Closer Jim Johnson
and the rest of Baltimore's bullpen deserve much of the credit for keeping the team in the hunt. (Icon SMI)
At a moment when the Nationals, A's and Pirates are all in line for postseason berths, one other surprise contender isn't receiving much love: the Orioles. Yet they continue to remain in the playoff picture. Thanks to a 57-51 record, they're second in the AL East, 6 1/2 games out of first place, and fourth in the Wild Card standings, one game out of a spot. Over the past week, they took two out of three from the division-leading Yankees in New York, and then two out of three from the Rays in Tampa Bay. All season long, they have held their own within the game's toughest division, going 22-19 (.537) against AL East opponents, compared to 24-20 (.545) for the Rays, 20-17 (.541) for the Yankees, 15-19 (.441) for the Blue Jays, and 18-24 (.429) for the Red Sox.
Though they're vying to be the franchise's first playoff team — and first team above .500 — since 1997, the Orioels haven't received the same amount of respect as the aforementioned surprise clubs because they've been outscored by 57 runs (501 to 444). Among AL clubs, the Rays (+19), Red Sox (+29), Blue Jays (+9) and Mariners (-3) all have better run differentials, even with the Sox, Jays and M's a combined 11 games below .500. In fact, at this point in the season no other team with a record above .500 has a run differential that's in the red. Baltimore's Pythagorean record — its record based upon runs scored and runs allowed, a better predictor of future performance than a team's actual record — is 48-60.
So how seriously should we take the Orioles? The Baseball Prospectus Playoff Odds, which are driven by remaining schedule and run differentials, adjusted for the quality of opposition, gives Baltimore just a 5.7 percent chance at making the playoffs. Even the Red Sox, who at 54-55 are 3 1/2 games below them in the standings, have a 10.2 percent shot. Those odds aren't simply theoretical, as history suggests the deck is strongly stacked against them. Few teams wind up exceeding their Pythagorean records by nine wins, and even fewer teams with negative run differentials reach the playoffs.
It's exceedingly rare for a team with a negative run differential to make the playoffs. In fact, just five have done so: the 1984 Royals (-13), 1987 Twins (-20), 1997 Giants (-9), 2005 Padres (-44) and 2007 Diamondbacks (-20). The Twins, who went 85-77, are the only one of those teams to win the World Series, with the Diamondbacks the only other team in that group to win a postseason series; they and the Giants are the only ones from among those teams who reached 90 victories. That said, none of those five teams played prior to the advent of division play (1969), and three of them hail from the Wild Card era (1995 onward). Given the addition of an extra Wild Card in each league this year, the likelihood that a team with a negative run differential might make the postseason increases, but it's hard to imagine that a team with a differential 30 percent worse than the previous record-setting outlier, the 2005 Padres, can even squeeze into the "coin-toss" game.
It's only slightly more common for a team to outdo its expected record by such a wide margin. Since 1901, 25 teams have exceeded their Pythagorean records by at least nine games in a non-strike season. Again, those teams aren't distributed evenly throughout history; eight of them have occurred since the beginning of the Wild Card era:
|YEAR||TM||W-L||Actual WPCT||Pythag. WPCT||Wins Above Pythag|
As denoted by the asterisks, six of those 26 teams made the playoffs, four of them in the Wild Card era, which raises the question as to why this should be something that's more common recently. The reason may have to do with the increasing load being carried by bullpens. Back in 2009, I studied the history of Pythagorean overachievers for Baseball Prospectus and found that teams most likely to exceed their expected records tended to have bullpens that ranked among the league's best. The focus in that study was on what BP calls third-order Pythagorean records, that is, records adjusted for expected runs scored and allowed, and the quality of opposition, as well as a measure of bullpen effectiveness that the site has since rendered obsolete.
Intuitively, the principle makes sense, however. We often talk of teams that over- or underperform their projected records as "lucky" or "unlucky," but it's a misnomer to chalk up the entirety of such discrepancies to luck. They generally stem from an irregular distribution of runs. "Randomness" may be a better term, but there's often a method to the madness, in that overachieving teams tend to win most of the close games — the ones where narrow leads are protected by a team's best relievers — but get blown out a few times, in games where the team's worst pitchers are most likely to be deployed.
The Orioles fit that pattern almost to a tee. They have the league's third-best bullpen ERA at 3.16, though that mark is offset somewhat by their league-high 36 percent rate of allowing inherited runners to score. Their Fair Run Average, which divides up the responsibility for inherited runners between starters and relievers according to the base-out situation, ranks fifth in the league at 4.18.
What's remarkable is that the Orioles are an astounding 21-6 in one-run games, producing the highest winning percentage (.778) in such games since 1901. The 1970 Orioles hold the record for the best winning percentage over the course of a 162-game season at .727 (40-15), though that mark was bettered by the 1981 Orioles (.750 at 21-7 in a strike-shortened season) and 1908 Pirates (.733 at 33-12 in a 154-game season).
Furthermore, these Orioles are an MLB-best 36-17 (.679) in one- and two-run games, with a winning percentage that would rank 12th since 1901. They're 19 games above .500 in a subset of games in which they have outscored opponents by just 23 runs; their Pythagorean winning percentage in such games is .552, meaning that they've won about seven more games than expected in that subset. Meanwhile, they're 13-19 in games decided by at least five runs, six games below .500 in a subset in which they've been outscored by 54 runs; even there, they've won about one-and-a-half more games than expected via their Pythagorean record. No other team with a winning record this season has a sub-.500 record in such blowouts, and in fact, one has to go back to the aforementioned 2007 Diamondbacks to find a team that lost more blowouts than they've won and still made the playoffs. Those Diamondbacks went 20-26 (.435) in such games while being outscored by 62 runs, compared to 47-29 (.618) in games decided by two-runs or fewer, and 32-20 (.615) in games decided by one run.
Even given the strong performance of their best relievers — at 3.47, closer Jim Johnson has a higher ERA than setup men Pedro Strop
(1.31), Luis Ayala
(2.68) and Darren O'Day
(2.68) — the weight of the historical evidence strongly suggests that the Orioles aren't likely to keep winning the close ones with such frequency. If they can't win the close ones, the O's aren't likely to remain in the crowded playoff hunt, or even sustain a winning record; indeed, after a 14-9 April, they've gone a combined 43-42 since then, while being outscored by 69 runs. Even so, while run differential is predictive, it is not destiny, and the Birds' ability to even maintain a competitive front given the imbalance between their runs scored and runs allowed is one of those fascinating anomalies that bears watching.