Sunday, April 13, 2008


Please excuse the title's political allusion but I couldn't resist.

The fierce debate that has divided Yankees fans and the baseball cognoscenti who cover the team recalled to mind the chasm between the "nattering nabobs," liberal and conservative, who wanted to impeach Bill Clinton and the wide swath of the American public from whom the President's escapades evoked little more than a shrug.

Of course the gravity of their consequences couldn't differ more. Then again, to the Yankees franchise, perhaps they compare after all. In 2008, the Yankees Front-Office is unlikely to face a dilemma more momentous for their fate than whether to keep Joba Chamberlain in his current role as the Yankees redoubtable 8th inning set-up reliever or whether to return him to the starting pitcher he once was, the role in which he awed the Yankees' organization last year and rocketed through their farm system.

Almost the entirety of the reporters, columnists, radio hosts, and ex-athletes turned television pundits who cover the Yankees favor leaving Joba in his current role as an 8th inning setup reliever, at least for the 2008 season. To name just a few who advocate as much: WFAN's Pope Mike and The Idio-Savant, John Heyman, Sweeny Murti; the YES Network's and 1050 Radio's Michael Kay and David Cone, Newsday's Jim Baumbach, The Record's Bob Klapisch, and former baseball players ranging from Goose Gossage to John Kruk, Fernando Vina, Eric Young and the rest of the punditocracy Red Sox Reich's House Organ, ESPN, employs.

Their argument proceeds something like this. Baseball increasingly has become a sport of specialized roles. Organization confine starters to pitch counts and innings limits. The complete game has vanished. And a "quality" start now consists of little more than six innings pitched surrendering three or less runs. Meanwhile, closers rarely pitch more than the 9th inning and the rare out or two in the 8th.

As a consequence, the pivotal moment in most games occurs in the 6th, 7th, and 8th inning after the starter has exited but before the Sandman has entered. (Naturally, the experts don't adduce statistics to demonstrate this point. In the meantime, the fan awaits a Jamesian study that derives a metric that identifies the game's watershed inning.)

Indeed, the talent margin between a league's best team and its worst often hinges on its middle relief. Teams that can resort to dominant relievers in the 7th and 8th innings, typically, are the teams that play in October: e.g, in 2007, the Angels with Scott Shields and Justin Speier; the Indians, with Betancourt and Perez; the Red Sox, with Okajima; the Rockies, with Herges and Fuentes. Teams like the 2007, Devil Rays, Orioles, Rangers, and Tigers (without a healthy Zumaya) in contrast, lost innumerable games in the late innings precisely because they lacked one or more consistently effective relivers who could preserve leads late in games.

Enter Joba. In the expert's view, he delivers the shock before the Sandman registers the kill. Indeed, four of the six Yankees wins to date have administered the Joba-Sandman prescription for throttling opponents' late-inning rallies. Thus the nostalgia it evokes in the experts who never tire of hearkening to 1996 to remind us that the Yankees haven't possessed this formidable a late-inning tandem since Rivera and Wettleland (who they conveniently forget Yankees fans called Sweat Land that year.) But they have a point: middle relief, the 7th and 8th innings in particular, has been a Yankees' Achilles Heel since 2001 and the swan song of Nelson, Stanton, Mendoza.

Moving Joba into the starting rotation, in their view, then is suicidal. It leaves the Yankees with two equally unpalatable alternatives, the one more noxious than the other: either (i) assigning the 8th inning role to Farnsworth or Hawkins, two pitchers who in this still nascent season already have demonstrated they can't handle the role (Farnsworth of the 2.00+ WHIP and Hawkins of 15.75 ERA) or (ii) entrusting the job to Ohlendorf, Bruney, Albaledejo, Veras, or one of the many relievers untested in the role. Of course, before last year, Joba himself was untested in this role as well. Actually, Joba not only had not navigated the shoals of a precarious 8th inning lead before, he hadn't ever worked as a reliever. In fact, he hadn't ever pitched in the major league in any capacity, besides.

Accordingly, a little imagination or exercise of logic might have prompted the experts to wonder whether there isn't someone else in the Yankees' system poised to rehearse Joba's trajectory before they rendered their verdict that moving Joba was tantamount to a death-wish. Alas the experts never do. The reason is because the great majority of New York's baseball media -- writers like Ken Davidoff and Pete Abraham genuinely interested in minor league baseball are the exceptions that prove the rule-- don't bother to familiarize themselves with the Yankees (or the Mets, for that matter) farm systems.

People like Francesa, Russo, Kay, and many of the senior writers, know the names, perhaps, of a few choice prospects but little else beyond that. They couldn't distinguish Steven White from Steven Jackson. They can't identify the pitches in Alan Horne's repertoire. They couldn't name any of the Yankees' elite pitching prospects recovering from Tommy John surgery. More damaging still, they don't know whether if a prospect with the composure and stuff to parallel Joba's remarkable ascension in '07 currently exists in the Yankees system. But that hasn't deterred them, nonetheless, from rendering judgment with a vehemence in direct proportion to their ignorance.


