Would Justin Verlander and other starting pitchers be more effective if they faced fewer batters?
Over the last half century, MLB has changed profoundly in the way pitching is handled. It used to be expected that starters would try to finish games—or at least pitch until they ran out of gas. Relievers would be summoned only if necessary—and bullpens were manned mainly by those pitchers not good enough to be starters.
Then saves became an official statistic—unfortunately—and the “closer” role evolved, and then the “set-up man” was invented, and now we have the full fad: a parade of bullpen specialists that clog up the end of a typical game.
Regardless of what you think of this, it’s clear that managers now routinely try to manage a game from the end backwards. They have the ninth inning set, the eighth inning predetermined, and options they will use in the seventh. And they hope the starting pitcher will get through six innings.
But because starters are allowed to throw fewer and fewer pitches to protect their arms. the problem, increasingly, is getting to the bullpen.
The World Series this year has amply demonstrated that the unsolved problem in today’s baseball is crossing the bridge from starter to relief specialists, and trying to shut down the other team’s offense in the middle innings. If Kansas City starters ran into trouble before the seventh inning, Ned Yost normally wasn’t willing to use his big three relievers much before their assigned innings. And his reluctance to do so led to disastrous results in Game Five. Bruce Bochy made the same sort of mistake in Game Two. Most managers use that same push-button, unwavering approach day after day.
But here’s the elephant in the room: Knowledgeable baseball minds have long known that batters figure out pitchers better the more times they face them in any particular game. Pitchers have a huge advantage over hitters the first time through the batting order, but after that the balance starts shifting to the hitters.
All manner of recent statistical studies have proven this to be true—and they indisputably show that the batter advantage increases slightly the second time through the order but dramatically the third time through the order. There’s even a handy acronym for this in sabermetric circles: TTOP, for “third time through the order penalty.”
The penalty holds for most pitchers in most games, with rare exceptions for the most stalwart performers or the most exceptional outings.
But baseball practice has lagged beyond statistical understanding of this phenomenon, leading to the middle innings dilemma. Clubs still pay starting pitchers enormous sums (even though they are pitching to fewer and fewer batters), and closers command big money now too. Set-up men are also starting to be paid more commensurately with their burdens and presumed importance. But the least well paid are the middle innings men—and they are almost always the worst relievers on the team.
Simply put, the earlier and later innings are pitched by the best and highest-paid hurlers available—but the middle innings are being sorely neglected. They’re being left to starters whose effectiveness is diminishing markedly as they go through the order the third time—or to the guys on the end of the bullpen bench. And as many games are decided in the middle innings as in any other inning—obviously.
The Giants are one team that seems to be recognizing this issue and sometimes addressing it by using a super starter-reliever, Yasmeiro Petit, to pitch in the middle innings when a starter doesn’t make it far enough, as he did in Game Four of the World Series. There are few analogues in the majors to Petit these days—and that helps make him a very valuable asset.
It usually takes about 100 pitches to get through the order three times—and that’s when starters are usually sent to the showers and relievers summoned because of concerns about fatigue. But TTOP shows that this is not the optimal strategy.
You can reasonably expect something like this from a starter: a WHIP around 1.25, meaning an average of five batters reach base in the first four innings (for a total of seventeen batters faced, or almost twice through the order). To avoid the TTOP, a reliever should be ready to be summoned by the fifth inning.
Smart clubs should therefore try to get their own versions of Petit to pitch the fifth and sixth innings. But there aren’t many relievers around these days who are prepared to pitch two innings or more, and it’s unclear how much rest they need if they do. Perhaps clubs should start experimenting with this option in their minor leagues—maybe seeing what happens when the limit on some relievers is not just one inning but one time through the order?
Another solution to the middle innings problem is simply to pull starters after two times through the order. This means they might throw 60-70 pitches. And if they do, shouldn’t they be able to pitch on three days’ rest? Clubs could find out the answer, again, by exploring this option in the minors.
In 2012, the Colorado Rockies tried a four-man rotation limited to 50-60 pitches a game—for a short time. It failed, but mainly because the Rockies had a bad starting rotation. They didn’t have the right men for the job. And the limit shouldn’t have been dictated by pitch count, but by TTOP.
If such a strategy did succeed, you could use the equivalent of a fifth starter to pitch the second half of some games, as well as a Petit type to work in the others. If the Tigers acquired or groomed a Petit clone and went to a TTOP-conscious rotation (without free agent Max Scherzer), it might look like this:
Game 1- Justin Verlander (4-5 innings), Kyle Lobstein (4-5 innings)
Game 2- Anibal Sanchez (4-5 innings), Petit clone (2-3 innings)
Game 3- David Price and other relievers
Game 4- Rick Porcello (4-5 innings), Petit clone
With occasional longer outings for Price afforded by off-days, this kind of rotation would be flexible, economic, and efficient, and it might yield real dividends in the middle innings, the neglected portion of today’s pitching template. Sooner or later, some team is going to try it again—why not Detroit?