Last Thursday, the New York Yankees and the Baltimore Orioles began a "big" four-game series and so of course I wrote something about it. Because that's what we do; we write things about things. But of course that big series, like most big series, didn't actually decide anything; they split those four games, leaving everything exactly as it was, except kicked a bit farther down the trail to October.
As it happens, this week the Orioles have another big series, this time against the Rays. How big? With all three American League contenders having Monday off, here are your Tuesday-morning standings:
This series probably won't be decisive, either ... but the Orioles could just about kill the Rays' chances, for a division title anyway, with a sweep. And the O's do seem to have a small edge here; after all, this series is in Baltimore and they've got a one-game edge in both the divisional and Wild Card tables.
So why does Baseball Prospectus's give the Rays a 55-percent chance of qualifying for the postseason, and the Orioles just a 29-percent chance.
Well, you know (or think you know) ... Stupid run differential. The Rays have outscored their opponents by 85 runs this season, while the Orioles have been outscored by 29 runs. As you know, it's exceptionally difficult to go 78-62 while being outscored by 29 runs, and the Orioles have done it largely on the strength of their practically miraculous 25-7 record in one-run games.*
* The Rays are 20-24 in one-run games, the Yankees 17-21. Despite excellent bullpens. Apparently their bullpens just don't know how to win the close ones.
Now, when we point out the Orioles' run differential, some subset of the cognoscenti will make an astute observation: These aren't the same Orioles who were routinely getting blown out in the first half of the season. Our fellow enthusiasts might even choose a date in July or August, and show that the Orioles have outscored their opponents since then. Which is interesting. The problem, of course, is that the thinner you slice the data, the less meaningful that data becomes. So let me choose an arbitrary endpoint, with no agenda other than to choose an endpoint: July 1.
From the July 1 through Monday, the Baltimore Orioles have been outscored by seven runs. The Yankees, over the same stretch, have outscored their opponents by only 34 runs. Meanwhile, the Rays lead the pack with a +84 run differential. They, and not the Yankees or the Orioles, have pretty obviously been the class of the American League East. And that is why the Playoff Odds Report likes the Rays as much as it does; their players have played better than the Orioles' players for an extended period of time: three months, five months, one year, two years ... It really doesn't matter which reasonable slice you prefer.
I should make a distinction ... As I understand the Playoff Odds Report, the figures are not based on run differential, which would be a clumsy tool indeed. Rather, they're based on something far more fundamental: the players on the roster, and their projected performances based on historical performances. Granted, projecting the performance of players and especially pitchers is tricky business, especially when we're projecting only three weeks of performance. But it's the best we can do. Otherwise, we might as well simply throw darts at a board, and go with that.