With minor league placements and pitching roles being decided for the organization's arms in the coming days, let's take a brief look back at how the righthanded candidates for organizational rotation spots pitched during the 2012 minor league season.
To review the performance of organizational righthanders who started in the minors during 2012 and quantify how well they pitched relative their peers based purely on analysis of statistics that are largely independent of defense and the locales where they were accumulated.
Righthanded pitchers in the Astros organization who pitched mostly as starters during the 2012 minor league season and in so doing faced at least 150 batters at or above the High Rookie League level.
These four stats explain much of the variation that exists between righthanded minor league starters in terms of run allowance and were computed for each pitcher and for each league they appeared in:
1. BB&HBP%. This is the sum of walks and hit batsmen divided by batters faced.
2. K%. This is strikeouts divided by batters faced.
3. LD&OFFB%. This is the sum of line drives surrendered and outfield flyballs surrendered divided by the number of non-bunted batted balls (and is equal to 100% minus the sum of the pitcher’s groundball rate on nonbunted batted balls and infield fly ball rate on nonbunted batted balls).
4. OFLD&OFFB Pull Rate. To generate this stat for a pitcher I filter out every fly ball and line drive that reached the outfield and segregate them into one of three groups based on which third of the outfield the ball wound up in per the play-by-play description of milb.com (the information referenced is furnished online by MLB Advanced Media): 1. the batter's pull-field third, 2. the batter's center-field third, or 3. the batter's opposite-field third. This is a stat that as yet cannot be looked up anywhere but seems to be very important in terms of the outcomes of batted balls. The table below summarizes the statistical consequences of where minor leaguers hit their outfield line drives or fly balls versus righthanded starters based on nearly ten thousand such events that I have logged from the last several minor league seasons (and treating sacrifice flies as at-bats).
Slugging percentage increases by a factor of two when a line drive or flyball goes to the pull-field third of the outfield versus either of the other two zones and batting average increases by a factor of 1.5. Whether one goes to the center-field third or opposite-field third seems relatively inconsequential, all things considered, so this stat from a computational standpoint is just the percentage of outfield line drives and outfield fly balls versus the pitcher that are hit to the batter’s pull-field third. Attention to this stat identifies pitchers who can consistently beat hitters who put their fastball into play and pitchers who excel at keeping their off-speed offerings down and/or out of the batter’s happy zone.
Grading Each Pitcher’s Performance Relative to Their Peers
Each pitcher’s peer group was defined to be the righthanded starters of their minor league level. The BB&HBP%, K%, and LD&OFFB% of every righthanded starter who faced 150+ batters in each league of a given level were pooled into a single set of data for that level (using minorleaguecentral.com's leaderboard feature), and the mean and standard deviation for each of the 3 stats were determined after small adjustments were made to account for differences that existed between leagues of the same level. Each prospective 2013 Astros righthanded starter was then rated based on how many standard deviations better or worse they were versus level average for the stat; a rating score of 50 connotes level average, with each 10 points on the scale equating to one standard deviation and any score above 50 being better than level average. Pitchers who appeared at multiple minor league levels had the rating scores for each stat from their two or three levels reduced down to one rating score by weighting their level scores based on the percentage of batters they faced at each stop. Since it was not practical to compute the OFLD&OFFB Pull Rate stat across all starters of a given minor league level, the mean and standard deviation for that stat were computed from a sample of 90 minor league seasons of righthanded starters on which I have data. Otherwise, each of the Astros’ righthanded starters was rated similarly as above on the stat.
Finally, an Overall Score was generated for each starter by giving 20% weight to their BB&HBP Score, 36% weight to their K Score, 22% weight to their OFLD&OFFB Score, and 22% weight to their OFLD&OFFB Pull Rate Score (those ratios were derived from results of multiple regression analyses involving batted ball SLG and run prevention); this Overall Score gets scaled away from 50 in either direction such that 10 points again amounts to one standard deviation of it. Lastly, to ease with comparisons, each pitcher’s seasonal BB&HBP Score, K Score, OFLD&OFFB Score, OFLD&OFFB Pull Rate Score, and Overall Score was rounded to the nearest number ending with a zero or five. As a frame of reference, appreciate that being one standard deviation better than average is very good (60 score) and being two standard deviations better than average is exceptional (70 score).