Ask a Yankees' fan, on the other hand. Consult one of the many blogs operated or frequented by the Yankees zealots, like Lohud or NoMaas, and you not only can obtain a quasi-professional scouting report on the top 25 Yankees prospects but an authoritative projection of their likely success. Question them about Joba and how he most profits the '08 Yankees. And the answer you receive is almost univerally the diametric opposite. Joba belongs in the starting rotation as soon as his innings limits allow. That is, at the first juncture in the season Joba can pitch 6 innings a start every fifth day through the schedule's final game (and perhaps October) without exceeding 150 total innings (accumulated in both his roles), Joba should start.

To them, it's axiomatic. Why confine a pitcher to 3 - 5 outs per game as an 8th inning reliever when he can provide you 15 to 21 outs per game as a starter. Why waste a commodity so scarce and precious as a pitcher with a repertoire of 4 above-average pitches-- two superb pitches, his fastball and slider, a third very good pitch, his curve, and a fourth pitch, his changeup, that's above average-- when the Yankees can draw from the surplus of hard-throwing relievers with which the organization abounds to replace Chamberlain.

For the setup role, the Yankees could call upon Ohlendorf or Bruney from the current major league roster, or Mark Melancon, Humberto Sanchez, Alan Horne, Jeff Marquez, or Jose Veras anyone of whom has at least a dominating fastball and another plus pitch besides. Only Joba however has four plus pitches. What's more when the Yankees need to turn to a sixth starter, as they will at some point, because of injuries or inning caps, they have no one to throw every fifth day who can duplicate Joba's talent, no other pitcher with the potential to develop into the staff's ace or to join Wang in that role. To match Beckett-Buchholz, Halliday-Burnett, King Felix-Bedard, or Sabathia-Carmona, the Yankees have one option and only one, Wang-Joba.

The precedent for transforming Joba from the bullpen to the rotation mid-season is, of course, Johan Santana's similar trajectory in 2003, as Joel Sherman reported last Sunday in The New York Post ("Jo-Joba's Witness," April 6, 2008) Through July 11, 2003, Santana pitched in the Twins bullpen, accumulating 66 innings. From July 11, onward, Santana started. Over those 90.3 innings as a starter, Santana went 8-2 and catapulted the Twins from a 44-46 to AL Central Division Title, with Santana recording the only Twins win in the 2003 ALDS against the Yankees.


One way to quantify this potential impact is by examining Joba's likely contribution in each role through through the prism of WARP or (Wins Over Replacement Player), a statistic Pecota and Baseball Prospectus utilized to measure a player's contribution to his team.

Assume Joba could throw approximately 150 innings this year in a mixed role of starter and reliever and approximately 70 innings in his role as set-up man, about the league average for a effective set-up man or closer on a winning team. Okajima threw 70 inning for the Red Sox in 2007; Scott Shields, 77 innings for the Angels; Casey Janssen, 72.7 for the Blue Jays; Rafael Betancourt, 79.3 innings for the Indians. (The median number Mariano Rivera has pitched the last four seasons is 75 innings)

The mixed role would confine Joba to throw approximately 110 innings as a starter, about 20-30 innings in the bullpen and 20-30 innings in the minors while they extend his arm.

At their best a dominating set-up man contributes under a 4.0 WARP. In 2007, Okajima, Shields, and Janssen amassed the following stats:

  • Betancourt-- 79.2 IPs, 1.48 ERA = 4.0 WARP
  • Okajima-- 69.0 IPs, 2.22 ERA = 3.3 WARP
  • Janssen-- 72.7 IPs, 2.35 ERA = 3.1 WARP
  • Shields -- 77 IPs, 3.86 ERA = 2.9 WARP

Now compare the difference to a starting pitcher, even one who throws half a season. The sample size for comparison is small obviously because few starters pitch less than 100 innings over a season. But in 2007, Jessie Litsch a rookie starter for the Toronto Blue Jays, threw 111 innings, about what the Yankees could expect of Joba once they transform him into a starter.

  • Jessie Litsch-- 111 IPs, 7-9, 4.49 ERA = 3.5 WARP

The ramifications of the above comparison are momentous. A mediocre starting pitcher, in just half a season, contributes about the same number of wins over a replacement player as some of the league's best set-up men over a season's entirety. If Joba pitches slightly better than Jessie Litsch did in 2007-- and there is great reason to expect his totals as a starter would excel Litsch-- he would contribute about as much to the Yankees success in 2008 than if he duplicated Rafael Betancourt's success last year as the AL's best set-up man.

Or consider Joba's likely WARP if his statistics as a starter parallel the dominance he's demonstrated as a reliever. Witness the impact in 2006 of Jered Weaver for the Anaheim Angels.

  • 2006- Jered Weaver, 123 innings, 11-2, 2.35 ERA = 5.9 WARP

Compare Weaver's WARP to the Angel's set-up man Scot Shield's WARP in the very same year.

  • 2006- Scot Shields, 87.7 IPs, 1.07 WHIP, 2.87 ERA = 4.9 WARP

The statistics would seem to bear out the Yankees' fan's contention that even the best set-up man contributes less to a team's success over the duration of an entire season than an above average starting pitcher will through one-half a season. A verity that the salary differences between dominant set-up men and even mediocre starter reflects.

The question is whether the Yankees Front-Office, when deciding Joba's fate, will ignore the onslaught of "expert" opinion and base their decision instead on the layman's impassioned reason.

No comments: