Giving Southpaws For Thought: Texas' Starting Five Pitchers
Josh Garoon
Matt Harrison deals against the Royals in Surprise on Saturday, March 6th.The Red Sox did it in 2004, and again in 2007 (with all due respect to an electric Jon Lester). The Cardinals did it in 2004, and again in 2006 (with all apologies to a fading Mark Mulder). The 2004 Yankees did it. So, coincidentally, did the last Rangers team to make the playoffs. What was it? Winning without consistently featuring a left-hander in the starting rotation.
The Rangers' desire to head into Opening Day with a southpaw starter has been something of a media and message-board meme this spring. Assuming right-handers Rich Harden, Scott Feldman, Colby Lewis, and Tommy Hunter have starting spots locked down (not necessarily a safe bet, granted, but one we'll make for the sake of argument), only one opening remains in the rotation. And even if it means losing the right-handed Brandon McCarthy to waivers, many folks apparently want to see Matt Harrison, Derek Holland, or C.J. Wilson get the nod for that final spot.
Why? Let's set aside residual resentment over the John Danks deal, the lingering worry that McCarthy will break (again), and the not unreasonable fear that even if McCarthy doesn't break (again), he simply won't be very good. The abiding concern is that Texas will struggle to contend without a lefty in its starting five. Over the course of a series, this argument runs, sending a left-hander to the mound every fifth day keeps opposing managers and batters honest, and helps limit the damage wrought by opponents' left-handed sluggers.
Or, as the Dallas Morning News' Tim Cowlishaw wrote last Saturday,"There's no question that having a left-hander in the mix is ideal. Going into series after series with three right-handers throwing cut fastballs ... makes Texas too predictable." (In fact, the scenario bothering Cowlishaw is likely to unfold regardless: Feldman, Lewis, and Hunter all feature cut fastballs, and McCarthy recently added the cutter to his arsenal.)
The importance of a left-handed starter is conventional wisdom -- but does it bear out? The examples we started with might suggest not, but they also come off as so much anecdotal evidence. Can we say anything more certain about the Rangers' 2010 starting situation, without resorting to such questionable comparisons? It's relatively straightforward to measure how well an individual starter is performing, but how should we measure the reciprocal interactions between a lefty starter and his right-handed colleagues (or, for that matter, the bullpen)?
Coming up with a definitive solution to the southpaw scenario is tricky. There's no obvious way to pull what Tom Tango calls a WOWY -- a "With or Without You" analysis, which compares outcomes in the presence of a factor, and then in its absence -- within a rotation or pitching staff. And there are too many confounding variables to cleanly analyze the effects of handedness in rotation composition across teams and seasons.
Resigning ourselves to these facts, what we can do is to examine the case for a lefty on its merits. Let's take a look at how the Rangers' four candidates for fifth starter have performed versus right- and left-handed batters over their careers. In the table below, we see selected platoon splits for Harrison, Holland, McCarthy, and Wilson. Juxtaposing Wilson's numbers with the rest is not exactly an apples-to-apples comparison, of course, since he's likely to fare differently as a starter than as a late-inning reliever. Still, these data give us a point of departure:
Harrison and Wilson both show fairly distinct platoon splits in these skill stats, which are the best approximations of pitchers' performance independent of defense and luck. In almost every category, Harrison and Wilson performed better against lefties than versus righties.
McCarthy was fairly consistent across the board -- his numbers aren't much worse (or, depending on your level of residual resentment, better) versus left-handed batters than against right-handed batters. In fact, his career xFIP shows a slightly reversed platoon split.
Holland's stats are more of a mixed bag. There are signs that in his rookie season, he struggled more versus left-handed than right-handed opponents. But the relatively small differential in Holland's xFIP platoon split is encouraging. In 2009, he was unusually prone to serving up gopher balls to right-handed batters on fly balls (and those batters hit lots of fly balls off him). xFIP accounts for this higher-than-expected HR/FB ratio against righties, which is likely to regress toward the league average of 10-12 percent in 2010.
