our illustrious editor-in-chief provided the news yesterday, and bgh broke down choate's stats. but the form of choate's contract - 3 years, $7.5m - bears some examination in its own right.
to briefly reiterate some of the general points from yesterday, the good news is that randy choate is a bona fide loogy. he holds a career 2.49 FIP against left-handed batters. unlike most of the supposed "specialists" we've toyed with over the years, this is actually an excellent split. mariano rivera's overall career FIP is 2.75, for reference. facing left-handers, randy choate is as effective a pitcher as you'll find in baseball.
the bad news is that randy choate is a bona fide loogy, and there's a reason he's a left-handed specialist, not just a plain old reliever. against right-handers, choate has a career 5.57 FIP, which compares to ervin santana's overall performance last year (i.e., the worst qualifying starting pitcher in MLB). mozeliak should tell matheny now that he'll be fined $1000 every time that choate faces a RHB.
to turn to the overall contract situation, we should first acknowledge that we're starting from the right position: with an actually effective left-hander. it's much better to be wondering whether it's worth blowing a bunch of money on an effective specialist, than to be trying to figure out how paying $1M to a reliever who's merely okay against lefties and dreadful against right-handers makes any sense at all.
after acknowledging that he has been a good-to-great specialist, the most troubling fact for me is choate's age. formerly good (trever miller) and even formerly great (arthur rhodes) left-handed relievers have seen the bottom fall out of their performance abruptly in just the phase of life we're discussing. all the analysis on his recent stats won't tell us if he's going to suddenly become useless at 38 or 39.
projecting reliever value is extremely dicey, because of the sample sizes involved. choate tends to pitch a very small number of innings in a huge number of appearances. he has faced almost exactly 2 batters per appearance for the last 4 seasons. not coincidentally, very careful use of choate in tampa, miami, and LA has correlated with improved stats and value, as compared to his years in arizona and new york, where he frequently saw 3, 4, 5, or more batters per appearnace.
during that time of cautious usage, he's been worth about 0.3-0.4 WAR on average. if he stays healthy and effective for three years, he will probably be worth about one win total, give or take a couple runs. $7.5m is a modest overpay on pure theoretical WAR terms.
however, the question is then whether we should pay randy choate slightly more than that pure theoretical WAR calculation. the theory being that choate's highest leverage innings, usually thrown against left-handers will be his most valuable, while he'll lose value in the games we don't care about - blowouts on one side or another, where his performance won't affect the likelihood of the team winning or losing.
i suspect that argument is frequently made in a vacuum of evidence. choate, for instance,pitched to only 13 batters in 2012 when his team was winning or losing by more than 4 runs, versus 155 when the game was within 4 runs. incidentally, the OPS against him in those 13 PAs was .408, so it wasn't the case that he was putting all his bad pitching into isolated blowout appearances.
you see similar trends in other years. in 2011, he pitched to 26 batters when his team led or trailed by more than four runs, versus 77 when the game was within four runs. he allowed a .626 OPS to those 26 batters. in 2010, he faced 26 batters with more than four runs separating the teams, facing 161 batters with four or fewer runs separating the winning and losing teams.
so, randy choate wasn't pitching 50% of the time in blowouts: more like 10-20% of the time. while pitching in blowouts may diminish his WAR value somewhat, it's not enough to account for major shifts in value.
the other problem with slightly overpaying for choate is that his contract reflects no discount whatever for his age and the likelihood of decline or injury, which is pretty likely.
i started writing this leaning toward favoring the deal, but I think it's a modest overpay, in light of the significant risk of injury or decline. I will say that - given the likely budget space in years 2 and 3 of the deal, it's a minimal overpay. I'm not sure i want to get real exercised about a modest overpay. and I'd rather see us slightly overpay for an actual specialist than pay league minimum for a sub-replacement value leftie.
the more appealing prospect would probably involve going with an internal left-handed option or chucking the two-lefty plan entirely and putting 6 excellent RHP in the bullpen along with Rzepczynski.