Starters who mostly pitched at the AAA level
The recently demoted Lyles pitched very well in AAA at the start of 2012 and remains incredibly young by level standards as we wait for that sort of performance to translate to MLB. White’s flyball and line drive stats were very similar to Lyles’ but his walk and strikeout stats trailed Lyles’ (this was true both in the PCL and in MLB). Peacock still has youth on his side but can he throw enough strikes, strike out enough batters, and limit the damage on fly balls to a level that would allow him to start regularly in the majors? Ely posted an impressive Overall Score but his high OFLD&OFFB Pull Rate would stand to be his Achilles’ heel as a big leaguer and a natural offshoot of his subpar fastball velocity; as was the case in 2012, he’ll have to excel at the other stats to minimize that vulnerability in the majors.
Starters who mostly pitched at the AA level
All four were noticeably young relative to their AA and AAA peers. Cosart’s 75 at LD&OFFB Score is the highest score on any stat in the complete sample of righthanded starters studied here (75 = 2.5 standard deviations better than level average). Cisnero’s high K rate, marginal control, and high line drive and flyball rate seem to point to late-inning reliever as a likely future occupation.
Starters who mostly pitched at the High A level
While the scores of these pitchers don’t exactly jump off the screen, all have at least a few things going for them statistically. Perez’ challenging promotion from the High Rookie level Appalachian League in 2011 to the California League in 2012 may have a lot to do with his rating poorest of the six on Overall Score.
Starters who mostly pitched at the Low A level
Beyond his elite K%, Tropeano also offsets his subpar line drive and flyball rate by being difficult to pull aerially. Foltynewicz, the 2012 organizational minor league pitcher of the year, rates consistently average to a bit better than average on all of the metrics and remains a full year young relative to his peers despite repeating Low A in 2012. Newcomer Gillingham posted batted ball stats that would rank among the best of any arm in the Astros system heading into 2013.
Starters who mostly pitched at the Short-Season A level
West joins Lyles and Ely as the third pitcher to earn a 65 Overall Score, rating well above average at both the non-batted ball and batted ball stats. That Velasquez rated so poorly at the two batted ball stats was one of the larger surprises, though it would be premature to overreact to the results given his relatively young age for a short-season A league and his missing of the entire 2011 season.
Starters who mostly pitched at the High Rookie level
Minor pops up on the radar as a bit of a surprise here, becoming the fourth hurler to achieve a 65 Overall Score and rating as the best starter at any level on OFLD&OFFB Pull Rate. Like Minor, Tiburcio and the oft-injured Rodriguez would be classified as a bit old age-wise for this level in posting very good overall stats. Comer rated very poor at OFLD&OFFB Pull Rate, suggesting both subpar fastball velocity and inopportune location of the breaking stuff in his first professional campaign. Lance McCullers, Joseph Musgrove, and Michael Feliz merit attention on these fronts but are lacking in 2012 data.
This analysis appears to do a reasonable job of identifying the strengths and weaknesses of the reviewed righthanders during 2012 in terms of four consequential statistics over which they have some semblance of control. Pre-rounding, the 6 righty starters who pitched best relative to their peers in the 2012 minor league season based on the Overall Score stat were Aaron West (67), Jordan Lyles (66), Daniel Minor (65), John Ely (63), Jarred Cosart (61), and Nick Tropeano (61). Remember that current performance is but one aspect of the evaluation of a pitcher and much can change over the next handful of seasons that could impact performance positively or negatively (fastball velocity, pitch effectiveness, arm health, etc.). Additionally, those at lower levels will see more elements of their pitching repertoire tested as they climb the organizational ladder, pitch deeper into games, and face a higher prevalence of advanced hitters and opposite-handed bats.