There *is* Such Thing as a Pitching Prospect: the Top 10 by Projected ERA at Peak

This is Jordan Rosenblum’s debut at Rotographs. 

I’ve spent the last couple of seasons building out a fully-fledged projection system. You may have seen the StuffPlus-fueled version of them, ppERA, published by Eno Sarris over at The Athletic, or else variations of them elsewhere. Like most projection systems, mine includes aging curves, major league equivalencies, park factors, league run environment adjustments, historical performance, and regression to the mean.

My pitching projections performed well in terms of predictive accuracy in 2023, generally holding their own against other more established projection systems, with my rookie projections performing especially well. I must, however, concede that Steamer dominated the field overall—all hail! This article unveils my top 10 projected pitching prospects for 2024 and beyond, highlighting names worth breaking the TINSTAAP pledge for.

The original inspiration for developing my own projections was to help analyze prospects. A slightly younger version of me would browse FanGraphs MILB leaderboards, filtering the data in different ways to try to get an idea of how an 18-year-old dominating Class A might project at their MLB peak—surely an exercise many readers will relate with. I found it difficult to do this exercise in my head, however—there are many factors to account for when trying to compare two minor leaguers with different performances at different ages in different leagues. With inspiration from Clay Davenport, my solution was to use major league equivalencies, aging curves, regression, park factors, and league scoring environment adjustments to convert all minor leaguer statistics onto the same peak MLB-equivalent scale, thus allowing for easier apples-to-apples comparisons between players in different leagues, at different ages. While this approach is not a prospect evaluation panacea, nor is it a replacement for scouting—especially for players with small samples, where it is less useful—it provides a fairly straightforward way to evaluate prospect performance that is easier than trying to account for everything you see on a player page or leaderboard in your head.

Moreover, even though my peak MLB projections do not account for everything, e.g., they don’t account for college performance, they don’t account for stuff or velocity for arms (although if I ask Eno nicely that may change!), and they don’t account for raw power or swing speed for bats (hopefully one day!), the peak projections still tend to correlate well with rankings from traditional prospect lists. For instance, last year’s top projected pitching prospects included Andrew Painter, Grayson Rodriguez, Eury Pérez, and Ricky Tiedemann in the top 10—a group also adored by scouts. This year’s top three projected hitting prospects are Junior Caminero, Samuel Basallo, and Jackson Holliday, all top 10 overall prospects per FanGraphs’ recent Top 100—I’ll save the rest of those for a later article!

My peak projections have a similar approach to past iterations but with a few methodological tweaks. For instance, one open question is how to regress prospects. Should you regress them all toward the major league mean—or is it better to regress them toward some minor league mean? After some deliberation, I have settled on regressing them toward a weighted average of both means based on their probability of making the majors, with their probability of making the majors derived from their age relative to the league they are in. For instance, using historical data, most 17-year-olds in Class A end up making the majors, while most 23-year-olds in Class A do not. So, a 23-year-old in Class A would mostly be regressed toward the minor league mean, while the 17-year-old is regressed toward the MLB mean. A second tweak is that I now use different regression amounts for each league based on the split-half reliability of each statistic in each league—something I’d not have been able to do without Bill Petti’s baseballr package. This was a minor tweak as the regression amounts generally don’t differ much from league to league.

With slightly further ado, the table below shows the game’s top 12 starting pitching prospects by peak projected MLB ERA (yes, I featured 12 because I wanted to include Cade Horton). The old 50-inning rookie eligibility threshold was used. As there is a lot of variation in how deep prospects pitch into games when they are in the minors, I take the peak projections for each prospect and then convert them all to the same 20 total batters faced per game baseline. So, for instance, Mason Miller’s projections as a pure reliever would look much better than what you see below, as what you see below assumes he will have a starter’s workload at peak (let me dream!). The projections assume a pitcher pitches in a neutral park in the American League. They are on an American League 2023 scale with the following league averages: 4.28 ERA, 23.2 K%, 8.5 BB%, and 41.9 GB%.

Each of these names has received considerable industry hype, with most making various top 100s this offseason. The top three, Tiedemann, Painter, and Harrison, match the recently published ZiPS top three pitching prospects (ignoring NPB and 2023 draftees, as ZiPS captures NPB and college statistics and I do not). Waldrep, Horton, Lin, and Thorpe each made the ZiPS top 100 as well. The names that didn’t make the ZiPS list were generally those with bigger relief risk or those that have already debuted in the majors (which might reflect different definitions of ‘prospect’ used by me and Dan).

Hall, Stone, and Brown were the only ones to not crack the FanGraphs 2024 Top 100, but Stone and Hall did make the Top 100 list for Baseball America. Stone had a rough MLB debut but he possesses a strong enough minor league track record to still project well. Lin was probably the most controversial name to make my top 10 last year as he does not have big stuff, but his age-relative-to-league performance has consistently been very good—and his velocity has improved to respectable levels, probably to be expected for someone so young. His performance has finally earned him some top 100 love from the more traditional lists this year.

Jackson Jobe and, to a lesser extent, Jacob Misiorowski are perhaps the highest-ranked guys on traditional lists that did not make my top 12. This is likely a reflection of a projection vs. scouting divide regarding these two prospects, as both have huge stuff, but their minor league performance has not been consistently great. ZiPS did not rank Misiorowski either, while it ranked Jobe 70th, well behind his placement on traditional lists (e.g., 16th on the FanGraphs list). Steamer also projects a big 2024 talent gap between Tiedemann and Jobe even though they are the same age. Traditional lists tend to rank them closely together. However, scouting reports highlight that Jobe’s stuff has taken a leap forward in 2023.

While only considering the most recent year of player performance is generally unwise from a predictive standpoint, it may be justified in certain circumstances. If I throw out Jobe’s more ordinary 2022 performance and only weight 2023, he’d slot in as the #2 pitching prospect (with a 3.88 peak projected ERA and an 18% peak projected K% minus BB%), basically tied with Painter. He’s also the number two pitching prospect for Dylan White’s RoboScout, which only considers 2023 performance. Additionally, while my peak projections don’t currently account for StuffPlus, my 2024 projections do, which would give Jobe and Misiorowski another boost.

Lastly, I highlighted earlier how the projections are much less useful for small sample players. For these players, scouting reports and pedigree ought to be weighed more heavily. Adding in regression makes it hard to differentiate between players with such a limited sample of performance. An enlightening exercise is to note which small sample players project well before regression is added in. The following players all have excellent regression-free projections reflecting a strong small sample of minor league performance: Noah Schultz, Henry Lalane, JR Ritchie, Teddy Sharkey, Kumar Rocker, Carson Whisenhunt, Paul Skenes, Noble Meyer, Brandon Barriera, Adam Serwinowski, Drue Hackenberg. Many of them are already highly touted prospects. The lesser-known names are a good bet to make some noise and rise up prospect lists in 2024.





23 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Greggmember
1 month ago

Very exciting that Jordan has joined Rotographs! Really looking forward to more great articles like this one!!