Multi-year Deals For Veterans: Do They Work?

The signing of John McDonald to a two-year deal means that Arizona now has on their books the two oldest players in the majors, in McDonald and Geoff Blum, to have been signed to contracts longer than one season. As last year with Blum, McDonald's signing provoked a fair bit of discussion - with the nay-sayers pointing to Blum's basic zero value to the team in 2010 as evidence against this kind of deal. I thought it might be interesting to take a look back at the history of players that age being signed for more than one year. Does that kind of contract work out for the team?

The following table lists all players, aged 37 or older, who have signed multi-year contracts since the 2006-07 off-season. It lists their age (on this date in the year in question), how much the contract was for, their production (all figures in this piece are bWAR), and what that production is valued at, using a straight rate of $4m per WAR to simplify things, given the five-year range covered.

The first thing to note is that multi-year deals for those of this age are really, really rare, especially for position players. During the previous five off-seasons, only three such men aged 37 or older (on this date in the year) received contracts for longer than a single campaign. None of them have come close to living up to the cost. In total, they have been paid more than $25 million, and in a total of five player-seasons [Geoff Blum still having his second one to go], produced a total of 2.4 WAR.

Admittedly, expectation are way lower for John McDonald than Frank Thomas or Ivan Rodriguez, two possible future Hall of Famers. But if John managed to reach even 0.8 WAR over the next two years, thereby justifying the $3 million cost in that time, he'll be the only position player his age to do so in recent baseball history. Things aren't much better for the nine pitchers on the list: Arthur Rhodes is the sole case whose production justified the cost, though if Rivera's 2012 is anything like his 2011, that deal should work out for the Yankees too.

In general, however, it's not too much of an exaggeration to say that old players...well, largely suck. Of the position players aged 37 or older who had any plate appearances in the majors from 2009-2011, more than 42% were replacement level or lower. That compares to 35.6% across all ages, but let's see exact;y how that number compares for other age brackets in the majors over the past three seasons.

It's an interesting chart, at first suggesting we need to stuff the team with those not yet old enough to drink [think of the money we'll also save by replacing playoff champagne with sparkling apple-juice!]. However, the fact that the 21 and under crowd are the least likely to blow chunks is easily explained - if you're playing in the majors at that age, you're a very special talent, making failure less likely. The names on the list include the likes of Mike Stanton, Justin Upton and Jason Heyward, After that, the level of suckage increases sharply, as the less talented rookies arrive, and experience their share of struggles.

The percentage who are replacement level or worse declines for every age group thereafter. That's not much of a surprise either, due to the self-selection involved. If you're no good in the majors at age 26, you probably won't be in the majors at 29. However, that appears only to go so far, and once a player reaches his late-30s, time appears to catch up, fast. This is nicely illustrated - and somewhat corroborated - by the following graph, which comes from a study on player aging done by Jeremy Greenhouse last year. As you see, once you're past 35, players tend to decline in value quite rapidly..

The growing old not-so gracefully applies not just at the bottom end of the spectrum. Over the same time-frame (2009-11), 219 players put up seasons at or above 3.0 WAR: a mere three were 37+: Manny Ramirez (2009 LAD); Chipper Jones (2010 ATL); and Jim Thome (2010 MIN). And at least one of those numbers probably deserves a big, fat, well-fertilized asterisk beside it - perhaps copied and pasted from the ones beside the trio of double-digit WAR seasons posted by B. Bonds, Esq, at ages 37-39 from 2002-04.

You might think reduced playing time is a factor there, but the median number of games played by the 64 veterans was a respectable 90 games, with about 290 plate appearances. While no-one expects a three WAR season from McDonald, that's still significantly more playing time than he will likely get, in his role as a back-up - over the past four years, he has averaged only 46 starts per season and 193 PAs. However, the Greenhouse study quoted above suggests that good defenders and bad hitters tend to age better than their opposites. On that basis, if you're going to give a 37-year old a multi-year contract, John McDonald might be the best candidate for the position.