KATOH Projection System from Chris Mitchell

Young players are the most valuable assets in baseball. They are incredibly cheap during their pre-arbitration and arbitration years, typically making a fraction of their market value. These years are also when they can be expected to appreciate in performance level. Good teams are built around this quality young talent and well-run organizations will hoard these assets.

More from Yankees News

• Yankees add to championship core with Wilmer Difo signing
• Aaron Judge’s ‘breadcrumbs’ showed he was never leaving Yankees, has more moves in mind
• Yankees analyst Cameron Maybin projects surprise landing spot for Gary Sánchez
• Yankees swipe intriguing minor-league FA lefty flamethrower from Braves
• What does Carlos Rodón’s new jersey number mean for Domingo Germán?

However, it is very difficult to project exactly how these young players will perform when they get to the Majors. Chris Mitchell developed his model for projecting Major League performance for young players and explained it in an article on the Hardball Times.

Mitchell’s model is named KATOH after Yankees’ prospect Gosuke Katoh and relies on observing correlations between past minor league stats and future major league performance.

First, he checked for statistical significance of various metrics at different levels of the minors. Said another way, he wanted to find which statistics recorded at certain levels of the minors portend success or failure in the Majors. The statistics include age, BB%, K%, ISO, BABIP, and SB% and almost all have predictive value at each level of the minors. As Mitchell notes, walk rates in the low levels of the minors are actually not statistically significant which makes sense because these players are so young and so far away from the Majors.

The idea that Rookie Ball and Low-A players are harder to project than AAA players is further proven in a table created by Mitchell that shows it’s both harder to predict if such players will reach the Majors and how they will perform if they reach the top level (using various WAR cutoffs as a measure of value). This conclusion seems obvious, but it is certainly instructive and valuable to see the numbers bear this out.

Mitchell readily admits that Minor League stats are not all that should be used when projecting Major League performance as scouting reports are valuable, too. Stats and scouting reports should be used together to generate information and optimize decisions. Mitchell does, however, think that KATOH has value and uses Yankees’ prospect Aaron Judge as an example. Judge batted .308/.419/.486 across Low-A and High-A last year and Keith Law placed him on his Top-50 prospect list.

However, KATOH projects that Judge has just a 54% chance of playing in The Show and just a 13% chance of generating more than 4 WAR through his age 28 season, approximately the period of time the Yankees will have control of him at reduced costs. Mitchell notes that Judge is old (22) for this level of the minors and that his batting line is heavily bases on balls driven which KATOH doesn’t believe has much predictive value in the low levels of the minors.

Finally, Mitchell notes that KATOH doesn’t include defense. He makes an adjustment for positional value (SS is more valuable than 1B) in one of his final tables. An adjustment for defense relative to average at the position is not included. He reveals what KATOH projects for several position players at the very end. Some of the best prospects in baseball are at the top of the projected WAR total through age 28 including Mookie Betts (21.6), Joc Pederson (18.3), and Kris Bryant (16.0).

The lone Yankee on the list is Alex Palma, an outfielder in Rookie Ball. KATOH pegs him at a 78% chance to make it to the Majors and a projected 9.1 WAR through age 28.

KATOH is a very interesting system and it will be interesting to see what it projects for other Yankees’ position player prospects if Mitchell rolls out more of these articles.