This started last night, when Patrick Corbin left the game in the fifth, having thrown 108 pitches to that point. GuruB asked if that was a season high for Arizona: checking, I found it was not quite, as the two Zacks had each thrown 109 pitches. But that in itself was interesting: 92 games into the season, the D-backs have not had a single man throw 110 pitches in a game. The last to do so was Zack Greinke, who threw 111 pitches on September 5 last year. Since then, Arizona have a streak of 105 consecutive games, and counting, with an apparent hard cap on their starting pitchers of 110 pitches.
The decline and fall
The chart below shows the number of 110+ pitch outings for each season in the D-backs history. There were seven last year, about the same number each campaign from 2014-17. Before that point, with various plateaus, the number has been in steady decline from an all-time high of sixty-two such outings in the 1999 campaign. There was a spell in July where eight of nine games passed 110, peaking with Brian Anderson and Randy Johnson throwing 127 and 139 on consecutive nights. You won’t be surprised to learn Johnson was responsible for 26 of these - so, only two fewer that season than the entire D-backs roster has made since the start of 2014. But Andy Benes had 17, and even Omar Daal had nine.
Is it just the D-backs?
They’re not entirely extinct. There have been 129 outings of 110 or more pitches in the majors this season, but they drop off sharply as we go higher. Only twenty-three of these went over 115 pitches, and six more than 120. [The season high is 127, by Trevor Bauer, and Tyson Ross against Arizona on April 20] But these are far from evenly distributed. Two teams, the Indians (19) and Nationals (18) are responsible for 29% of all such starts. At the other end, the Brewers are even stricter than the D-backs. They have also not reached 110 pitches, with a high of 107. Only ten times has a Milwaukee starter reached even a hundred pitches; Arizona have thrown that many on 22 occasions.
Overall, the D-backs have averaged 92 pitches per start this year: that’s actually two above the National League average this year. The range is from 96 (Washington) all the way down to 83 (Los Angeles). Interestingly, the raw number of pitches hasn’t changed all that much over the years. The table below shows, at five year intervals, the average pitches per start, innings per start and median number of 110+ pitch outings per team in the NL. You can see that pitches has decreased slightly; innings somewhat more. But the number of 110+ pitch outings has collapsed entirely, even in just the past five years.
NL starter outings, 1998-2018
There is a perception that outings of over 100 pitches become significantly more stressful on the arm. This appears to stem from the legendary Dr. James Andrews, who co-wrote a study in 1999 called “Elbow injuries in young baseball players,” which came up with 100 pitches as the threshold. It didn’t happen immediately, but the figure appears to have percolated into the mainstream of baseball thinking. Now, most teams (Cleveland and Washington aside, apparently - oh, that wacky Trevor Bauer!) are reluctant to push their starters very far beyond that. And why would you want to do so, when you can turn to one of your army of flame-throwing relievers instead?
90 is the new 100
The new border appears to be 18 PA, which would be 83 pitches at last year’s NL average of 4.63 pitches per plate appearance, with managers increasingly reluctant to let opposing starters see their pitchers for a third time. Shoewizard looked at this last month, pointing out how D-backs starters were struggling there, and we’re still seeing it. After OPS’s of .655 and .695 the first and second time, it spikes to .919 for the third PA. But it’s not something to which the D-backs publicly pay attention, Lovullo saying, “If you’re making pitches and you get into that fifth or sixth inning and you’re facing them for a third time, I’m not going to say that I’m worried about it.”
It is somewhere teams are increasingly reluctant to go. Over the past decade, the percentage of plate appearances for a starter going three or more times through the order has dropped from 28.7% in 2008 to 22.3% this season. The other side of the equation is an increase in bullpen use, and shoe noted that we’re on pace to see 40% of innings thrown by relievers for the first time ever. A factor in this may be the creative use of bullpen churn, or what I’ve just decided to call the “Bracho Bus-pass effect”. There’s no easy way to check, but it certainly feels like we’re seeing much more use of the taxi squad: Silvino Bracho, especially, along with Jake Barrett, Jimmie Sherfy and Braden Shipley.
Those four have combined for 25 appearances and 30 innings - the same as Jorge De La Rosa and around the workload of Andrew Chafin and Brad Boxberger (33.1 each). In effect, the shuttling up and down from Reno has given the Diamondbacks an additional bullpen arm. It might be one generally not given high-leverage innings, but somebody has to throw them, and it allows the team the ability to have a quicker hook, rather than having to “save the bullpen”. Now we’ve got Randall Delgado back, with he and T.J. McFarland both capable of throwing multiple innings, this is a trend which may even accelerate in the second half of the season.
But overall, we may simply need to re-calibrate our expectations as the game of baseball shifts around us. This should happen across the board. For example, as recently as 2007, batting .268 made you an MLB average hitter; this year, you only need to hit .245 to reach the same mark. On the pitching side, in the same year, the strikeout rate was 6.7 per nine innings; now it’s 8.5. We might need to begin doing the same with regard to what we consider an “average” starting performance: right now, 16 outs and 90 pitches appears to be the new normal.