, because evolving strategies ask them to. Measure the energy exerted in 40 Drysdale starts and 33 Max Scherzer starts, and I bet it’s pretty close—or, perhaps, not all that close at all, with the modern pitchers exerting far more energy. (I’d bet confidently that Scherzer’s between-starts routine to keep conditioned for those pitchers is greater than Drysdale’s was, too.)
The league-leader last year threw 232 innings. A “max” season has gone from 250 innings to 200, with 28 pitchers on 30 teams meeting that measure last year. Sixty-five percent of 200 would be 130 innings. In 1962, 130 wouldn’t be enough to weed out relievers. (Well, in 1962 it would have, barely. But a few pitchers most years would have cleared that threshold while starting 10 or fewer—even zero—games.) Now it is. And if the minimum innings needed to qualify for leaderboards was 130 innings, pitchers, our ratios would be almost perfectly proportioned to the standards for qualification we had in 1962:
- 11+ starts: 59 percent
- 15+ starts: 70 percent
- 20+ starts: 84 percent
- Qualifiers per team: 3.5
It’s not quite so cute as "one inning per game in the season," but it's still perfect. Hit publish.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now