Monday, April 19, 2010
BP: More on power rankings
By Jay Jaffe, Baseball Prospectus
For the past six seasons, I've been charged with generating Baseball Prospectus' power rankings via the weekly Prospectus Hit List. Rather than a subjective evaluation, the rankings are based on the Hit List Factor, an objective formula I devised that incorporates a team's actual winning percentage as well as a variety of Pythagorean-based winning percentages from our Adjusted Standings page. (Such factors included are runs scored, runs against and adjustments for park and league.) While still granting some credit for good ol' W's and L's, the idea is that the predictable relationship between a team's runs scored and runs allowed -- and the elements that go into those totals -- is more fundamentally indicative of team quality. Keep this in mind as you read today's power rankings, which are a snapshot of today, not future performance.
Here's a quick look at some of the teams with the largest gaps between their actual winning percentages and their league-adjusted Hit List Factors -- that is, teams either much better or much worse than they appear in the standings.
Better than they look
Orioles (2-11, .154 winning percentage, .361 Hit List Factor)
Though PECOTA projected the Orioles to win 78 games, they're off to a wretched start, in part because they're 1-4 in one-run games. Closer Mike Gonzalez is one culprit; he turned two save opportunities into losses before going on the disabled list. Frequent success or failure in one-run games can distort teams' winning percentages considerably; teams that exceed their Pythagorean projections tend to get there by winning many close ones while losing a few blowouts, and vice versa.
White Sox (4-9, .308 winning percentage, .483 Hit List Factor)
Despite cumulatively outscoring opponents by one run (49 scored, 48 allowed), the Sox lost two-thirds of their first 12 games. Their run differential is distorted by last week's 11-1 drubbing of Toronto, and to a lesser extent by their 6-0 opening day win over Cleveland. Beyond those, they're 2-6 in games decided by two runs or less.
Red Sox (4-8, .333 winning percentage, .485 Hit List Factor)
Projected by PECOTA to win an MLB-best 95 games, the Sox have lost three of their first four series, including two to the Yankees and Rays. They're 2-5 in games decided by two runs or less, with an offense that has basically been league average (.262 True Average), and a rotation that's been shaky, with just five quality starts in 11, none by Jon Lester or Clay Buchholz.
Worse than they look
Pirates (7-5, .583 winning percentage, .393 Hit List Factor)
PECOTA pegged the Pirates for an NL-worst 70 wins, but instead they're 6-5, including 5-0 in games decided by one or two runs. Considering that opponents were batting .294/.371/.500 against their staff coming into Sunday, something's got to give, with the historical deck stacked against a team with 17 straight losing seasons.
Rays (9-3, .750 winning percentage, .614 Hit List Factor)
We projected the Rays to win a handsome 92 games, not a sizzling 117. They're 5-1 in one- and two-run games because of extraordinary clutch success by their otherwise underperforming offense; they've hit .344/.421/.547 with runners in scoring position, .223/.272/.363 in all other situations.
Nationals (6-6, .500 winning percentage, .420 Hit List Factor)
Over the past three years, Livan Hernandez put up a 5.45 ERA and allowed more hits per nine (11.5) than any other pitcher. He's tossed 16 scoreless innings thus far, while the rest of the Nationals' rotation had yielded an 8.92 ERA -- and that was before Jason Marquis' seven-run, no-out debacle Sunday. 'Tis the season when such anomalies abound; check back in a few weeks and you'll find Livan's corpulent ERA more closely resembling his mates', sinking the Nats below .500.
Jay Jaffe is an author of Baseball Prospectus.