It's a bummer that KATOH won't be available to the public anymore, since detailed information about its calculation never left the statistical black box and the creator and curator, Chris Mitchell, has moved on to something presumably better. I did my best to fill some tiny portion of the hole recently by making some minor league hitter leaderboards, based on league z-scores of the main inputs to KATOH, since Mitchell did explain in a general sense what went into KATOH and gave some hints about how important those inputs were, relative to each other.
Now I'd like to do the same thing for minor league arms. Mitchell showed us that Age, K rate, and Games Started % are always important statistical correlates of MLB success. BB% and HR% becoming significant somewhere up the minor league ladder. Here is the chart from Mitchell's reveal article:
For my z-scores I'll be making the following judgment calls. Remember that I want to keep this simple.
Handedness and K^2 will be ignored;
BB% will be used with full weight from A ball and up;
HR% will be given a 0.5 weight from A- to A+ ball, then full weight in AA and AAA (Mitchell said its importance varies up the ladder and it only has "some" predictive value in A-; the "hole" at A+ defies common sense so I will count HR% at that level)
The IP line for inclusion is 30 for short season and rookie, and 50 for A ball through AAA. I wanted to include relief pitchers and partial seasons but not very small samples. The samples range from 68 to 156 players in the various leagues.
Note that HR% is not HR/FB (which is displayed on FanGraphs). For HR% you need to multiply HR/FB by FB%, and it ends up being the rate of batted balls that leave the park.
The way GS% works out, on the tables, is kind of satisfying. The mean GS% tends to be around 60%, and the SD tends to be about 40%, in my league samples. This means by z-score the prospects who started every game get about a +1.0 bonus and the full-time relievers got knocked by about -1.5. I was worried about how GS% would work under my lazy z-score method, but I think it works out fine - relievers with fantastic numbers can still rank relatively well, but prospects pitching mostly in relief get a notable penalty without being completely disregarded.
For the hitters, I slow-rolled the results over five articles. This time everything is on one spreadsheet (LINK) that you can view right now, and I'll discuss the rankings in just two pieces. This one will deal with Rookie advanced all the way through high A.
Rookie
In the Appalachian League, Eric Pardinho is the #1 prospect by my numbers, thanks to his well below-average age and his nearly-elite K rate. At this level, he's not even getting a bonus for his walk rate, which would only have increased his hold on the #1 spot. This guy is legit.
Coming in at #4 is the Jays' rule 5 pick, Elvis Luciano. His score is based on starting games at a young age and having a respectable K rate, which sounds kind of boring but a prospect holding their own at a young age is a significant thing.
The Yankees place a couple of Luises in the Appy top five - Luis Gil (#2) and Luis Medina (#5). Both of these prospects have sick stuff but their walk rates in 2018 were revolting. Mitchell said that walk rates weren't significant at this level, so the Luises will be interesting to follow as they climb the Yankees' minor league ladder.
The Pioneer league is lead by Ryan Feltner, a 2018 4th rounder by the Rockies. Feltner was actually an unsigned Blue Jays pick in 2015. Reports on his stuff seem strong (touches 98 with some promising secondaries) but his numbers in college were not impressive.
Low A
Friend of the podcast Josh Winckowski was the 4th best pitcher in the NWL by the numbers, trailing a trio of teenagers in Hans Crouse, Ramon Perez, and Gregory Santos. When we interviewed Josh recently he mentioned that Crouse should maybe have not been eligible for Baseball America's league top 20 list, but Crouse only made four less starts than Winckowski and his numbers were nuts. Gregory Santos gets plus grades on his fastball and slider and the changeup shows potential, so he'll be a SP prospect to watch. As if the Padres don't have enough pitching prospects already, the Cuban Ramon Perez seems to be a lefty with a well-rounded profile.
The other Vancouver Canadian to rank well was Jose Espada, who was repeating the level. You might remember him as a popular inclusion on deeper Blue Jays prospect rankings a couple of years ago - he was a 2015 pick by Toronto and his numbers were strong in his rookie league debut that year, but he scuffled in 2016 and 2017.
