Here, I will update the recent American-League-reliever-centric FanPost with an eye cast towards how the Astros' offseason 40-man-roster relief additions would compare versus AL relievers in general based on their 2013 major league data and the resulting Fielding and Ballpark Independent Outcomes scores.
Methodology in Brief
All non-bunt, non-foulout, non-pitcher-batting plate appearances against the pitcher in the 2013 MLB season were sorted into 12 generic event type outcome categories (either a strikeout, walk-or-hit-batsman, infield flyball, or one of the 9 permutations involving whether the batted ball was a groundball, line drive, or outfield flyball and which third of the field the ball was hit to in relation to the batter’s handedness). Each of those 12 event types has an expected number of runs associated with it based on what is average for the event type given the league (here, the AL) and season being evaluated and the pitcher gets charged with the corresponding number of expected runs for each such event; that sum is divided by the total number of plate appearances to quantify what an average plate appearance against them was worth in terms of expected runs (the more negative that number, the more run-averse that pitcher’s generic outcomes were and the better their expected performance would be).
Each pitcher’s expected runs per plate appearance value is then compared to a reference dataset of 101 values that was compiled from a sample of 101 2013 American League relief pitchers and they are assigned a Performance Score based on where their value would rate versus the average of the sample of 101 values on a standard deviation (SD) basis. A 50 score indicates league-average performance (pitcher topped or equaled 50% of peers), a 60 is 1 SD better than league-average (topped/equaled 83% of peers), a 70 is 2 SD better than league-average (topped/equaled 97% of peers), a 40 is 1 SD worse than league-average (topped/equaled 17%), a 30 is 2 SD worse than league-average (topped/equaled 3% of peers), and so on with any 10-point change in score amounting to 1 SD. Likewise, the pitcher’s performance on control ("Control Subscore", based on their walk-plus-hit-by-pitch rate), strikeouts ("Strikeout Subscore"), and batted balls ("Batted Ball Subscore", based on their value of expected runs per batted ball) relative to the same sample of 101 AL relievers can be quantified on the same 20-to-80 scale, as can their youth ("Age Score").
The tables will include all relief-biased pitchers who pitched for Houston during 2013 plus any pitchers who have been added to the 40-man roster this offseason, if the pitcher faced at least 80 major league batters and did not average more than 10 batters faced per game. Pitchers added this offseason are highlighted in bold text in the tables and asterisks denote lefthanded throwers. Green values beat league-average by at least 1 SD ("very good", or better); red values trailed league-average by at least 1 SD ("very bad", or worse).
As shown previously, this system smiles rather favorably on waiver-claim Downs' 2013 Detroit data. Qualls would grade out as better than average at Performance Score based on his 2013 Marlins campaign, equaling or beating 59% of his would-be AL relief peers; Qualls' 62 Batted Ball Subscore would beat 89% of AL relievers while his 43 Strikeout Subscore would beat just 25% of them. Though Valdes grades out rather well at control and just above league-average at strikeouts, his extremely low Batted Ball Subscore of 20 as a Phillie confirms that his poor 2013 results were fueled by surrendering a high frequency of dangerous batted ball types (like Qualls, Valdes is no spring chicken). Bass’ 2013 Padres performance grades out rather poorly on these measures, save for control. Ideally, with pitchers like Valdes, Bass, and several of these Astros who split their 2013 season between the majors and minors, one would also perform similar analyses with that minor league data and with other recent professional data to help quantify just how typical for them their small-sample 2013 MLB results computed via this system are (the same goes for non-roster additions like Peter Moylan, who faced only 70 big league batters in 2013). Albers joins Qualls and Downs as a third offseason relief addition who excelled at avoiding increasingly run-associated batted ball event types in 2013; a potential fourth one was lost when Veras inked with the Cubs last week after rejecting an offer from Houston. [UPDATE: A healthy Crain seems like the logical candidate to slot into Veras' projected closer role; unlike Veras and those 3 newcomers, Crain is a strikeout standout rather than a batted balls standout.]
The table below shows how these pitchers’ plate appearances break down by event type frequency, with the denominator for the batted ball ones being the total number of batted balls rather than plate appearances.
And below are the results of applying this system to their results against only lefthanded batters (LHB) and their results against only righthanded batters (RHB). Note that in several cases, already small statistical samples are being split still further.
One can squint a bit at Downs' ratings and see a potential late-blooming, equivalently-aged, economical alternative to free agent lefty Boone Logan, who recently signed a 3-year $16 million contract with Colorado; the splits portion of the duo's 2013 scores and subscores were similar across the board save for Logan's LHB Strikeout Subscore being 1.6 SD (16 points) better [Logan grades out on the splits table left-to-right as follows: 68(96%)=55/64/57, 72(99%)=57/70/56, 51(54%)=50/49/53]. Qualls is essentially the same guy versus LHB as he is versus RHB per his 2013 subscores. If Albers could reclaim some of his lost strikeout ability and elevate his Strikeout Subscore by roughly 1 SD, he would end up rating very similarly to how Qualls did versus LHB and versus RHB and closer to league-average at overall Performance Score. Valdes’ 2013 splits values cast him as the prototypical one-and-done LOOGY who would stand to be rather successful against LHB but equally vulnerable on batted balls struck by RHB (especially in MinuteMaid Park). The righty Bass rates poorly against LHB on all three subscores. It is no surprise to see young and/or inexperienced arms like Fields, Clemens, Cisnero, Lo, Bass, Cruz, and Chapman posting such low Performance Scores against opposite-handed bats given that they have not faced many major leaguers in their careers, were not considered outstanding prospects as minor leaguers, and likely would still be starting if they were projected to be average or better against opposite-handed major league batters. [UPDATE: Among 2013 AL relievers Crain trailed only Koji Uehara and Greg Holland on Performance Score vs RHB while grading out a bit below average against LHB given his slightly subpar Batted Ball Subscore versus them].
Glimpsing more generally at the above splits table, one can see that despite the considerable efforts that Jeff Luhnow and his team have committed to reshaping the bullpen for 2014, much room for improvement remains as evidenced by the preponderance of red numbers in the Performance Score column. The bulk of that work may have to be orchestrated via routes other than free agency, inclusive of further developing this and other organizational talent for the role. In a perfect world, the club would execute trades that involve pitchers who fall towards the bottom end of the table and seem less apt to improve on these measures and in return receive a pitcher or two who has demonstrated a recent capacity for rating better on these sorts of generic parameters at the major league level (or projects to do the same). [UPDATE: A healthy Crain gives the top of this bullpen a strong boost that should offset the recent miss on signing Veras.]