Here, I will examine new free agent signee Scott Feldman of the Houston Astros using the Fielding and Ballpark Independent Outcomes system that I've been publishing data from here at The Crawfish Boxes. If you desire more detail as to how the system works, this recent Fanpost on American League starting pitchers would be your best source of information.
All pitchers who faced at least 250 batters as an American Leaguer during 2013 and averaged at least 10 batters faced per appearance. Plate appearances that resulted in a bunt or foulout or were taken by a pitcher were omitted from the data sample under review.
Methodology in Brief
Using run expectancies and play-by-play data from AL stadiums in 2013, I determined what each of these 12 unique plate appearance outcomes were worth in terms of runs, on average: 1. walk or hit-batsmen, 2. strikeout, 3. infield flyball, 4. flyball to batter's pull-field third of the outfield, 5. flyball to center-third of the outfield, 6. flyball to batter's opposite-field third of the outfield, 7. line drive to pull-third, 8. line drive to center-third, 9. line drive to opposite-third, 10. groundball to pull-third, 11. groundball to center-third, and 12. groundball to opposite-third.
The first step of evaluating a pitcher is to count up how many of each event type he allowed and charge him with the corresponding number of runs that is typical for each event; this yields a sum that gets divided by the total number of plate appearances to arrive at a single number that quantifies what a 2013 plate appearance against the pitcher was worth in terms of runs (the more negative the better).
Then the pitchers in the study sample get scored based on how many standard deviations they were high or low versus the league's average number of runs per plate appearance for starters. I call this quantity their Performance Score, with a 50 score being league-average (pitcher topped or equaled 50% of peers), a 60 being 1 SD better than league-average (topped/equaled 83% of peers), a 70 being 2 SD better than league-average (topped/equaled 97% of peers), a 40 being 1 SD worse than league-average (topped/equaled 17%), a 30 being 2 SD worse than league-average (topped/equaled 3% of peers), and so on with 1 SD amounting to 10 points. To understand the origins of their Performance Score, three other subscores are computed on the 20-to-80 scale: 1. Control Subscore, 2. Strikeout Subscore, and 3. Batted Ball Subscore. An Age Score is also computed to reflect how many SD the pitcher is young relative to league-average.
In the tables to come, values that bettered the league average for that quantity by at least 1 SD are highlighted in green text whereas those that trailed it by at least 1 SD are in red text. Asterisks indicate lefthanded throwers.
Feldman the National Leaguer versus Feldman the American Leaguer
Here I'm scoring Feldman's NL data as if it was accumulated in the AL, since I have not compiled the data in full for the NL season. I'm also assuming that the 12 event types have a similar run value, on average, in NL play as in AL play.
Feldman was better per this system in the first half of the season as a Cub (376 batters faced) than he was in the latter half as an Oriole (382 batters faced), with his Control Subscore and his Batted Ball Subscore each dropping some as an ALer. His seasonal Performance Score of 54 amounts to being nearly half a SD better than league average at that stat (better or equal to 66% of league peers); the 49 Control Subscore means he was essentially league-average at avoiding walks and hit batsmen; the 47 Strikeout Subscore indicates he was a bit below average at accumulating strikeouts; the 62 Batted Ball Subscore conveys that he was more than 1 SD better than average at avoiding the batted ball types that are most strongly associated with runs (better or equal to 89% of peers) .
Feldman versus 2013 Astros Starting Pitchers
So, yes, Feldman would be the best of the returning (and non-returning) candidates for the starting rotation based on 2013 Performance Scores. The spray table below shows that Feldman's high Batted Ball Subscore was fueled by moderate success at both avoiding line drives and inducing groundballs (particularly pulled groundballs, which are the best type from a run prevention standpoint).
Feldman versus Free Agents who were American Leaguers in 2013
And here you can see how Feldman stacks up against other starting pitchers who were eligible to be inked by the Astros as a free agent this offseason (though perhaps unlikely to).
Feldman rates as the 6th best at Performance Score of these twenty-two 2013 American Leaguers. That Feldman rates so well on batted balls could have a fair bit to do with the Astros' pursuit of him.
To what extent does this analysis of 2013 data sway your opinion on the signing of Feldman?