The Rays’ Modified Four-Man Rotation

The prospect of a six-man rotation has either been discussed or confirmed by a number of clubs this offseason. The Angels will use one to help ease along Shohei Ohtani’s development as a pitcher. The Rangers also have plans to experiment with one (although Cole Hamels isn’t an advocate). Mickey Callaway mentioned at the beginning of February that the Mets might utilize six starters at points of the season.

The Rays, as they often do, are trying something different. In this case, they’ve announced plans to use just a four-man rotation in 2018.

#Rays Cash just said they could go with 4-man rotation all season, filling in with bullpen guys when needed to start

— Marc Topkin (@TBTimes_Rays) March 7, 2018

 

 

 

 

 

As manager Kevin Cash suggests here, it wouldn’t be pure four-man rotation. Instead, when a hypothetical fifth starter was needed, the club would just utilize relievers exclusively. So it’s four starters plus a bullpen game.

Given trends going in the opposite direction, this plan could lead to disaster, although it isn’t clear that the five-man rotation is obviously superior. Tampa Bay is in a fairly unique personnel situation, so there’s some logic behind the decision. The move isn’t likely to work, but it might be worth a shot.

When discussing the possibility of the Los Angeles Angels’ use of a six-man rotation, I noted the importance of having starting pitching depth and no ace. To effectively deploy a four-man rotation, the opposite is true; indeed, it’s the presence of an ace and little rotation depth behind him that give rise to the unique possibility. A team also requires a deep bullpen and multiple players with minor-league options, so that, whatever starting depth the club does possess, can be easily moved back and forth between the majors and minors.

The Rays fit those criteria nicely. They have an ace in Chris Archer. After the trade of Jake Odorizzi and injuries to Brent Honeywell and Jose De Leon, the team has also taken a hit to its starting-pitching depth. The team does have a lot of bullpen pieces and features some starting depth, as well, as seen by our Depth Chart projections below.

The depth here doesn’t actually look too bad. Matt Andriese appears to be a capable fifth starter, while Yonny Chirinos and Ryan Yarbrough both experienced some success in Triple-A last season. Newcomer Anthony Banda, meanwhile, produced strong fielding-independent numbers in a brief MLB debut last season.

With eight pitchers capable of providing innings, doing away with the conventional five-man rotation doesn’t seem incredibly necessary. There is, however, a pretty distinct drop from the first five pitchers here to the next three. That drop-off, plus extra off-days built into the schedule, does make the four-man rotation possible.

The extra days built into the schedule by the most recent CBA are actually important here. In 1968, teams played 162 games in 172 days. A decade later, five more days were added. By the late-80s, the number of days in the season was up to around 180, increasing to the 183 of the last few years. This year, as part of the new CBA, the players sought and received a few more days, with Tampa’s season now scheduled to be played over the course of 186 days. Those 24 off-days decrease the number of bullpen games needed.

The table below shows just how a four-man rotation would work over the course of the season.

That’s a total of 27 bullpen games, which amounts to exactly one-sixth of the season. Chris Archer, meanwhile, would take the ball 34 times as part of this strategy. Over the last four years, Archer has taken the ball 133 times, including 34 last season, so the 34 starts he would get in this scenario shouldn’t be overly burdensome given the ace’s track record

If the team wanted to get a bit more radical, they could simply start Archer every fifth day, like we see below.

In this scenario, the bullpen would still be responsible for 27 games, but Chris Archer would receive 36 starts with the No. 4 spot in the rotation appearing on just 31 occasions. This implementation might be worth an extra half-win over the course of the season, but the throwing schedules for starters No. 2 through 4 would be disrupted quite a bit and might not end up providing the gains.

Let’s assume the team goes with the first schedule above, then, giving the top-four pitchers essentially one extra start. In order for the strategy to pay off, the bullpen games would need to outperform an expected fifth starter without negatively impacting the bullpen in the rest of the games. This will be tough to pull off.

Let’s start by assuming that starters average six innings per start, so that, in a typical five-man rotation, starters pitch 972 innings and relievers pitch 486 innings. In the Rays’ scenario, we will have starters go six innings 135 times, or 810 innings. Relievers, meanwhile, will go nine innings 27 times and three innings 135 times, or 648 innings. That’s a a lot of innings for a bullpen to cover. Let’s take a look at the Rays’ bullpen Depth Chart under a normal innings scenario.

The Rays have five pitchers who can’t be sent to the minor leagues. In a presumtive eight-man bullpen, that leaves three open spots with which the team can move pitchers between the minors and majors to ensure sufficient rest. To start the season, Matt Andriese seems assured of a spot, so we are talking about two spots with flexibility. With Andrew Kittredge representing one of the teams better pieces here, that flexibility might be even further challenged. The team might need to carry nine relievers to make this work.

One potential workaround comes from the bumped starters. While the team might not want to commit to Banda, Chirinos, or Yarbrough as a full-time fifth starter, they could choose to alternate the bullpen starts among those players and let each pitcher go three innings before sending him back down to the minors.

American League starting pitchers put up a 4.56 FIP last season. The first time through the order, AL starters produced a 4.23 FIP, though, with a sharp increase to 4.76 the second time through and then 4.79 the third time. (The last figure also benefits from some selection bias: ineffective starters tend not to go through the lineup a third time, resulting in one-third fewer innings). Banda, Chirinos, and Yarbrough are all projected as average pitchers, but the first time through a lineup, they would presumably be slightly above average. By letting those three pitchers go three innings in their bullpen starts, we’ve added 81 innings for which the bullpen isn’t responsible, and they’ve been covered at a decent level of quality.

Even with those 81 innings from minor-league starters, the bullpen still has to come up with an extra 10% in the innings category compared to what we see above. To make up that gap, we are going to see more innings from the bottom of that reliever table. Jose Alvarado, Chih-Wei Hu, Austin Pruitt, and Ryne Stanek are likely to see more innings than represented above. Jaime Schultz, too. The good thing for Tampa Bay is that most of those relievers are projected to record better numbers than a fifth starter would the second and third times through the order. The bad thing is that injuries are likely, which will mess with the rotation plan and the available bullpen arms. When there are bullpen games followed by or preceding a start where the pitcher didn’t go deep into the game, the pen is going to be strained.

We are left with a lot of questions. Will the team be able to stick with this setup if one of their top-four starters — or top-five including Andriese — goes down with an injury? Would a jump in performance by Banda, Chirinos, or Yarbrough change the dynamic? Will the team be disciplined enough to go through the churn of continuously optioning relievers even when their performance dictates a stay in the majors? Will Kevin Cash be able to exercise similar discipline not to wear down the better relief arms he does have?

These questions will likely be answered as the season goes along, but keeping a rotation healthy and bullpen arms rested is nearly impossible in the best of circumstances. This alignment probably won’t last, but for the Rays, it is probably worth trying, anyways.