I like doing statistical analysis for two primary reasons. There are two main biases that it combats and we are all prey to those biases at one time or another. The first form of bias is what I would call immediacy bias. A great example of that is the string of short starts that the Astros are currently going through. The temptation is to assume that this has been constant throughout the season when that is almost never the case. There are ebbs and flows to these things, so looking at overall numbers
can help us combat that bias.
The second bias is a local bias. We watch the Astros. Some of us watch them religiously. So, when we are watching one team 162 times a season and the other teams maybe a handful of times, it is easy to look at what your local group is doing and assume that it is either really good or really bad. We simply don’t have anything to compare it to. This becomes important when we are looking at relative strengths and weaknesses of a team. This can come into play in terms of game to game strategy or long-term questions about how the team should be structured.
We call this the lab because I like to start with observations that can be formed into a testable hypothesis. In this case, the hypothesis is that the Astros overly rely on their bullpen for innings because their starters do not go deep enough into games. We can test that simply by comparing them with the rest of the American League to see if that is true comparatively. We will start with the starting pitchers only in terms of innings pitched and ERA.
I should note that the most significant thing here is not that the Astros are 14th in starter ERA, but the gap between them and the Twins for 13th place. As my grandfather used to say, they are so far down in the cellar that they have to pipe light to them. Certainly, if one were to apply park effects to these numbers you could make an argument that the Astros rotation has been the worst in the American League this season when adjusted for inflation.
It shouldn’t be any surprise that they are 13th in starter innings. The White Sox low total is jarring, but the rest of the group kind of settles in. There are three teams within 15 innings of the Astros on the season. Over the course of 120 games, 15 innings is not a completely significant sum. These numbers serve to answer our main question: are our observations valid? The simple answer is yes. This team doesn’t pitch as deeply into games as most of the other teams in the league. The overall numbers include those first couple of months when things were rough, but on balance this has not been even an average starting rotation.
Obviously, we need some more information before we decide whether the lack of starting innings is a good thing or bad thing. Most would reflexively state it isn’t, but I would caution people before making that conclusion. Throughout baseball history, bullpen ERAs typically outpace starting pitching ERA when taken collectively. So, we should expect a positive differential. I’ll throw in the starting to reliever differentials so we see what is happening throughout the league.
Someone once said there are lies, damn lies, and statistics but that was obviously a gross simplification. Numbers are what they are and there is nothing that will change them. The question comes in how we interpret them. The Astros are third in bullpen innings. That is a fact and there is nothing that can change that. The question comes in how we choose to interpret that information.
Conventional wisdom would say that leading in bullpen innings is a bad thing. This is usually where the old-timers wax poetic about how pitchers used to finish their games back when men were men and sheep were nervous. I’d offer a different outlook. I’d simply say that you want to maximize your best performers as much as you possibly can. That means every manager has a different task, The managers with a negative differential are seeing their pens do worse than their starters. So, they should maximize starter innings as much as possible. The teams with positive differentials should maximize bullpen innings.
The best managers are the ones that can adapt to what their teams are actually doing. The Astros have the largest differential in the American League and likely the biggest in baseball. It may seem counterintuitive, but they should probably lean into that as they have lately. Obviously, there are workload concerns that can’t be ignored, but as a simple proposition, the pen is the best part of their pitching staff and that has been true even more so since May 1st.
Of course, this has broader implications in terms of roster construction. What is peculiar about this team is that you have a group of pitchers that have been effective in relief, but not effective as starters. That list includes Tatsuya Imai, Cristian Javier, and Kei-wei Teng. There is another group of pitchers that have worked heavier workloads in multi-inning opportunities. It’s a challenge for any manager, but this team may actually be better sticking to four and five inning starts and letting the pen carry the load. They run the risk of running out of gas, but the alternative is throwing out the worst rotation in the American League and hoping things work out.
