Once upon a time, or so the story goes, Billy Beane’s A’s were a motley collection of unwanted, damaged, washed up, one-dimensional players who played for pennies on the dollar relative to their potential production.
From Jim Mecir to Jeremy Giambi to Scott Hatteberg to Matt Stairs, the early-period Moneyballers were signs, primarily, not of Beane’s miserliness nor of his desire to field a beer-league softball team.
Rather, they were all players who could make some specific, productive on-field contribution (in many cases, one and only one kind of contribution) while being markedly deficient in some specific other baseball skills (in many cases, multiple marked deficiencies).
This, to me, was always the more compelling story than Michael Lewis’s abstruse, economically filtered “undervalued” explanation.
Beane was able thereby to assemble a complete and complementary team – not, as teams traditionally were built, on the basis of gathering as many complete “five-tool” individual players as possible and filling in the gaps as necessary, but backwards: filling as many lineup, defensive, and pitching gaps with cheap, one- or two-tools players first, and then rolling the dice on a couple of more complete players becoming available via draft or trade.
We’ve all heard variants on this tale over and over again. But I think that this specific model of roster-building has gone somewhat under-remarked.
And I think it’s a successful model that Beane has been moving away from – and moving away from unnecessarily.
[Continued after the jump.]
There’s certainly a degree of truth in “the market having caught up to Beane,” in that certain skills once undervalued by the talent market (e.g., OBP for hitters and GB% for pitchers) are now trading at a premium, with even some one-dimensional players out of Beane’s reach (such that a Kevin Youkilis is now a relatively untouchable commodity).
But more and more, Beane the last several years has been building the core his roster around (and committing long-term dollars to) a mix of putatively “complete” players (Chavez, Swisher, Ellis, Harden, Street) and traditional old-school intangible types (Kotsay, Kendall) – and only then looking to fill in the gaps with the one-dimensional castoff role players.
Ironically enough, it’s still been those one-dimensional castoff role players (whether they be OPS machines like Thomas, Piazza, and Cust, or complementary .300 line-drive hitters like Stewart and Payton) who have continued to represent Beane’s greatest success stories.
As I said above, I don’t think that the conventional story of the market catching up to Beane’s wily evaluative ways is entirely on-target.
I also don’t think that it’s necessarily entirely due to institutional hubris, of trusting The System and its heuristics over facts on the ground.
Instead, I think it’s more a case of the System itself evolving in unforeseen and uncontrolled ways: If the System is, in the eyes of its agents, designed to focus on undervalued skills and players, then the System will be allowed to follow undervaluations into potential blind alleys.
And with no adjustments to the System to accommodate for when undervaluations lead to either players too similar to be complementary on the field, skills not powerful enough to generate requisite on-field results, or deficiencies inherent in the undervaluations but not reflected in rate projections, the relatively simple System will spawn a host of complex problems.
The roster may be assembled logically, and that logic itself may seem a simple and straightforward extension of the steps that have preceded it – but the System itself may break down or run out of control.
In short, we’ve gone from the Island of Misfit Toys to Jurassic Park.