keyboard_arrow_uptop

There weren't many defenders of the Jason Bay deal when it happened. Bay was a very good player, coming off a 5.9-WARP season, but to justify his four-year, $66 million contract he'd need to continue to produce about four wins per year. That seemed just a bit too much for a player who was an excellent hitter but (by most metrics) lousy on defense. Maybe three wins per year, but not quite four. That's why the Red Sox had refused to offer more than four years and $60 million. Those extra $6 million that the Mets offered sure looked like an overpay. 

That's how precisely we tend to analyze these deals. If we set the price of a win at, say, $5 million, then a move that promises wins for $4.5 million look pretty good and moves that promise wins for $5.5 million look kind of bad. Sure, Jason Bay is good, and every team would love to have him, nobody denies that; it's just those extra $6 million (sometimes extra $1 million, sometimes extra $30 million) that really makes that signing look bad.

Of course, now we know that Jason Bay was worth nothing, and that no contract would have been good; the Mets essentially released him today, letting Bay walk away in exchange for a bit of money deferred. Despite the precise talk we use to evaluate player contracts when they're signed, it's rare that players' actual performance lands in that margin between pretty good and kind of bad. Here are the 10 top free agents that winter, according to MLBTradeRumors' rankings at the time: 

  1. Matt Holliday
  2. John Lackey
  3. Jason Bay
  4. Chone Figgins
  5. Randy Wolf
  6. Andy Pettitte
  7. Jose Valverde
  8. Marco Scutaro
  9. Adrian Beltre
  10. Rich Harden

And here's how much each of those players has been paid per win since then. The blogosphere-accepted rate for a win at the time was around $4 million.

  1. Matt Holliday: $4.6 million per win*
  2. John Lackey: $24.1 million per win*
  3. Jason Bay: $55 million per win
  4. Chone Figgins: ∞ ($27 million; no seasons above replacement level)*
  5. Randy Wolf: $6.9 million per win
  6. Andy Pettitte: $6.9 million per win
  7. Jose Valverde: $11.5 million per win
  8. Marco Scutaro: $2.6 million per win
  9. Adrian Beltre: $1.2 million per win
  10. Rich Harden: ∞ ($7.5 million, sub-replacement level)

It's perfectly fair to ding a team for overpaying. If a player's market value should be $8 million, and a team pays him $10 million, that's $2 million that it couldn't spend anywhere else. Those $2 million are the proverbial 2 percent, and they matter, and as far as what teams can actually control they might be all that separates good organizations from bad. That's what we talk about when we talk about the day's signings.

But this offseason, you're going to read a lot of analysis about player contracts, and which ones are good and bad based on the narrow sliver of space between projected dollars per win and the market rate for wins. The truth is that that narrow sliver of space doesn't often come into play for expensive free agents. Most players don't end up all that near their mean forecast, and most teams don't actually get the player they're expecting. The real winner is the team that signs an unexpected MVP candidate; the real loser is the one who signs a very good player and watches him turn into nothing almost immediately. And predicting those is beyond most of our abilities.  

*player still under contract