« Baltimore's diamond constellation | Main | 2005 midseason analysis, part 2: team stats »

2005 midseason analysis, part 1: the standings

At the All-Star break, the 47–40 Orioles find themselves at a pivotal stage in their drive for a playoff spot. Their tumble over the past three weeks dropped them two games below first-place Boston in the American League East and a game and a half behind Minnesota in the wild-card race. And the Orioles aren't the only team contending for that last playoff spot: Cleveland, New York, and Texas trail the Twins by two games.

Early in the season, the Orioles dominated opponents like a division leader should, winning nearly two out of every three games while their main rivals, the Red Sox and Yankees, battled injuries, slumps, and decline. Indeed, for the season's first two months, Baltimore appeared to be cruising toward a postseason berth, and as recently as a month ago, the O's still sat comfortably atop their division with a three-game lead over Boston. Housing in Jimmyville (the fictional municipality of optimistic Oriole fans) was booked solid.

But as usually happens, the breaks began to even out. Injuries took critical players out of the Orioles' lineup, and the reinforcements were not as good as the men they replaced. The pitching rotation regressed, the bullpen blew a few leads, and some hitters fell into slumps. Meanwhile, other teams caught up to or passed the Birds. Baltimore's current trajectory is uncertain: from here the team could remain in contention or sink further still.

How have the O's reached this precarious position, and where do the signposts point for the second half? A survey of the team's first-half statistics may help uncover the answer.

Of records and run differentials

As of the break, the Orioles have scored 431 runs and allowed 409. Per game, they have scored 4.95 runs (fourth most in the AL) and allowed 4.70 (ninth least). Their league ranks in those categories are the same when park effects are factored in. After converting those runs into wins and losses using Bill James's Pythagorean method, Baltimore's run differential translates into an expected record of 46–41, about the same as their actual record. While that is respectable, eight(!) other AL teams have better Pythagorean records. So it appears that Baltimore's current place in the standings is reflective of the team's true quality: while a winning team, the O's have not been as good as Boston and have fallen back to the crowded pack of wild-card competitors.

Not satisfied with Pythagoras and his broad-brush ways? Going another level deeper, the adjusted standings page at the Baseball Prospectus web site paints a slightly rosier picture of the Orioles. According to a calculation based on Clay Davenport's Equivalent Runs (a stat that accounts for offensive events, park effects, and league run-scoring levels) and his Pythagenport extension of James's method, the Orioles' second-order won-lost record is 49–38. This two-game discrepancy from their actual record is hardly unusual, but the 49 wins are tied with Boston as the second most in the league. Cleveland, with the benefit of an additional game played, tops the pack in second-order wins at 50–38. And the Chicago White Sox appear to be due for a fall because their second-order record of 46–39 is eleven games worse than their actual record. Meanwhile, Baltimore's third-order W–L record (which adds in quality of competition) is also 49–38. That mark is fifth best in the league.

This year's Birds have shown significant improvement over the stragglers that started last year 37–48. A year ago, the O's had underperformed Davenport's adjusted records by a significant margin—at the 2004 All-Star break, Baltimore's second- and third-order winning percentages were close to .500, but its actual winning percentage was just .435. So expectations of a second-half improvement were entirely within reason, and the team fulfilled those expectations by rebounding with a record of 41–36 after the break to finish at 78–84. It appears that the success the Orioles enjoyed in the second half of 2004 has carried over to this year.

So far in 2005, the run-scoring data indicate that the Orioles have played up to their won-lost record, so they should be able to maintain that level of performance in the second half of this season. But will that be good enough to make the playoffs?

In the year 2005...

Another interesting experiment by the BP crew is their daily updated Postseason Odds report, which projects the season's final standings by means of a Monte Carlo numerical simulation. The accuracy of this approach is somewhat questionable given the simplicity of the model, but it's worth a look if you are at all curious about the season's outcome. To produce the report, the model calculates an estimate of each team's true quality from the third-order winning percentages in BP's Adjusted Standings, then makes a few other assumptions and proceeds to simulate the remainder of the season's schedule one million times. The report tallies the number of times each team won its division or the wild card in the simulations and expresses the totals as a percentage. (A deeper explanation of the model's methodology is available on the BP web site. Last fall, Davenport added two articles describing revisions he made to improve the model. It's still a work in progress.)

At the moment, the Postseason Odds report gives Baltimore a 27 percent chance of making the playoffs. That breaks down into an 18 percent chance of winning the division and a 9 percent chance of taking the wild card. In the latest simulation, Boston made the playoffs two-thirds of the time (58% division champion + 8% wild card) and the Yankees were right behind the O's at 26 percent (18% DC + 8% WC). According to the report, the AL wild card is likely to come from the Central division—a Central team took the wild card spot in about 56% of the simulations, with Cleveland the heavy favorite at 26% (44% overall). In contrast, an East team claimed the wild card just 28% of the time and a West team 16% of the time. On average, a team needed 91.6 wins to claim the AL wild card and 93.3 to win the AL East.

If these simulated results have any real-life validity, then the O's have their work cut out for them. In their remaining 75 games, the Birds will probably have to play at least .600 ball (45–30) to make the playoffs. Though not out of the question, that type of performance is a tall order to fill given the way the team has played lately. To make the postseason, Baltimore needs to win the way it did in April and May, not tread water and collapse the way it did in June and early July.

Can it do that? Stay tuned for the next installment.


This page contains a single entry from the blog posted on July 13, 2005 2:56 PM.

The previous post in this blog was Baltimore's diamond constellation.

The next post in this blog is 2005 midseason analysis, part 2: team stats.

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.33