FanPost

Fielding- and Ballpark-Independent Outcome Stats of Sickels’ Pre-2013 A, A-, B+, and B Starting Pitching Prospects

Jose Fernandez - USA TODAY Sports

.

Introduction

This is a system that I have developed to grade the performance of starting pitchers using only statistics that are marginally influenced by hitting environment, fielding, and luck. Four stats are involved in the analysis and they have zero to weak correlations with each other limiting the potential for a pitcher's performance at one to influence their performance at another. Via multiple regression equations, I can explain nearly 50% of the variation in run allowance over large samples of minor league starters with these 4 stats (the remaining "unexplained" 50% would stand to be driven by factors that are not included in the model like luck, fielding, ballpark effects, sequencing of batted ball events, etc.). I mostly use this system with pitchers of one organization but to apply it requires evaluating tons of pitchers from other organizations since each pitcher winds up being compared to their level’s pitching peers. Thus, I have accumulated quite a bit of this sort of performance data on prospects and much of it can’t be found on internet stat sites at present. Here, I’ll share some of that data on more highly revered starting pitcher prospects.

.

Study Group

Pitchers who: 1) received an A, A-, B+, or B grade from John Sickels in his pre-2013 organizational reviews, 2) faced at least 200 minor league batters during 2012 above the two transitional rookie leagues (Gulf Coast League and Arizona League), and 3) faced at least 15 batters per game in so doing (a "starter"). Only their 2012 data accumulated above the two transitional rookie leagues will be reported here. Yes, it is 2012 data. About a third of the data was manually compiled from the MLB Advanced Media archived web data. The rest was pulled from minorleaguecentral which references the same source.

.

Statistics Assessed for Each Pitcher

  1. BB&HBP%: The sum of walks and hit batsmen, divided by plate appearances against. Here, we’re assessing control.
  2. K%: Strikeouts, divided by plate appearances against. Here, we’re evaluating the pitcher’s ability to miss bats.
  3. LD&OFFB%: The sum of line drives and outfield flyballs surrendered, divided by the number of nonbunted batted balls. Batters hit for a higher average and slug more on these batted ball types than on the other two nonbunted types, and logically that drives run allowance up (here's what minor league batters did versus these 60 pitchers in 2012 by nonbunted batted ball type: LD+OFFB = 0.457AVG/0.765SLG, GB+IFFB = 0.195AVG/0.216SLG).
  4. OFLD&OFFB Pull%: The number of line drives and flyballs hit to the batter’s pull-field third of the outfield, divided by the total number of line drives and flyballs hit to outfield. Batters hit for a higher average and slug more when they hit a line drive or fly ball to the pull-third of the outfield versus when they hit a line drive or flyball to the center-field or opposite-field third of the outfield (pull-field third = 0.652AVG/1.286SLG, center-field third = 0.419AVG/0.625SLG, opposite-field third = 0.407AVG/0.615 SLG ), which in turn impacts run allowance.

.

Computation of Performance Scores

Performance scores are generated for each stat based on how many standard deviations (SD) better or worse the pitcher was versus the mean of same-handed "starters" who pitched at the same level (High A, for example) during 2012, following some slight mathematical corrections for differences between leagues of the same level. That’s true except for the OFLD&OFFB Pull stat; given the lack of instant accessibility to those values I am currently resigned to comparing each pitcher’s value to a mean and standard deviation that has been compiled over about 100 (for lefties) to 200 (for righties) pitchers of the same dexterity who pitched at various levels of the minor leagues rather than at a specific level. Otherwise, a 50 performance score signifies level-average performance, with a score beating 50 indicating better than level-average performance and each 10 points equaling 1 SD. For those who prefer percentiles over that scouting-rooted grading scheme, percentiles can be found on all of the graphs and tables that follow. Being 2 SD better than level-average (70 Score) amounts to the 97th percentile (pitcher is equaling or bettering 97% of level peers), 1 SD better than level-average (60 Score) amounts to the 84th percentile, level-average (50 Score) amounts to the 50th percentile, 1 SD worse than level-average (40 Score) amounts to the 16th percentile, 2 SD worse than level-average (30 Score) amounts to the 3rd percentile. Besides those 4 performance scores, an Overall Score is generated by weighting the 4 scores disproportionately based on their statistical associations with run allowance (20% BB&HBP Score, 36% K Score, 21% LD&OFFB Score, 23% OFLD&OFFB Pull Score) and scaling the result away from 50 such that 10 points of it amounts to 1 SD.

.

Results

.

BB&HBP Score

2012spborbetter-bbhbpscore_zpsb9f30b5a_large

With 31 of the 60 evaluated pitchers falling below level-average and only 6 pitchers being a SD better than level-average, you can appreciate that control isn’t an overly prioritized parameter in terms of how prospects are valued in the public domain. Clayton Blackburn and Nick Maronde were the two standouts in this category among the study group during 2012. Archie Bradley and Danny Hultzen were definitely laggards.

