Why Do GMs Hold On to Albatrosses?

A report suggesting that the San Francisco Giants were considering releasing Barry Zito appear to be bunk, according to Andrew Baggarly at MercuryNews.com. This isn’t particularly surprising or necessarily even newsworthy. I only bring this item up because it has resurrected my interest in a particularly intriguing baseball topic (at least to me): the reluctance of a general manager to release a replacement-level player with an albatross contract.

The key concept in these considerations (and all else regarding baseball transactions) is that player contracts are guaranteed. So, even with terrible players like Jeff Suppan or Oliver Perez, the team gains absolutely nothing by cutting a player if a suitable replacement can’t be dug up. For this reason, the idea of cutting Barry Zito is silly. Unless something drastic happened between the end of last season and this very second, Zito looks to still have something left in the tank. He’s projected for a 4.23 FIP by Marcel and even accounting for age and decline, Zito’s median expectation should be somewhere between one and two wins above replacement. Unless the Giants find a willing taker for Zito’s contract (not happening unless Arte Moreno and/or Tony Reagins go crazy again), Zito is just part of one complete $110 million commitment the Giants franchise has already made.

There are the obvious examples like Zito, Vernon Wells, and Alfonso Soriano (among others) where the cost of the player is unfortunate and damaging to the franchise, but releasing them would simply be cutting off the nose to spite the face. But these are not the cases I find interesting.

Instead, the interesting cases are the players that hover around replacement level, or perhaps perform just a bit below. Not the Oliver Perezes of the world – he’s so bad that the roster spot could be used better with 100% certainty – but instead the Carlos Lees (or maybe the Chan Ho Parks circa 2004) grab my attention here.

In order for the release of the player to be worth it, purely from a wins standpoint, the team has to find a suitable replacement. This hypothetical replacement may or may not exist, via the farm system, or minor league free agency, or the rule V draft, or a trade for organizational depth. In the case of many minor leaguers, though, the extra seasoning from time in the minors would benefit the organization more than a marginal win gain over the poor, overpaid player. In the other cases, the organization would have to spend opportunity costs such as scouting time, a rule V draft pick, and any other costs that come from acquiring a replacement from outside the organization.

And this all comes before organization politics and other non-wins factors come into play. By releasing high-profile, high-earning players, the organization is effectively admitting that it made a mistake. Even if the mistake is obvious, admittance doesn’t tend to be a good look for organizations. Perhaps the owner isn’t willing to see such a big investment released with no return, even if that return clearly isn’t coming. And what if the player goes on to resurrect his career with his new team? That would be an unmitigated PR disaster.

These reasons help me understand why teams and owners tend to hold on to bad, near-replacement contracts for what seems like dear life. Although it seems like the risk is extremely low for such moves, the reward can be even lower and the risk is likely higher than it appears at first glance. Is this decision-making process among GMs completely rational? Perhaps not, but given the risk aversion many people express to begin with, I don’t find it particularly surprising.