If Valentine had the 2004 roster, or the 2007 roster, he could make some of his questionable moves and the team would be fine. He doesn't have those rosters..
If Valentine had the 2004 or 2007 roster, the moves he's making wouldn't be all that questionable. That's why you can't separate the quality of the options from the quality of the decision-making process. In 2004, you could have ordered the starting 9 any way you wanted to and been able to defend it. In the 8th inning of a game in 2004, you could got to any of Timlin, Embree, or Foulke and defend it, regardless of the situation -- Do people really not understand the gulf in talent, sustained success, and experience between those three and even the best case scenario of Melancon, Morales, and Bailey?
What JakeRae is basically asking people to do is look at the splits of the hitters and pitchers and decide whether Valentine went to the option that minimized the probability of an adverse outcome ON PAPER, IN APRIL. The defense for what Valentine not using that objective function is that his task is figure out, fairly quickly, who his best pitchers are going to be IN THE PRESSURE COOKER OF OCTOBER.
Further, that job is made more complicated by a full 40-man roster, the very similar track records of a whole bunch of pitchers who have no minor league options, and the need to sort them out quickly before Hill and Matsuzaka and Cook and Miller need to be activated. If you don't accumulate information on the dreck, in order to figure out which dreck is slight less dreck, then you very probably will lose the wrong players when the time comes to waive somebody, or, in this case, 4 somebodys, before Memorial Day. For instance, Miller and Morales have incredibly similar tools and more similar production to date than some want to admit, while Thomas has good minor league splits. Before I jettison Thomas for Miller, who's frighteningly wild and inconsistent but has more upside, I want to see if Thomas can be a reliable loogy who doesn't threaten to walk 25 percent of the batters he faces--If Thomas is marginal but reliably marginal, maybe that's better than Miller's upside with a heavy downside. But, if I like Thomas, then before I waive Miller, I want to see if I'm getting at least the Boston 2011 version of Morales, or if I'm getting the version of Morales that Colorado waived because he looked a lot like Miller. And Morales, like Miller, also has upside, so I want to see if he can be more than what I saw in Boston last year.
And, for all the bowing done around here to major league equivalents (I think they're fine for hitters), they're just not that great for relievers. You can't decide that Justin Thomas will get major league lefties out because he got minor league lefties out -- the key difference between major league lefties and minor league lefties is that those minor leaguers are often helpless against lefthanders. A good AAA loogie is not necessarily going to be productive in the majors, but watching him in the minors or spring training is not going to give you any additional information. So, because you can't stash any of these guys in AAA (even though Thomas has an option, the 40-man roster is full), put someone else on the major league roster, and then wait around day after day for the perfect opportunitiees to use the guy in the majors and see what he's capable of, you have to make what spreadsheet readers call "suboptimal decisions." Given that the Red Sox are desperate for anything resembling quality and time constrained in their decision making, it may make sense to let Thomas face a righthander if the benefit is that you get to watch him against another lefthander -- as was the case in one of the instances people are pointing to as a "mistake."
The lack of good options, and the lack of obvious separation in talent between the bad options, means that "optimizing" in April entails maybe sacrificing wins now to end up with the best of a bad lot in September. If you have good options or clear bad options, you don't have to do that. That's why, again, the composition and quality of the roster is intertwined with the assessment of the decision-making process.
Edited by Plympton91, 22 April 2012 - 08:44 AM.