Sean Wymer is on mlb.com's Blue Jays top 30 list, but he only ranks 27th out of 67 qualified league prospects.
From the New York Pen league the guy you want to buy, if you can, in your dynasty league is Cleveland's Luis Oviedo. His league-relative z-score sum of +4.58 was one of the best of this entire exercise.
A Ball
The only Lugnut worth mentioning from the Midwest league list is Maximo Castillo, who ranked 28th out of 155 qualified arms (about the 82nd percentile). This big guy was able to not suck while starting a lot of games at a very young age. I've mentioned him before and reports on his stuff have not been consistent, so who really knows what his chances are.
The MWL top six makes me feel good about my method here: Luis Patino, Hunter Greene, Matt Manning, Jordan Balazovic, Jairo Solis, and MacKenzie Gore. Four of the best pitching prospects in the game, plus a riser in Solis who is already getting noticed in a good system, and a potential sleeper in Balazovic who barely cracks the Twins' top 30 list on MLB Pipeline.
Balazovic is actually from Mississauga. It kind of bites to see a different team get really good value on a prospect from Toronto's backyard.
The South Atlantic League list is lead by a reliever, Demarcus Evans from the Rangers system. This is the only reliever to rise to the top of a league ranking - actually, he's the only pure reliever to rank higher than 7th (a Rays reliever in the International League slid into 7th). Evans had an absurd strikeout rate. He doesn't even appear to have elite velocity, but Longenhagen says he has "hellacious movement" on his curveball, which is a fun description.
A+ Ball
Yennsy Diaz is the highest ranked player from Dunedin. He placed 16th out of 102 players, and his numerical profile is similar to Maximo Castillo - he scores okay because he held his own while being young. Three other Dunedin starters sneak into the league top 30: Zach Logue, Patrick Murphy, and Maverik Buffo. Murphy's stuff sounds exciting and it's great that he managed to stay healthy this year, but perhaps tap the breaks before aggressively sliding him way up your personal Blue Jays prospect list; the Jays put him on the 40 man roster to protect him but his production in 2018 doesn't exactly signal a role 50 player at this point.
In the cold water category, former organizational sleeper favourite Angel Perdomo had a below average score and ranked 70th out of the 102 qualified arms. Despite the reportedly good stuff, age is working against him at this point and the walk rate was more than a standard deviation worse than my sample mean, which is not a good recipe when paired with a season pitched partially as a reliever.
Brusdar Graterol, Ian Anderson, and Matt Manning are the household names forming the top three in the Florida State League but a Bailey Falter fellow is right there with them at #4. FanGraphs scouts are only giving him a future value of "35+" so he's your standout statistical sleeper in the FSL.
In the Carolina League two Rangers prospects, Edgar Arredondo and Jonathan Hernandez, lead the way with very strong z-score sums of over +4.50. Hecter Perez' numbers from the Astros A+ team are not very strong, ranking him 41st out of 103 qualified arms.
The California League was kind to new Blue Jay Andrew Sopko, whose very strong A+ numbers rank him 8th. We also see the best league relative score in the entire exercise here, a +7.21 mark from Chris Paddack, who is getting some legitimate recognition now but still might be underrated and is probably one of the best pitching prospects in the game.
Blue Jays - System Comments
I like Pardinho and Winckowski a lot, but this system is years away from being a pipeline of pitching prospects. Lansing and Dunedin were particularly devoid of high-end and productive arms in 2018, unless you really like Yennsy or Maximo (both seem like longshots to be good big leaguers to me), so the system sort of has a two level gap in the middle of it. There are a few bright lights in AA and AAA, which I'll highlight in the next blog post.
Here is the link again for the full spreadsheet.
Thanks for reading Stat Scouting the Minors: Z-Scores for Pitchers (Rookie to A+) by Nick Hill. If you have any questions or comments relating to this article, we encourage you to leave them below. For all general inquiries, we can be reached at the following:
Twitter https://twitter.com/radio_scouts
Opmerkingen