All in all, the numbers suggest that Harrison and Wilson are good to go against lefties, but could labor against righties. They imply that McCarthy probably is what he is -- and that's roughly the same against hitters on either side of the plate. And they confirm that Holland could very well be in for a breakout season, especially if his long-ball tendencies normalize. But even these observations still don't provide as much clarification as we might like about the importance of including a left-hander among Texas' starters.
For more, let's take a look at the platoon splits for the two right-handed starters Texas expects to lead its rotation in 2010: Scott Feldman and Rich Harden. We have both Feldman's career and 2009 numbers listed, since considering his 2009 breakout season separately seems useful:
The small differences stand out. Neither Harden nor Feldman (especially in 2009) have been hit especially hard by lefties, even in comparison to Harrison, Holland, and Wilson. It's worth noting that the cutter Feldman broke out last season -- according to FanGraphs, the most effective pitch of its type in the majors last season -- was noted to be especially tough on left-handed batters. Also notable: Feldman's relative success against right-handed batters in 2009 may have been misleading. Only six percent of the fly balls Feldman gave up to right-handed batters went for home runs, which is much lower than expected – and, as a result, so were his HR/9 and FIP versus right-handers. (This is the opposite of what we saw above with Holland.)
What, in the end, do these two tables tell us about the Rangers' need for a left-handed starter entering 2010? First, they indicate that in Harden and Feldman, the Rangers have two right-handers who present distinct challenges to hitters of either handedness. This pokes a hole in the argument that Texas absolutely needs a left-handed starter to keep opposing lineups off balance. Second, it indicates that whatever Texas would stand to gain against left-handers by going with Harrison, Holland, or Wilson, it would likely give something away against right-handed batters. The Rangers have to ask themselves exactly how much of a penalty against righties they would pay in games started by any one of these pitchers. Given what Harden and Feldman already bring to the table, would that cost exceed the benefits relative to McCarthy?
There are no easy answers to these questions ... but here's one take. For the Rangers to maximize their chances to reach the post-season for the first time in over a decade, they need to break camp with the starters they believe will perform best from the word go. Maybe that means giving the fifth spot to Holland, whose potential and peripherals are undoubtedly promising. Maybe it goes to Harrison, boasting a newly svelte silhouette and spring success. Perhaps it's even Wilson's, given his talent and tenacity.
Handedness, however, shouldn't be a leg up for any of the trio -- not least because it's very difficult to believe the Rangers will rely on only five starters the entire season, especially if the Opening Day rotation features both Harden and McCarthy. The 2007 Jon Lester may prove an instructive example, after all. It's certainly not a stretch to anticipate Holland (or Harrison) starting the season in Triple-A, getting the call to replace an injured or ineffective righty, and giving the Rangers a second-half shot in the arm.
So if the club should decide that the best choice is McCarthy -- whose new mechanics and early results have impressed thus far, and who has an edge in experience -- so be it. Even if that means series after series of four right-handed starters throwing cut fastballs. Even if that means contention rides on five right arms as of April 5th. And if regular-season history then repeats itself, Rangers fans can party like it's, well, 1999.
[Editor's note: If you're wondering exactly who Josh Garoon is, click here to view the explanatory website announcement, and/or click here to view his bio. Josh, a post-doctorate scholar at Johns Hopkins University in Baltimore, makes his debut appearance in the BBTiA writing lineup today. Welcome aboard, Josh, and try to keep your on-base percentage up.]
Reader Comments (14)
How about 4 IP, 8K's & 2 hits? Does that make a Southpaw, Matt Harrison look like a better deal in the rotation?
Good article Josh. I agree, pick the best 5 SPs regardless of which hand they use.
Good stuff Josh, that was an awesome read.
I've been saying for awhile now, both here and on LSB that I'd prefer to see the 5 best pitchers we have break camp in the rotation regardless of handedness, options, or spring training numbers (unless those ST numbers are egregiously bad to the point of being a red flag).
