Examining possible ranges of outcomes: Relief Pitchers

With that out of the way, I’m going to spend the rest of this article talk about projections, because what else do we have?! There are a lot of different projection systems out there these days, which is a good thing because no one system is perfect but the more data points you can throw into a sample the more useful the information you do have should be. Baseball Prospectus’s PECOTA system is one of the most famous systems and has been around forever at this point, and again while it’s not perfect it is one of my favorites.

Specifically, I love that PECOTA doesn’t just publish the 50 percent projection. That is still their main output, but BP will also publish different percentile projections for every hitter, essentially creating a ceiling and a floor for a player. They go mostly by tens of percentage points, but also include the one percent range and 99 percent range projections. Essentially, using their model they simulate the season tens of thousands of times and these are the performances that happened 90 percent of the time, 10 percent of the time, et cetera.

I’m going to take a look at those ranges for the players on the Red Sox roster (click here for BP’s depth chart and links to individual player projections) today and see what the more relatively realistic possible range of outcomes are for everyone, looking at the 10th and 90th percentile outcomes. Today I’ll look at the relievers, and I’ll look at who I would expect to hold the top seven spots in the unit.

Brandon Workman

10th Percentile: 0.0 WARP

90th Percentile: 0.8 WARP

Before I talk about Workman specifically, I should probably mention the grain of salt with which you should take these projections. We know the limitations of projections to begin with, and they become more extreme as the samples get smaller. Relievers, of course, live on a small sample size year over year. That being said, Workman’s projection is interesting to me. He was obviously phenomenal last year, and that was despite walking almost six batters per nine innings. A lot of that was based on success on balls in play, both in terms of limiting hits and home runs. Although much of that was earned from a great curveball/fastball combination, a projection system will never buy that on a one-year sample. PECOTA does see him keeping a high strikeout rate, which surprised me since 2019 was his first year with over a strikeout per inning, but his walk rate jumps from 3.8 to 5.5 per nine from the 10th percentile to the 90th, and with it his ERA goes from 3.24 to 5.26 while his DRA- goes from 82 to 108.

Matt Barnes

10th Percentile: 0.1 WARP

90th Percentile: 0.9 WARP

PECOTA sees Barnes as the reliever on this roster with the highest ceiling, and I have a very hard time arguing with that. The righty is undoubtedly frustrating and seems to have a six-week period just about every year where he looks unusable, but he also has longer stretches of looking elite. He is truly one of the best strikeout pitchers in baseball, and PECOTA actually underrates him here. On his 90th percentile projection, he is projected to strike out 13.6 batters per nine innings. That’s a huge number, and it’s also lower than each of his last two seasons. Even his 99th percentile strikeout rate of 14.1 per nine is over a strikeout less than his rate in 2019. Anyway, as with Workman, the difference here comes down to control, with his walk rates ranging from 4.2 to 6.0 per nine, though even at the 10th percentile projection Barnes is nearly league-average with a 102 DRA-. His 90th percentile mark, by the way, is 79, which would be his worst since 2016.

Josh Taylor

10th Percentile: 0.0 WARP

90th Percentile: 0.7 WARP

This was the projection I was most interested in seeing. While relievers are fungible year-to-year and neither Workman nor Barnes are close to established as consistent back-end arms, they at least have enough of a track record that we have some idea who they are. Taylor has just one major-league season under his belt, and PECOTA is generally a conservative system, particularly when there’s less data to judge from. Taylor’s 2019 was really, really good, and as I’ve talked about before there’s not much to look at that screams regression, but I figured PECOTA would be a bit more conservative. Instead, on a rate basis, they have him being very slightly better than Workman with a DRA- range of 80 to 108. His projection is for 12 fewer innings. I would certainly take Workman if given the choice, but this comparison is more interesting than some may realize, and PECOTA giving Taylor the higher ceiling has me re-thinking my initial impressions here.

Photo by Ronald Martinez/Getty Images

Darwinzon Hernandez

10th Percentile: -0.1 WARP

90th Percentile: 0.8 WARP

If you were to ask me before looking at the numbers here who I thought had the widest range of possible outcomes in the bullpen, I would have went with Hernandez without thinking about it. As it turns out, he’s tied with the next guy by PECOTA’s formula. I would still go with Hernandez. Now, this projections is a little out of whack because they project him for five starts, but we’ll call those opener starts. The overall inning total is still 56, which sounds about right assuming a full season in a reliever role. As one would expect, control is a major issue here with a 10th percentile projection of nearly seven walks per nine innings. Ultimately, I think these projections might actually be a bit too conservative — which, again, is PECOTA’s MO — because I think the realistic ceiling is higher than an 82 DRA- and the realistic floor is lower than 113.

Marcus Walden

10th Percentile: -0.3 WARP

90th Percentile: 0.6 WARP

I really don’t have much to say here besides simply that I disagree. While acknowledging that all relievers are unpredictable and have a relatively wide range of possible outcomes, Walden would have been last on my list if I was ordering these guys in terms of widest ranges of possible outcomes. Instead, he ties with Hernandez with the widest range. Over 92 ²⁄₃ career innings, Walden’s career DRA- is 77.4. His 90th percentile outcome here is 86. Even his 99th percentile is 78. I think his ground ball style gives him a floor that stays at least at replacement level, and based on what he’s done so far in his career I can’t believe his ceiling isn’t a little higher. They’re very different pitchers, but I’d probably have Walden’s range of WARP the same as Taylor’s above.

Heath Hembree

10th Percentile: 0.3 WARP

90th Percentile: 0.8 WARP

Again, this is just a flat-out disagreement between me and PECOTA here, as they have Hembree as the safest reliever in this bullpen. While recognizing the difficulties of projecting relievers, I can’t fathom how a system would come to this conclusion. At the 10th percentile, Hembree is projected for a 97 DRA-. He finished last year with a mark of 134 and has been above 100 in three of his five seasons with at least 25 innings. The floor here should simply be much, much lower. In fact, I’d probably switch this range with Walden’s and I’d have an easier time buying it for both players. I will concede that Hembree’s ceiling of a 74 DRA- is buyable, though I would certainly bet against it. He’s had stretches where he’s looked outstanding plenty of times in the past and in 2017 he finished with a mark of 72.

Ryan Brasier

10th Percentile: 0.2 WARP

90th Percentile: 0.9 WARP

Without the aid of numbers, Brasier’s range doesn’t seem terribly off to me, though I think him having the highest ceiling is a bit out of whack. I’d call that more of an error with the ceilings on the upper-echelon guys being too low than having an issue here, though. That said, I’m not sure how a system can accurately project a guy like Brasier, who was an out-of-nowhere success story in 2018 before pitching poorly enough to be demoted to Triple-A in 2019. I’m a believer in Brasier in that I think he can get back to being a very useful major-league reliever in 2020. I’m just not quite as high on the floor as PECOTA.

Austin Brice

10th Percentile: N/A

90th Percentile: N/A

I included Brice here because I think he pretty clearly was on track to make the Opening Day roster before baseball was shut down, but he’s not even on this projected depth chart. I’m not entirely sure why.

Overall

10th Percentile: 0.2 WARP

90th Percentile: 5.7 WARP

Overall Pitchers

10th Percentile: -0.6 WARP

90th Percentile: 12.5 WARP