Can you have too much (relief) pitching?

The Giants set a new franchise record for innings pitched by the bullpen this season. Good, bad, or who cares?

I need to state my philosophy about bullpens and relief pitchers right up front: relief pitchers are, objectively, the worst baseball players in uniform at any given point. Put it another way, the worst players on any baseball team can be found in the bullpen. Here is my one data point to validate this opinion: position players can be used to do their job.

So, from a human perspective, that’s pretty gross. Yes, it can be fun to put numbers into a spreadsheet, but those numbers, good or bad, are generated by discrete event generators known as baseball players, who are, exclusively, human beings. On the other hand, it really sucks when a reliever walks in the winning run or puts two guys on and then gives up a three-run home run to lose the game for your favorite team.

But we need relief pitching, as painful as it can be. And as I’ve written about already this season, the San Francisco Giants have deployed their bullpen game strategy to great positive effect. The Tampa Bay Rays have become a juggernaut because they’ve taken their financial limitations and turned them into heavily game-planned pitching strategies and organizational philosophies that have made the team an annual division contender thanks to relievers.

The Giants don’t have the financial constraints of the Rays or the tanking aspirations of the Pirates, Reds, etc. but they do have a pitching strategy that works awfully well for them, which means they’re going to rely on their bullpen a lot more year after year, especially if they keep letting aces walk (first Gausman, now Rodon). The Zaidi era has seen a massive, permanent jump in bullpen usage:

Here are all the times in major league history where a team’s bullpen threw 600+ innings, which is generally what gets us into the 40% range of total innings pitched by a team in a (non-strike shortened) season.

Last year saw quite a jump in those relief innings leaguewide and that’s largely because of getting all the pitchers to ramp up to a full season workload after the pandemic. Makes a whole lot of sense, but just surveying that list you see a lot of bad teams in the years prior and since using their bullpen a whole lot, and so my premise is that a 60-40/58-42% split in starting pitching versus relief pitching innings is not usually the goal of a good team.

None of that really matters if you’re getting a strong performance out of those arms, but there aren’t more variable players year over year than relief pitchers, and so the idea that a team should just remake its entire pitching staff every season seems wacky to borderline irresponsible, though it is much cheaper.

The Giants have been blamed by some and even themselves for simply CTRL+V’ing last year into this year’s spreadsheet and patting themselves on the back. The data shows that the Giants have been pursuing this trend of more bullpen work regardless of season and even if I grant them all the exigent circumstances surrounding Zaidi’s four seasons as President of Baseball Operations.

This year, the Giants’ arm barn was 18th in baseball, with just 2.8 fWAR in value on a 4.05 xFIP across 650 IP. Last year, it was 5.4 fWAR on a 4.20 xFIP across 623.2 IP. A really high xFIP but in a slightly different run environment and, as such, that 5.4 fWAR was sixth-best in baseball. These numbers are based on just 11 out of 29 relievers returning.

So, to a large degree, blaming themselves for simply returning the same team feels like an emotional dodge. 11/29 returnees is hardly the same as running it back. Their process simply didn’t work out. There’s not a whole lot of evidence in the 21st century that an analytics-driven front office will ditch their process. That usually requires a change in leadership, which isn’t happening in San Francisco anytime soon.

With every sign pointing towards the Giants waving goodbye to Carlos Rodon this offseason, it’s worth looking into what that means for their pitching strategy in 2023. Camilo Doval and Tyler Rogers seem to be the only real anchors in the bullpen, unless they have been able to verify that John Brebbia’s downturn towards the end of the season was largely the result of overuse. And maybe Scott Alexander makes sense in that group, too. Jakob Junis, I’ve speculated is fungible enough for upgrade, but maybe the Giants’ lack of pitching depth means Junis and Cole Waites and other minor league et cetera will wind up playing critical roles, to say nothing of the Spring Training reliever NRIs.

But back to Rodon and the pitching strategy. Ditching Rodon suggests the team is more inclined to stand pat with their rotation, banking on healthy turns from DeSclafani, Cobb, and Wood along with Kyle Harrison’s arrival. They’ve been able to turn up gems like DeSclafani and Cobb, but like Alex Wood and Rodon, they have been injury risks their whole careers, too. So, by virtue of maximizing value there (lower cost deals for potentially high upside results) with these 150-165 IP injury guys, they have built a 40+% bullpen usage rate into their pitching plan.

As far as I know, the Giants have not revamped their organizational approach to pitching like the Rays have done, so it’s not really baked into their processes that any pitcher can fill any role. It’s also really clear that the Giants simply do not have the quality pitching depth of the Rays to make a staff by committee approach work to any great effect. Even with the modest success from their bullpen games this year, the maximum value the Giants got from those starts was exactly equivalent to the collective performance of those relievers. If the relievers aren’t great, then their bullpen games aren’t going to be better. A good starter can outperform a good bullpen game with more consistency.

Dominic Leone and Zack Littell were good finds who helped the team in 2021. They were able to get great leverage innings out of Tyler Rogers in years past. Bullpens are all about the circle of life, and for as much as analytics and biometrics have helped close the uncertainty gap, they’re still pretty much a coin flip. Implementing a high variance strategy year after year seems like an obvious setup for high variance results year after year, and if the goal is to be in the playoffs every season, leaving it all to a coin flip seems silly.

If you want to know what this article’s really about, it’s about re-signing Carlos Rodón. It’s not true that he was a five and dive guy. He pitched 6+ innings in 17 of 31 starts and into the 6th inning in three more starts. Put a better defense behind him and he looks even better. Put a better defense on the field and maybe the relievers are both 1) not as exposed via workload and 2) hosed by bad defense.

But if the objection to a long-term deal with Rodón is the injury risk, it’s not like it’d be impossible to work out a deal that mitigates that risk to some extent. There’s just a number the Giants won’t go to with guys like that, even if they will gladly pay long-time injury risks like Cobb, DeSclafani, and Wood to fill critical innings (knowing the overall cost will be lower, even if the risk is exactly the same). The willingness to play to the middle or at the margins rather than pursue the sure thing, answers this article’s question: the quantity of relievers relievers is a more valuable strategy than the quality of pitching.