In my mind one of those is Holland, but he unfortunately doesn't appear to be at the top of the Rangers list right now thanks to his early-spring setback (though perhaps the Hunter injury lends him some hope).
Good article, welcome aboard.
Is Hunter really the right choice? I like all four contenders better than him.
First, good article, and it would be a highly involved simulation process to determine the impact of a left-handed pitcher compared to a right-handed pitcher in that there would be the need to understand the transition probabilities of how teams fared in games in which a lefty started after a righty, a righty after a righty, and righty after a lefty. And that still wouldn't get at the question of replacing the actual configuration with a hypothetical one.
Unfortunately, such clear-headed, empirically based thinking probably doesn't do anything to convince those who believe in conventional wisdom.
However, I take heart in Wash's statement that he wants the top 5 in the rotation. It's too early for management to tip its hand on who the top 5 are, but I think Harrison isn't quite there, as he hasn't received any starts.
One of the reasons I think we haven't seen Harrison with any starts is the issue of experience. CJ and BMac have many, many more innings, and I think that plays heavily into Wash's Gut.
What if the Rangers' scouts and FO think the best 5, based upon recent history of offseason work and spring training, Harden, Feldman, Lewis, CJ, and Harrison, but Wash disagrees and feels BMac should be in Harrison's place?
I wonder who wins that discussion. The FO could say, we disagree, but it is your team and you need to have the utmost confidence in it. If it turns out to not be the case, it's your butt that will be shown the door. The FO could also say, we're going with the consensus opinion here, and BMac is not one of our top 5 and will not be in the rotation. I tend to think they would do the former rather than the latter.
And, so, here's where I think the rotation and bullpend is headed. Hunter's injury could significantly change the equation, and I really think CJ was one outing away from winning a spot before the injury, because I think the organizational philosophy is to have pitchers who can get as many outs as possible by K's and GB's. If the injury keeps Hunter from consideration in the opening day rotation, I think the Rangers are on a trajectory for the following setup:
Harden, Feldman, Lewis, CJ, BMac
Nippert, Harrison, O'Day, Ray, Oliver, frankfranKKKKK
I think BMac has a tenuous hold on the fifth spot, based largely upon his experience, which is the approach taken last year with Benson/Feldman. And, I think the Rangers FO might have a CJ career path in mind with Harrison. That is, let Harrison refine his craft, which has suddenly shown a lot more promise, in the bullpen and eventually move him back into the rotation.
That leaves a AAA rotation of:
Holland, Feliz, Hunter (rehab), Moscosco, Kiker
While it appears to me, based upon past FO decisions and current performance, that this is the likely final mix, I still think a trade or two is very likely. At the very least, I think Mendoza will be traded, and I still get the feeling BMac might as well, but it would have to involve a significant ML piece. I have a hard time imagining what the return on BMac would be and who would need to trade for him, which makes me much less confident about a trade involving him compared to Mendoza.
One other thing....
If Hunter's injury doesn't keep him out of opening day consideration, I think BMac's experience keeps him on the team for at least a month, unless the Rangers are able to find a trade partner. And, I think the Rangers would be willing to take less than ML talent in a BMac trade under this scenario. If the could get a couple of intriguing A-ballers that need some development time (say guys who are just off the BA top 100), I believe they believe they have enough depth to take that risk. I don't know where they would put those A-ballers, assuming they were pitchers and not hitters, but I think they would bite on such a deal in the event that Hunter is healthy.
Assuming no BMac trade and a healthy Hunter, my current guess of the opening day pitching staff, and probably the staff for about the first month, is
Harden, Feldman, Lewis, CJ, Hunter
Nippert, BMac, O'Day, Ray, Oliver, frankfranKKKKK
AAA:
Holland, Feliz, Harrison, Moscosco, Kiker
With a trade of BMac, I would think Harrison takes BMac's place in the bullpen.
