a few months ago, Baseball Analysts ran a guest column by dan levitt analyzing terry ryan's performance as the minnesota gm. using win-shares as his unit of measure, levitt counted up --- one transaction at a time --- how much talent ryan brought into the organization (via trade, free agency, etc) and how much talent he let get away. then he added it all up and concluded that ryan's roster management netted the twins about 200 win shares between 1994 and 2005 --- about 18 win shares per year, which translates into an average of 6 wins per season. that's a phenomenal performance --- an 18 win-share player is a borderline all-star. in the comments section, somebody wondered how walt jocketty would fare in such an analysis; you'll find the answer right here.

but before i show you the bottom line, i have to address some problems i have with this methodology. first and foremost, it doesn't account for payroll. since walt jocketty had vastly more financial resources to work with than terry ryan, his decision-making process was completely different; he could afford to trade young players for walk-year stars such as mark mcgwire, jim edmonds, and scott rolen because he knew his owner would bankroll long-term extensions. ryan couldn't even consider trades of that type; he simply didn't have that type of option available. which leads to a second limitation of the methodology, ie it doesn't capture the ability to choose the best move from among a range of available options, which is perhaps the defining measure of a gm's performance. to take one example from walt's career: under this method, jock gets credit for netting 26.4 win shares for the organization via the signing of tino martinez in 2002. in my opinion, that transaction should show up as a debit on the ledger rather than a credit, because jocketty passed up other, better options --- e.g., he could have moved pujols from left field (his home at that time) to 1st base, then signed a free-agent to replace albert in left field. reggie sanders, kenny lofton, johnny damon, and moises alou were available that year; sanders and lofton signed for far less money than tino, alou signed for about the same in dollars and years, and damon signed for a lower annual salary but for one extra season. the most accurate way to evaluate that transaction is to say that jocketty only added 26.4 win shares, but this method doesn't reflect the fact that walt failed to maximize the use of the roster spot and the dollars. it simply credits him for acquiring 26.4 win shares.

finally, this method assigns the gm no credit at all for talent development --- ie, for players who come up through the farm system. i don't think that's fair. while the gm doesn't make individual decisions regarding whom to draft, when to promote them from A to AA, and so forth, he does hire the personnel who make those decisions, and he helps to set the organization's overall priorities and philosophy. he also sets the budget for player procurement at the draft, which as we have seen recently has a major impact on the draft results. and the gm can create or foreclose opportunities for young players. when jocketty brought in sidney ponson in 2006, he closed off an opportunity for wainwright / reyes; when he let supps and weaver and marquis go in 2007, he created opportunities for the same two pitchers. player development is an important part of a gm's responsibility, and this method simply ignores it; talents such as pujols, morris, and drew are treated as accidents, rather than as products of the pipeline the gm helped to create and manage.

levitt acknowledges some of these limitations at the end of his Baseball Analysts piece:

one final note to clarify for you before i show you these numbers: for each player involved in each transaction, this method counts career win-shares remaining. to return to the tino martinez example: when jock signed him in 2002, tino had 48.9 win shares left in his career; that number goes into the "to st louis" column. when tino was traded away in 2004, he had 22.5 win shares remaining; that goes into the "from st louis" column. the net is 26.4 win shares --- that's the number tino delivered while wearing a st louis uniform. got it? i hope so. here's how this method breaks down walt jocketty's tenure as the gm:

that's counting win shares through the 2007 season. i ran those numbers twice and got the same result. in 13 years as the gm, jocketty's transactions netted 725 win shares for the franchse --- an average annual gain of 56 win shares, or roughly 19 wins a season. now, before we get carried away with our praise of the departed gm, realize that the vast majority of that surplus --- a net gain of 532 win shares --- is accounted for by free-agent comings/goings, which are as much the owner's doing as the gm's. i'll also note that jocketty's stellar record in the trade market --- plus 234 win shares --- was built largely on 2 deals for players who were headed into free agency and only stayed with st louis because the owners ponied up for costly extensions (ie, they were de facto free-agent acquisitions). those 2 trades --- for mcgwire and edmonds --- netted 162 win shares for the cardinals. all the rest of walt's exchanges brought an aggregate of 72 win shares into the organization, which is still an outstanding performance (worth nearly 2 wins per year) but not quite so breathtaking.

now here's the next caveat: the ticker is still running on many of these transactions, and the drift is not in jocketty's favor. for most players, the effects will cancel each other out; each win share that jeff suppan earns from here on out will be counted both as a win share that jocketty gained (by signing supps in 2004) and a win share lost (by allowing supps to leave in 2007). guys like rolen and edmonds and eckstein fall into that same boat. but the talent walt traded away from within the organization --- that's gonna kill him. dan haren and daric barton will be adding to his debit column for years and years and years; there's a chance those two alone will wipe out his entire win-share advantage in trades. players like coco crisp and jd drew and placido polanco and even dmitri young and matt morris are still out there racking up win shares in jocketty's "from st louis" column while adding nothing to his "to st louis" total. of course, wainwright will help to counteract that, and whatever carpenter has left will also accrue to walt's credit, along with the win-share scraps kicked in by ludwick, looper, miles, spiezio and franklin. if there is a single, solitary win share left in mark mulder's arm, the greater walt jocketty's glory.

on thursday i'll take a closer look at this record, take it apart year by year and identify the most advantageous and disadvantageous individual transactions.