December 18, 2009
Checking the Numbers
Swinging Role Reversals
Last week, we took a look at swingmen, those pitchers that spend a decent amount of time in both the rotation and bullpen during the same season, doing so as a means of gauging the true expected performance differential when a pitcher shifts roles in either direction. The number of pitchers of this ilk have declined over the last few decades, but they still surface from time to time for one reason or another. Some are young prospects who, when called up, are instantly installed in the pen, to develop confidence, to get exposure, just help out in middle relief, or a combination of the three. At other times, putting young starters in the pen aids the team's efforts to limit their workloads. In certain situations, the ability to serve as both a specialist and emergency starter provides some additional utility to teams, as they don't need to sign Josh Towers to take a start, or dip into the farm system in the event of a doubleheader or an injury.
Overall, the results indicated that pitchers improve their ERA, FRA, and strikeout rate as relievers. Their unintentional walk rates stayed the same, suggesting that the added ability to rear back and let it rip, as the old axiom goes, may result in more empty swings, but it does not translate to increased control.
The pitchers were broken down based on propensities for issuing free passes and whiffing opponents and classified as either power, neutral, or finesse. In each of these three categories, the pitchers experienced the same results: improved ERA, improved FRA, improved strikeout rate, and sustained walk rate when in the bullpen relative to the rotation. Finesse pitchers experienced the most positive change, which makes sense intuitively, as power and neutral pitchers have less "distance" to travel in terms of throwing harder, and have less to gain by not having to pace themselves and worry solely on hitting spots. One curious finding was the number of pitchers that changed classifications between their two roles as swingmen: of the 897 pitcher seasons since 1974, there were 382 instances of such a shift occurring, but a mere 12 of those situations involved the extreme reversal, of a power starter turning into a finesse reliever.
Even imagining that is tough, given that the terms themselves suggest velocity. Such juxtaposition should not be as automatic as it is, but the fact remains that it occurs. It's easier to assume Bobby Jenks is a power pitcher because he throws 95 miles per hour, when he has actually been more neutral or finesse lately, just like it is hard to fathom someone throwing 88-89 miles per hour as a power pitcher. Because there is such a vast perceived gap between these two classifications, it really piqued my interest that 12 pitchers over the last three and a half decades managed to put up completely different numbers in each role-not necessarily worse numbers, but different, and different enough that they fell into the polar opposite bin.
For an overall perspective, here is a comparison of the statistics for all power pitchers involved in this study, as well as the Fellowship of the Severe Strong-yet-Soft Swingmen:
As SP As RP Power Power/Finesse Power Power/Finesse FRA 4.86 4.88 4.34 6.10 ERA 4.37 4.19 3.99 5.33 K/PA 0.19 0.17 0.21 0.12 UBB/PA 0.11 0.11 0.11 0.08
Unlike all of the previously discussed bins, the group of swingmen who went from power starters to finesse relievers performed markedly worse pitching from the pen. Granted, we are talking about a small group of pitchers that logged four times more frames as starters than as relievers, but for their relief rates to be in such huge opposition to the larger starting sample invites a few questions. Why would they perform better as starters than as relievers? How exactly would they "downgrade" in status? Is it necessarily a downgrade to go from a power starter to a finesse reliever? And who are the dirty dozen in the aforementioned fellowship?
Answering the last question first, the tables below show the 12 men and their stats, separated by whether or not they started:
Starter Year Age IP K/PA UBB/PA FRA ERA Miguel Batista 1998 27 71.1 0.16 0.12 3.21 3.03 Scott Eyre 1998 26 87.0 0.15 0.14 6.98 5.79 Bobby M. Jones 1999 27 100.0 0.14 0.15 7.21 6.03 Robert Person 1996 27 72.1 0.21 0.08 4.86 4.48 Darryl Kile 1991 22 129.0 0.16 0.11 4.15 3.07 Dave Stewart 1986 29 118.1 0.18 0.10 3.73 3.35 Storm Davis 1987 25 75.1 0.18 0.11 5.21 4.54 Ramon Garcia 1997 27 117.0 0.19 0.08 4.06 3.46 Len Barker 1979 23 110.1 0.16 0.12 4.98 4.65 Rudy May 1974 29 116.0 0.18 0.11 3.57 3.03 Garrett Mock 2009 26 78.1 0.18 0.09 6.57 5.40 Clayton Richard 2009 25 136.1 0.18 0.11 5.03 4.42
Reliever Year Age IP K/PA UBB/PA FRA ERA Miguel Batista 1998 27 63.2 0.15 0.07 5.37 4.67 Scott Eyre 1998 26 20.0 0.13 0.09 6.20 3.60 Bobby M. Jones 1999 27 12.1 0.13 0.10 9.23 8.76 Robert Person 1996 27 17.1 0.13 0.09 4.37 4.67 Darryl Kile 1991 22 24.2 0.06 0.13 8.38 6.93 Dave Stewart 1986 29 43.1 0.11 0.09 5.81 5.61 Storm Davis 1987 25 17.2 0.08 0.07 10.63 8.15 Ramon Garcia 1997 27 41.2 0.14 0.06 5.18 4.32 Len Barker 1979 23 27.0 0.12 0.09 6.34 6.00 Rudy May 1974 29 25.1 0.14 0.06 3.48 3.91 Garrett Mock 2009 26 13.0 0.10 0.13 7.08 6.92 Clayton Richard 2009 25 16.2 0.13 0.07 5.97 4.32
Two aspects of these tables stick out initially: first, that only three of the pitchers accrued more than 150 innings in the season, and second, that two of these pitchers experienced the swing reversal just last year. Aside from Dave Stewart's 161
Miguel Batista had bounced around a few organizations in the early part of the nineties, but 1998 marked his first full season in the majors. Pitching for the Expos, Batista almost literally split his 135 innings between the rotation and the bullpen, thriving as a starter in the run-prevention department while struggling a bit with his controllable skills; in the bullpen the exact opposite occurred, as his strikeout and walk rates improved while his ERA and FRA fell. Batista is also the perfect archetype for describing swingmen to someone unfamiliar with the term, as he has appeared in 530 games over 15 seasons, just 237 of which have been starts. In just two of his 12 full seasons, he was utilized strictly as a reliever, with teams either not knowing the best way to implement his talent or basking in the fact that he could go both ways. For his career, Batista is almost dead even in both roles, with a 4.51 ERA as a starter and a 4.68 ERA as a reliever, K/BB ratios of 1.43 and 1.40, and WHIPs of 1.49 and 1.54. While Batista has been used as a swingman for the last decade, it does not appear that he has actually fallen in line with the aggregate by improving as a penman.
Raise your hand if you remember Scott Eyre originally came up as a starting pitcher with the White Sox. On August 1, 1997, Eyre debuted for the White Sox with the quite the uninspiring line: 4.1 IP, 6 H, 6 R, 6 ER, 5 BB, 4 K, losing 9-1 to the Angels. Over his next seven starts, he managed a 3.79 RA over 38 innings, and he pitched well in nine of his 11 starts. The next season, Eyre made 33 appearances, of which 17 were starts, but he never seemed to have a defined role. He struggled mightily as a starter, walking as many as he whiffed, providing little relief in preventing runners from crossing the dish, and while his numbers improved in the bullpen, his transformation from power to finesse isn't all that miraculous given that he was borderline neutral in both instances, the lowest tier of power and the highest tier of finesse. Since 1998-a year in which his bullpen outings lasted multiple innings more often than not-Eyre has made four starts, none since 2002, and he has been limited in relief to lefty specialist duties.
Apologies to the Bobby Jones fans out there-especially the ones that remember, without looking it up, which one I mentioned-but his numbers were so awful in 1999 that no matter what role he was given, his outings should have come with that old Marge Simpson warning about the severe horror in store for viewers on Halloween. Robert Person's season mirrors the 1998 of Batista, in that a kid with raw talent had made it to the big leagues but his employers were not sure in which capacity to max his performance. Person split time in both roles in 1996 before starting on a full-time basis for the Blue Jays the next season. In 1998, Person could not crack the Blue Jays rotation, one featuring Roger Clemens, a useful Woody Williams, stalwarts Pat Hentgen and Juan Guzman, and a 23-year-old Chris Carpenter. A 21-year-old Roy Halladay and all-but-washed-up Erik Hanson also made starts, and the Jays seemed more interested in rookie Kelvim Escobar handling swingman duties. Person struggled to the tune of a 7.04 ERA in 27 innings and did not again crack the rotation until switching to the Phillies in the 1999 season.
Darryl Kile was a rookie in 1991, starting 22 of his 37 games, but his swingman season was more along the lines of a team trying to break in a young pitcher as opposed to being unsure of his role. With someone like Mock, it is too early to make such a judgment, as 2009 marked just his second season. His 2009 swingman compadre, Clayton Richard, is perhaps more like Kile in his inclusion, as he split time as a member of the White Sox, especially as they started skipping the fifth slot down the stretch in 2008; he was used strictly as a starter after being shipped to the Padres in the Jake Peavy deal.
The really interesting cases are those of Dave Stewart and Storm Davis, as the swingmen seasons tabled above came in the middle of their careers, and seemingly came out of nowhere. Take a look at Dave Stewart's 1978-85 and tell me his role. Go ahead. It's hard to tell if he was a swingman or if he was a reliever who made a start or two or if he was a full-time starter. In 1986, he split time between the Phillies and Athletics. After fishing Stewart out of the free-talent pool, the A's eventually found himmore suitable for the rotation with the rest, as they say, being history, as Stewart not only became an effective starter, he led the league in starts every year from 1988-1991. Stewart was always known for throwing hard, but he also threw his fastball very straight, and it took being paired up with Dave Duncan in Oakland as well as the mastery of a splitter to salvage what very well could have been a lost career.
The presence of Davis is misleading here, as he was a starter before and after the 1987 season. After four years with the Orioles (including a brief '82 debut in which he swung, but not long enough to be captured in the sample), the Padres acquired him and wound up using him in both roles. A mid-season deal that sent him to the A's-to meet Mr. Duncan and team up with Mr. Stewart-resulted in his resumed post as a starter. This was part of Sandy Alderson's grabbing of every conceivable starter in the mid-late '80s to shore up his team's rotation. It worked: by 1988, the rotation featured Stewart and Davis, Bob Welch, Curt Young, and Todd Burns.
As you can see, there are many ways a pitcher can both be a swingman and change classifications dramatically. For someone like Davis, he was never pointed towards a career in the pen beforehand. For Stewart, he didn't really take off until he learned a splitter, and at that point he was a full-time starter who never looked back. Looking at pitchers like Eyre, we see that certain pitchers may swing for a bit before settling into a specific, different role for the rest of their careers, while hurlers like Batista will swing for their entire career. Pitchers like Batista are truly the ones of interest here, as they were swingmen for much of their career. Alas, a sample size of one isn't even worth mentioning, let alone using to make any sort of judgments. Based on the overall group of extreme classification changers, however, it seems that extenuating circumstances and the appropriate context shed more light on the reasons for changing from power to finesse than anything related to the pitcher's talent.
Thanks to Christina Kahrl for the assistance with things that happened when I was three years old or under; I didn't catch many A's games in the womb.