.

K Score

2012spborbetter-kscore_zpsa4bd8ca5_large

This is a skill that clearly carries lots of weight in terms of how starting pitcher prospects are valued; only 6 of 60 pitchers fell below level-average at K%. Dan Straily and Kyle Smith stood out here during 2012, with several others beating level-average by 2 SD: Robbie Erlin, Henry Owens, Tony Cingrani, Jose Fernandez, Trevor Bauer, Shelby Miller, Roberto Osuna, Dylan Bundy, Noah Syndergaard, Chris Archer. Pulling up the rear were Sonny Gray, Luis Heredia, and Martin Perez.

.

LD&OFFB Score

2012spborbetter-ldoffbscore_zps2ad52a9b_large

This is another skill that the group isn’t overly strong at, with 34 of 60 rating below level-average. Jarred Cosart and Taylor Guerrieri stand out by a good margin versus the rest at avoiding line drives and outfield flyballs during 2012. Jake Odorizzi ranked last, just behind Kyle Smith.

.

OFLD&OFFB Pull Score

2012spborbetter-ofldoffbpullscore_zps6764b0bb_large

While this is a rather unfamiliar stat, you can appreciate that this collection of arms rates rather well at it with only 18 of 60 falling below level-average. This is probably mostly a function of that these sorts of prospects tend to have very good fastballs which are difficult to turn on, or at worst they do a good job at keeping the offspeed stuff out of the batter’s wheelhouse. At one end of the spectrum you see Guerrieri in the top spot making another appearance with Hultzen and Jose Fernandez nipping at his heels. At the other end you see Victor Sanchez which isn’t terribly distressing given that he is extremely young and pitched only in short-season ball, though a Trevor Bauer sighting there raises questions about his prospective batted ball fortunes considering his level of physical maturity and nearness to becoming a full-time big leaguer.

.

Overall Score

2012spborbetter-overallscore_zps9eafb328_large

As a reminder K Score gets the most weight in the computation at 36% followed by OFLD&OFFB Pull Score at 23%, LD&OFFB Score at 21%, and BB&HBP Score at 20%. Not surprisingly, this group scores well with only 6 rating below level-average. The top score for 2012 belongs to Blackburn with Syndergaard, Fernandez, and Guerrieri bunched a few points behind him.

.

Age Score

The other factor to take a look at is the pitcher’s age relative to the level or levels at which he competed. An Age Score can be computed for each pitcher just as with the above scores by determining how many SD younger each was versus what was average for each level that they appeared at in 2012. An Age Score over 50 indicates that the pitcher was younger than level-average. Not surprisingly, only 3 of these 60 prospects were older than the average for their level(s): James Paxton, Asher Wojciechowski, and Alex Meyer.

2012spborbetter-agescore_zps2d20de69_large

.

Wrap-up: All Scores

The tables below summarize each pitcher’s scores (and relative percentiles of pitching peers equaled or bettered in parentheses) for 2012. Scores that beat level-average by at least 2 SD are in highlighted in dark green, those that beat level-average by 1 SD are in light green, those that trailed level-average by 1 SD are in light yellow, and those that trailed level-average by 2 SD are in bright yellow.

2012sp-fabioscorestablesickelsborbetter1st30-corrected_zpsca4df3a2_large

2012sp-fabioscorestablesickelsborbetter2nd30crop-corrected_zps20c8aada_large

As a disclaimer, this method is not intended to predict which prospects are best or which ones are most likely to succeed as major league starting pitchers. Rather it is an approach that grades the statistical performance of starting pitchers versus same-handed starting pitchers who competed at the same level(s) during the same season. As such it provides an opportunity to indirectly compare how different prospects rate versus their respective level peers and perhaps more usefully it allows one to track how a specific pitcher’s performance at specific fielding-, ballpark-, and luck-independent stats of consequence change over time as they mature and face stiffer plate competition.

.

Future Plans

Many of the reviewed pitchers who have yet to graduate the minors are just now reaching batters faced totals for 2013 where the numbers become worthy of review. And other B or better starter prospects who failed to meet that 200 batters faced requirement a year ago have now or soon will. So compilation of the 2013 data is ongoing. Based on the 2013 numbers to date, the chances of a pitcher's 2013 score on a stat changing versus the 2012 score by 16 points or more would be 8%, the chances of a score changing by 6 to 15 points would be 42%, and the chances of score changing by 5 points or less would be 50%. So subtle steps forward or backward would stand to be more the norm than giant leaps from year to year.

.

In This FanPost

Topics
Teams
Players
X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Minor League Ball

You must be a member of Minor League Ball to participate.

We have our own Community Guidelines at Minor League Ball. You should read them.

Join Minor League Ball

You must be a member of Minor League Ball to participate.

We have our own Community Guidelines at Minor League Ball. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker