Speculating on on Rookie Arms

The release of Baseball Prospectus PECOTA projections is exciting for many reasons, not the least of which is finding the player comparisons that helped inform the projection system. Since PECOTA (and its redesigns) is based in part on finding comparable players in order to assess “true ability” and design aging curves suitable to each prospect and MLB player, these player comparisons are thrilling because they offer in-the-flesh examples to think about when assessing a player. This is especially helpful when assessing prospects, who often provide little evidence for fans and analysts (save for those that rigorously follow Minor League TV, or those that scout minor league games in person). Take Trey Supak, as one example; I’ve not seen much of Supak, but when I see PECOTA comparisons like Lucas Luetge (92 Score) or Vance Worley (92 Score), that puts potential career trajectories and roles in my mind, to be read alongside the scouting reports that will arrive as he advances this summer. This gets even more fun with prospects entering the hype cycle, like Corbin Burnes and Freddy Peralta, who were recently cited as MLB-roster considerations by Brewers manager Craig Counsell: if Burnes’s 2018 development pattern is comparable to Zack Wheeler (90) or Carl Edwards Jr. (93), and Freddy Peralta is comparable to Tommy Hanson (95) or Fautino De Los Santos (91), that provides wide variety of role risk and potential ceiling that can be applied to their statistics and scouting profiles.

Stepping into speculative histories of MLB players, I’ve been thinking about what I’d expect from players like Burnes or Peralta if they reached the MLB in 2018. Both pitchers provide extreme challenges to the observer, and for different reasons.

• Burnes provides a challenge because the righty’s quick ascent to the advanced minors has out-paced the non-proprietary information available about the pitcher, which manifested itself during the 2017 season in the form of widely varying fastball, slider, and delivery reports. From early in the season, when on-the-ground reports placed Burnes as a potentially high reliever-risk profile who could make it work as a middle rotation starter, to late in the season when delivery adjustments and improved fastball reports began leaking out, fans and analysts were essentially given several different pitchers’ worth of information. Entering 2018, then, it’s worth asking whether Burnes is the fastball / slider reliever with middle rotation question marks, the middle rotation arm who could decrease reliever risk by continuing to refine his delivery, or the surging pitching prospect who simply continues to redefine roles by refining mechanics and therefore allowing his stuff to take the next step. All of this can happen with Burnes, or none of it; thus the righty prospect is an acute example of information asymmetry.

• Peralta faces different questions about his stuff and profile, especially due to his diminished stature (compared to Burnes) and his pitching approach (he’s not quite a command-and-deception guy, but he twists and turns his fastballs in order to help his stuff “play up”). In one sense, Peralta is much more conventional than Burnes, insofar as Peralta will be lauded as “the small framed righty who proved everyone wrong” should his frame stick in a functional mid-rotation role. On the other hand, given the frame and lack of a strong fastball, it is worth questioning how Peralta’s arsenal and command will play as he advances in the system. The significance of the righty’s pitch sequencing and location will not be understated. It’s tough to say whether Peralta faces “reliever risk” in the same sense Burnes does, as Peralta’s top role is even murkier than his Advanced Minors teammate.

So, if Burnes and Peralta reach the MLB in 2018, what will their potential roles be? What performance levels might fans expect? With Counsell looking at both pitchers as potential midseason reinforcements for the staff, it is worth digging into potential production expectations.

Role risk will be nothing new to the 2018 Brewers; role risk is all over the roster, even beyond the rookie class. But, in attempting to form expectations about Burnes and Peralta, I turned to other rookie pitchers that are on the MLB roster or in Spring Training camp as non-roster invitees. By my count, these rookies comprise (in order of height) Peralta, Taylor Williams, Marcos Diplan, Jorge Lopez, Aaron Wilkerson, Erik Davis, Burnes, Luis Ortiz, Adrian Houser, Brandon Woodruff, Tyler Webb, and Jon Perrin.

• Several of these arms are already established as MLB relievers or very likely to have relief roles in 2018 (see Williams, Lopez, Davis, Houser, and Webb).

• A few of these arms are organizational depth with questionable futures in terms of organizational plans (see Diplan, Lopez, Wilkerson, Davis, Ortiz, and Perrin).

• Finally, there’s a deep group of potential rotation depth, in terms of starters who could work as rotational replacements or MLB emergency starters in 2018 (Peralta, Diplan, Lopez, Wilkerson, Burnes, Ortiz, and Perrin).

• Woodruff is likely the only pitcher here who is a rotational lock in 2018 (and even writing that leads me to raise some questions, as “There is no such thing as a pitching prospect,” there is no such thing as a surefire pitching role).

Available PECOTA projections (March 2, 2018) for 2018 Brewers rookies (Erik Davis and Jon Perrin projections unavailable). This table features IP projections, as well as Runs Allowed and Runs Prevented figures drawn from projected Deserved Runs Average. Runs Prevented figures are drawn from an average of the 2016-2017 Miller Park / National League. PECOTA strike outs, walks, and WARP are also featured.

With these dozen pitchers, I’ve already laid out extensive role risk simply by categorizing these arms. So, in order to form potential expectations for 2018, I turned to speculative history: namely, how many pitchers in the MLB expansion era matched these arms’ precise height, weight range, handedness, and age as rookies? That is to ask, if these pitchers reach the MLB as rookies in 2018 (e.g., their precise 2018 age), how does their physical profile fare (e.g., their handedness, height, and weight?).

To answer this question, and engage in this speculative exercise, I used the indispensable (and highly recommended) Baseball Reference Play Index to search comparable expansion era players. I searched the expansion era because it is a largely integrated era that also includes a relatively similar form of baseball over the decades in terms of pitching arsenals or prototypes (it tracks the diminishing knuckleball, rise of the splitter / forkball, rise of the fastball / slider profile, etc.). Additionally, as I attempted to search other time frames, I found the reasoning to be rather arbitrary (e.g., if I search the Wild Card Era, should I simply search the Contemporary PED era? If I only search the expanded Wild Card Era (2012-present) will I have a large enough group of players to analyze? And so on).

Using the expansion era, here are the search results from Baseball Reference Play Index. For each player, I searched exact height, handedness, age, and rookie status, while also searching a five percent range in weight (there are some issues with using listed weight). The “Comps” column shows the number of historical rookie comparisons for each potential 2018 Brewers rookie.

This table shows the basic summary of the searches that I conducted using the Play Index tool. What immediately surprised me was the lack of comparisons for some players (like Peralta, Diplan, and Webb), and the large number of comparisons available for “prototypical” starting pitching frames (particularly Lopez, Burnes, and Woodruff). Alternately, for a supposedly “old” prospect, the number of comparisons available for Perrin also surprised me, which suggests to me that there has been a path in MLB history for righties of his size and draft pedigree to reach the MLB, even if it takes a while.

Meandering through the comparisons, there are simply some fun and interesting names that appear.

• 2018 rookie Jon Perrin would reach the MLB at the same time as 2012 rookie Yu Darvish, who shares Perrin’s handedness, age, height, and basic weight range; among older comps, Doug Brocail and Todd Worrell are my favorites for Perrin.

• A few names jump off of 2018 rookie Brandon Woodruff’s spreadsheet, like Corey Kluber (!), Ryan Vogelsong, and (of course), Tim Worrell.

• Corbin Burnes has Michael Fulmer and Steve Bedrosian, Houser has Kevin Quackenbush, while Neftali Feliz and Shelby Miller belong to Luis Ortiz. Fulmer is a fascinating comp for Burnes not because of his excellent MLB performance, but because he shared Burnes’s information asymmetry in terms of advanced minors scouting roles and MLB adjustments to exceed expectations. Feliz and Miller are interesting comps for Ortiz, for as the righty faces questions about innings workload and relief risk, it’s good to remember that successful arms have also shared the righty’s frame, height, and age as rookies.

• The lone comps for Diplan and Peralta? Peralta matches 2000 rookie Byung-Hyun Kim (!), while Diplan matches 1965 rookie Dick Selma.

• Obviously, there is a world of gray area here, as these comparisons say nothing about a player’s draft or international development status, stuff, place in the minor league system, etc. Moreover, for my search, I did not use these comparisons to create trends or aging curves, so what I am attempting should not be read alongside the (much more methodologically sound) PECOTA system.

From these searches, a batch of 257 rookie comparison seasons resulted. In order to analyze potential performance markers in environments comparable to the 2016-2017 MLB, I indexed each season according to Runs Allowed per Game (RA/G), and sampled post-Strike rookies that played during a season with a run environment within 5 percent of the 2017 RA/G. This analysis produced a sample of 85 rookie comparison seasons. The following table shows the basic production range for this 85 rookie population:

For this analysis, I focused on Innings Pitched and basic league-level Runs Prevented (I did not investigate historical park factors for this analysis). To compensate for the lack of park factors, I produced a range of runs prevented estimates. Overall, this group of pitchers did not average high innings pitched totals (presumably due to their rookie status, in many cases), with the average pitcher in this sample working nearly 39.0 innings.

This sample can be further categorized to focus on each specific Brewers rookie’s physical profile. In the next table, a range of runs prevented (RnPrv, LowRnPrv, and HighRnPrv) accompany Innings Pitched and Runs averages and standard deviation (IP, IP_StDev; Rn, Rn_StDev).

This table undoubtedly demonstrates a wide range of potentialities for these Brewers rookie pitchers. The innings pitched variance alone suggests that these players could range anywhere from “Did Not Play” to 70-to-80 IP seasons. It’s easy to get excited about some of these results; for example, 75 innings pitched and five runs prevented from Corbin Burnes, or 56.3 IP and six runs prevented from Taylor Williams would represent excellent rotational and bullpen support for Milwaukee.

Yet, the basic averages might provide some hint as to why the Brewers front office has remained quiet on the pitching free agency market thus far: even averaging around 35 innings and providing moderate runs prevented totals, this group of pitchers has a physical pedigree that suggests competent replacement depth can come from rookie ranks. Of course, it is impossible to fully extrapolate from these historical rookie comparisons to the Brewers 2018 pitching staff; this is not a correlative relationship. Yet, through this speculative historical analysis, one can find that Milwaukee has some solid physical characteristics with plenty of historical counterparts (especially Lopez, Burnes, and Woodruff); there are also some nearly one-of-a-kind profiles (see Diplan, Peralta, and Webb). Ultimately, these dozen rookies present Milwaukee’s front office and field management with an array of options to collect outs and prevent runs, even if they’re a group of wide-ranging replacements.

Baseball Prospectus. 2018 PECOTA (March 2, 2018) [CSV]. Retrieved March 4, 2018 from baseballprospectus.com.

Baseball Prospectus. “More PECOTA” (Player Cards). Retrieved March 4, 2018 from baseballprospectus.com.

Baseball Reference. Play Index. Sports Reference, LLC., Sports Reference LLC, 2000-2018. Searches conducted March 3, 2018 from baseball-reference.com.