Innings limits: Can a pitcher have too much rest heading into October?
During the first week of the 2008 season, the Freakonomics blog published an interview it had conducted with Bill James. Readers could submit questions for the authors to relay to James. One exchange therein went as follows:
Q: Do you feel, given the right personnel, that some teams should try a four man rotation. If not, why not? If so, which team do you think is best suited and why?
A: I think it is plausible that that could happen and could succeed. I would explain my feelings about it this way: that between 1975 and 1990, two changes were made to reduce the workload of starting pitchers in an effort to reduce injuries. First, we switched from a four-man to a five-man rotation. Second, we imposed pitch-count limits on starting pitchers, starting at about 140 and then gradually reducing that to about 110.
I think it is clear that at least one of those changes was unnecessary, and accomplished nothing. It is possible that both of them were unnecessary and accomplished nothing, but the better evidence is on the side of the pitch limits. I think it is possible, based on what I know, that the starting rotations could go back to four pitchers with no negative consequences.
Obviously, eight seasons later, James looks less than prophetic there. During each of the past two seasons, at least one team has given at least a short trial to a six-man rotation. This year, 51.9 percent of all starts have been made by pitchers on at least five days of rest -- the highest share ever.
Percentage of games started on five days' rest or more (since 2000)
| Season | Percentage |
| 2000 | 43.3 |
| 2001* | 47.7 |
| 2002 | 44.2 |
| 2003 | 46.0 |
| 2004 | 45.7 |
| 2005 | 43.7 |
| 2006 | 45.9 |
| 2007 | 45.7 |
| 2008 | 46.7 |
| 2009 | 47.0 |
| 2010 | 47.8 |
| 2011 | 50.2 |
| 2012 | 50.9 |
| 2013 | 49.3 |
| 2014 | 51.3 |
| 2015** | 51.9 |
| (* 2001 numbers affected by 9/11 attacks) | |
| (** 2015 stats through Sept. 20) |
James wasn't exactly wrong. He said that a four-man rotation was feasible and possible, and could be successful. He didn't say it was probable. His hypothesis remains untested. What's jarring is how it seems certain not to be tested, maybe ever again. Injuries haven't been prevented more successfully as teams have gradually decreased their reliance on each individual pitcher on their staff, and while run scoring is down, "pitchers are better" is perhaps fourth or fifth on a good list of reasons for that.
This season, Phillies right-hander Jerome Williams is the only starter to make more than seven consecutive starts on four or fewer days of rest. (He made 10 consecutive starts that way, mostly because the Phillies are terrible and he wasn't hurt.) Only Garrett Richards and Roenis Elias have made as many as seven consecutive starts every fifth day. There have been nine streaks of six starts on four days, and 13 of five straight starts. Of note:
The Rays (who have famously limited their starters within games more than any other team in the league, in keeping with what James believed to be the more credible workload management paradigm of modern baseball) have managed to send their guys out there on regular rest consistently. Nate Karns and Alex Colome have both had streaks of six straight starts on four days. Chris Archer has twice had five-start streaks of regular rotation work, and Colome has had another five-gamer. The Rays lead the league in starts on what we'll soon call short rest (four days of it), with 86, through Sunday.
Precisely two prospective playoff starters -- Gerrit Cole and Dallas Keuchel -- have made as many as five consecutive starts on four days of rest this year. (Cole, in fact, made six straight starts that way.) Cole's 18 starts on four days this year tie him for the fourth-most in the league, and would make him easily the most acclimated pitcher in the playoffs. Cubs co-aces Jake Arrieta and Jon Lester have just 14 such starts apiece. Jacob deGrom, the unquestioned ace of the October Mets, has only made 10 such starts. Jaime Garcia has made only six of his 17 starts on regular rest, never in consecutive trips through the rotation, and his ERA when he has tried it is 4.89.
It's now so common to have a fifth day of rest between starts that it's nearly certain that teams need to reconsider the way they have pitchers prepare between starts. The playoffs loom for many pitchers who are only halfway (at most) in the habit of pitching every fifth day.
They might have had some bullets saved by teams who elected to give them 31 or 32 starts in the regular season, instead of 34, but they will have to find a comfort zone with a new, even less predictable routine, and quickly, if they want to succeed in October.
As James said, it's not even clear that saving bullets by eliminating starts is effective, so it feels as though many teams have given away regular-season wins (or the fractions of them we can fairly assign to any given player on any given day) for the hazy hope that they can ride a pitcher deep into the fall without driving up his chance of getting hurt. It's a tradeoff with which risk-averse team-builders and coaching staffs seem comfortable, but I suspect it's a bad deal. We'll start to see in October, when the ability of several hurlers to change their rest patterns and preparatory habits without missing a beat will be tested on a big stage.