Great analysis Josh.... we are going to love your contributions. Thanks for separating Feldman's 2009 season from the career numbers; IMO it offers a better projection of his 2010. I am wondering what you think about Harrison's spring training numbers indicating a watershed for him, especially his domination of hitters on both sides of the plate?
Congrats on joining the team Josh! I enjoyed the first article. I agree with your findings (at least for the most part). Being one myself, I am a little biased towards southpaws; however, I've always believed in filling your rotation with your five best starters. With the Rangers depth right now, I would rule Tommy Hunter out for the opening day rotation. For the other two starting roles it just so happens that I think our three best candidates are lefties. Personally I would throw Holland into the third slot in the rotation; It would show confidence in Wonderboy and take a little pressure off of Lewis. For the last slot, I'd leave it up to Harrison and Wilson. Right now it would be a coin flip for me, but if pressed into a choice, I'd go with CJ. A lot of us seem to only remember the bad outings for CJ, but those were usually on no rest or only one day. On three or more days of rest, CJ was virtually unhittable. Can he keep that up for 6-8 innings? I have no clue, but I'd like to see.
Good to read something that's actually baseball related....
Tough decisions for Rangers' management. That's for sure. Even though C.J. has done well as a starter this spring, I'm thinking it would be best for the TEAM if he was returned to the bullpen as a late inning guy.
I like Holland and Harrison for the final rotation spot....How do you not select Harrison after the performances he has shown? On the other hand, I think the Rangers, and myself, would love to see Holland take some progressive steps against Major League hitters this season. He seems to be poised to do just that. I don't know what is going to happen but I'm very excited to watch how the Rangers' season will unfold for the first time in a long time.
Thanks to all for the warm welcome and great comments. Some quick responses:
Hunter's injury certainly throws a wrench into my scenario. Like badspellr and Dave H, I'm not quite as high on Hunter as some ~ but I do think a Rangers organization expecting to contend this season would've had to think had and long (especially given a rotation this uncertain) about not staying with a likely innings-eater who's had major-league success. Wilson and McCarthy should be searched for a voodoo doll.
rooster, McCarthy having a great start to the season and building up his trade value would be a beautiful thing ~ but would that sort of performance, accompanying early success team-wide, force Daniels to think twice about trading him? The Harrison-to-the-bullpen idea's an intriguing one, worth more consideration. texasliftr, Carl, Cliff, I love how Harrison's performing, but like Jon Page, I'm wary of making too much out of Spring Training performances. (See: Kata, Matt.). And I think folks forget Harrison's just 14 months older than Holland.
And speaking of Holland: He's my bet to be the Rangers' version of the 2007 Jon Lester. And ain't nothing wrong with that.
If they are going to go with the 5 best starters they have, then Tanner Scheppers should be in there as I have stated since November.
Oops, welcome aboard, Josh. You joined the bes and you are great cmpany with the writers here. Looking forward to more good stuff.
Add my welcome aboard, as well as a rousing well done for your first effort. As an old school baseball guy (read older, unfortunately not necessarily better) I am less a consumer of all the statistical crunching but do appreciate the effort and understand the value that it has in bringing a different breed of fan to the great game. And that's a good thing.
What's not mentioned is this ballparks' tendency to favor lefties, particularly GB inducing lefties. I've got no issue with McCarthy, (Danks trade? It's water under the bridge) but in this park, with these particular arms, and all else being nearly equal, I'm going with the GB lefty over the flyball righty until performance dictates otherwise.
Best of luck going forward.
Weirdly enough, the 2 righties battling for back end spots, performed better at the Ballpark
BMac WHIP: Home 1.087 Away 1.597 ERA: Home 3.13 Away 5.96
Hunter WHIP: Home 1.216 Away 1.481 ERA: Home 3.36 Away 5.01
Conversely, Holland- WHIP: Home 1.691 Away 1.277 ERA: Home 7.12 Away 4.98
Those are 2009 numbers, not sure if they're outliers, or the Ballpark advantage vis-a-vis handedness is a myth.
Anyways, great work Josh. Best 5 works for me.