May 6, 2009
Prospectus Hit and Run
Cinco do Samplo, AL
Drawing conclusions from one month's play is a challenge. PECOTA projections, run differentials, strength of schedule and batted-ball results all fit into what I'll call the Rorschach Test of Impending Correction: selectively viewed blots of data which can justify nearly every regression-to-the-mean scenario under the sun. We know that the Blue Jays were projected to win 75 games in a brutally difficult division, that they're nonetheless 18-10 thanks to league leads in Equivalent Average, Defensive Efficiency, and run differential, and that they've played one of the easiest schedules to date, so how seriously should we take them?
To grapple with the question of which results are significant thus far, I've called upon two tools to estimate final winning percentages. The first, likely more familiar to our readers, is derived via the PECOTA-based Playoff Odds Report. Each team's current record and third-order Pythagorean record are factored into a Monte Carlo simulation of the rest of the season, with their records regressing not to .500 but to their projected winning percentages. Though it doesn't account for big injuries that have happened since Opening Day, this method obviously incorporates a great deal of data regarding current rosters.
The second is the outcome of Rany Jazayerli's Royally-inspired three-part investigation into hot starts circa 2003. Via multivariate regression, Rany found that a team's final winning percentage after a given number of games could be estimated by Y = P + ((S - P) × (.0415 + (.0096 × G))), where P is a team's projected winning percentage based on a weighted average of three previous seasons, S is their current winning percentage, and G is the number of games thus far. It's a nasty-looking formula, but it's based on over 70 years of historical data.
Below are the two estimates for each AL team's final winning percentage based upon the results through Monday (tomorrow I'll hit the NL), sorted by the Odds method:
AL East Odds Hist Red Sox .595 .572 Yankees .594 .536 Rays .538 .486 Blue Jays .493 .558 Orioles .435 .429 AL Central Odds Hist Tigers .522 .502 Royals .494 .495 Indians .493 .474 Twins .474 .518 White Sox .462 .508 AL West Odds Hist Mariners .501 .486 Athletics .490 .462 Angels .490 .539 Rangers .451 .498
With no team bolting quickly from the gate, the key take-home point is relatively subtle: the hot-starting interlopers in the AL Central and AL West races, the Royals and Mariners, certainly have shots in their tightly packed divisions, while the Blue Jays' bid for contention is harder to take seriously.
The Royals boast the league's top run-prevention unit, thanks in large part to Zack Greinke (6-0, 0.40 ERA) and in spite of the decision to start Aruban knightmare Sidney Ponson while stashing Luke Hochevar at Triple-A. Ponson's disasterpieces aside (0-4, 7.16 ERA), the ranking is genuine, in that the Royals rank second in all three defense-independent categories (strikeout, walk, and homer rates) without a fluky batting average on balls in play. They have their problems, namely a subpar offense (.253 EqA, 10th in the league), top-flight closer Joakim Soria dealing with shoulder woes, and a manager, Trey Hillman, who has made enough questionable bullpen decisions to suggest that he's secretly a depressed goth girl with a tendency toward self-mutilation.
At least the Royals look better than the division favorites in Cleveland, who appear to have spit the bit for the third year out of four. While Cliff Lee has overcome his ugly spring with four straight quality starts, Fausto Carmona (6.11 ERA) remains lost enough to suggest that his 2007 was the fluke, and reclamation projects Carl Pavano and Anthony Reyes have scarcely looked worth salvaging. Oh, and that's not the Cuyahoga burning, that's Cleveland's bullpen. Meanwhile, the Tigers look considerably more robust simply for weathering a slew of pitching injuries and a fair bit of late-inning arson themselves.
Like the Royals, the Mariners top their division because of exceptional pitching and despite execrable offense. Felix Hernandez is living up to his regal nickname, Erik Bedard has bounced back from the injuries that curtailed his 2008, and the bullpen has withstood Brandon Morrow's early-season troubles. Overall, Seattle ranks second in run prevention thanks to the lowest homer rate and third-lowest walk rate. What's more, thanks in part to an excellent defensive outfield of Endy Chavez, Franklin Gutierrez, and Ichiro Suzuki, the team's Defensive Efficiency has improved by 17 points over last year's, jumping from 12th to third. The offense, which ranks second-to-last in EqA, is another matter, though it's improved considerably since Ichiro returned from the bleeding ulcer which cost him the season's first eight games.
In the East, history suggests that we ignore Toronto's hot start at our peril; 18-10 teams with three straight seasons above .500 tend to keep the good times rolling. On the other hand, particularly with three teams forecast to win at least 94 games, the PECOTA-based odds suggest a deck still stacked heavily against the Blue Jays, and last week I identified a handful of reasons they might regress. Forecast to have the league's lowest-scoring offense, they're suddenly and improbably the highest-scoring unit, fueled by an infield that's hitting a combined .303/.380/.479, with Aaron Hill (.360/.404/.552) and Marco Scutaro (.262/.400/.458) both particularly over their heads. Their rotation has been decimated by injuries, and it's possible that three starters who helped them post the league's top ERA last year-Dustin McGowan and Shaun Marcum, both rehabbing from off-season arm surgeries, as well as departed free agent A.J. Burnett-won't throw a single pitch for them this year. Through the end of April they had played the league's second-easiest schedule (.474, based on PECOTA-projections), but they'll face the AL's second-hardest (.513) overall. Not helping the Jays is the fact that the indicators suggest that neither the Yankees nor Rays have scuffled enough to rule them out, and both have substantial upgrades waiting in the wings-the former in the form of Alex Rodriguez, the latter via David Price, the game's top pitching prospect.
In all, it's still too early to discard those pre-season predictions simply because of a few teams markedly exceeding expectations. As Rany's work shows, it's not until the 30-game mark that the die is cast even for teams at the extremes, and only around the 48-game mark where current season performance outweighs projected performance. All of which is a polite way of saying check back in a few weeks and we'll really know what's going on.
A version of this story originally appeared on ESPN Insider .