Starter Fatigue and Coors Field

The Rockies have spent more than $100 million on relief pitchers this offseason. It’s a strategy with an obvious eye turned toward the postseason, when relievers are becoming increasingly important to team success. But Dave Cameron made a compelling case that the Rockies may not be ready for that sort of roster refinement. They overachieved their expected win total by five games last season and have yet to replace several key contributors—including Carlos Gonzalez, Jonathan Lucroy, and Mark Reynolds—that helped them eke into the playoffs with 87 wins. I don’t dispute that they will still need to fill some of the holes on their roster, but I did stumble upon some evidence that the Rockies’ bolstered bullpen may pay major dividends in the regular season, not just the postseason.

The principle reason teams have relied more heavily on their bullpens in the playoffs is because batters have more and more success the more times they see a pitcher. There is a ton of great research out there that shows how pitcher performance declines the second and third times through the order. David Laurila interviewed a bunch of different mangers on the topic at the 2015 Winter Meetings, and most of them agreed that it’s a concern they have, but one they have to balance with other considerations like workload. The simple evidence I’ll provide is that starters allow more home runs per plate appearance (HR%) the second time through the order than the first time and the third time through the order than the second time.

Coors Field is a home run haven, so it is logical to assume that (1) the home run rates of Rockies starters in home games are higher than the MLB average and (2) those increases are consistent from the different MLB averages the first, second, and third time through the order. That first assumption is obviously true. One glance at the park indices will demonstrate that. But I was surprised to find that Rockies pitchers did not see a consistent bump in their home runs allowed from the MLB averages each time through the order. Instead, they have been pretty much average in their home runs allowed per plate appearance the first and second time through the order before collapsing the third time through. There, they have allowed 1.1 percent more home runs per batter faced than an average starter. It is tied for the 3rd-biggest increase for a team’s home starters since 2015.

Notice how most of the other teams that allow a lot of home runs the third time through the order also allow a lot of home runs the first and second time through. That’s especially true of the Phillies’ and White Sox’s starters, who see between a 0.6 percent and 1.2 percent increase in their home runs per batter faced every time through the lineup. Why don’t the Rockies’ starters follow that pattern? Doesn’t the altitude carry flyballs in the first inning the way it does in the sixth inning? I’m pretty sure that answer is yes, but I have a theory that that isn’t the only thing the altitude does.

In football, it’s pretty common knowledge that teams that travel to play in Denver tend to grow unusually exhausted by the fourth quarter. Football is a sport of endurance, and research has shown that endurance-type athletes perform worse as altitude increases. Baseball is start and stop. Most players need only expend heavy energy when they run the bases or chase a ball hit to their part of the field. They have time to rest between those bursts. But pitchers don’t. They have to continue to throw to each batter every other half inning. It may not seem as physically demanding as running around on a football field, but it is undoubtedly tiring, and I speculate the Rockies starters become fatigued more quickly than other pitchers in their home starts because of the altitude.

Last season, the Rockies had seven starters who threw at least 80 innings. All seven of them saw a higher percentage of their home runs come their third time through the order than their percentage of batters faced the third time through the order. Tyler Anderson was their only starter where it was even a contest between the two.

If the altitude really does increase the fatigue of their starters, then the Rockies are smart to move away from the traditional model of starter and bullpen deployment. And that would be a new development. Last year, their starters saw 24.8 percent of their total batters faced the third time through the order. It was the eighth highest rate in baseball.

Fantasy owners were already shy about using Rockies starters, especially in their home starts. If the team plans to dramatically cut down on the average innings per start, then that is good and bad news. On one hand, it will cut down on their wins and strikeouts. On the other, their ratios could improve dramatically if it proves true that those starters see disproportionate damage in the sixth and seventh innings. Either way, the players with the real opportunity for added value are Rockies relievers in deeper formats. The perception could become that the team’s abundance of quality relievers will hold back all of them from racking up big innings totals. I don’t expect that to be the case. Look instead for the Rockies starters to be throttled back to leave plenty of batters for each of Wade Davis, Jake McGee, Bryan Shaw, Adam Ottavino, Mike Dunn and company to face.