July 5, 2005
Can Of Corn
Pythag and the Nats
At the exact midpoint of the current season, the profoundly surprising Washington Nationals held a five-and-a-half game lead in the NL East and were on pace for an even 100 wins. As surprising and impressive as that might be, the real oddity about this team is that, if trends hold, they'll win 100 games and a division title despite a run differential of -2. Calling such a confluence of events "historically unprecedented" would be to indulge in understatement of criminal proportions.
Since 1900, 90 teams have won at least 100 games in a season. Those 90 teams had an average run differential of +214.9. The worst run differential by a 100-win team belongs to the 2004 Yankees, who finished with a +89 mark. Only two other teams, the '69 Mets and '70 Reds, have posted run differentials of worse than +100 while still winning 100 games for the season. If the Nats were able to reach the century mark in victories while burdened by a negative run differential it would be, suffice it to say, stretching the depth and breadth of improbability.
The Nats are certainly overperforming to date, but it's a bit facile to say their success is merely the product of good fortune. In fact, there's some evidence to suggest that the team's meager run differential isn't indicative of its genuine, long-term quality. To wit, Washington has a strong and properly deployed bullpen, and they've been exceptional at home. Let's take these one at a time …
Much has been made of the fact that the Nationals have been outstanding this year in one-run games (22-7 on the season). In many analytical circles, success or failure under such circumstances is often dismissed as a matter of luck. However, if a team makes a habit of giving high-leverage innings to its best relievers (and, by extension, low-leverage outings to its worst relievers), they can thrive in one-run contests and, hence, exceed the various Pythagorean-inspired projections. In D.C.'s case, their four top relievers all rank in the top 50 for Relievers' Expected Wins Added (REWA). Also, those four relievers--Chad Cordero, Luis Ayala, Hector Carrasco and Gary Majewski--have worked almost two-thirds of the team's total bullpen innings. Now here's how they fare in terms of leverage:
Pitcher IP R/G REWA Rank Leverage (Rank on Team)* Cordero 44.1 1.62 1 1.97 (1) Ayala 48.0 3.38 13 1.49 (2) Carrasco 29.0 1.86 37 1.22 (3) Majewski 35.1 3.57 49 1.15 (4) * - Among Washington relievers with at least 20 IPAs you can see, manager Frank Robinson is deploying his relievers in exact accordance with their REWA rank--that's to say, the best relievers are working the most critical innings. That they've also thrived in nip-and-tuck affairs is no accident. When you have four relievers the quality of those above and you're using them in critical situations, you're going to have a strong record in one-run games.
There's also the matter of the club's success in home games. At the halfway point, the Nats' record at RFK was 29-10, which comes to a winning percentage of .744. That puts them on pace for 60 home wins this season, which would mean this year's Nats model would tie the eighth-best single-season home-win total of all-time. Here's the full list:
Rank Team Home Wins 1. '61 Yankees 65 2. '75 Reds 64 3. '32 Yankees 62 3. '98 Yankees 62 5. '46 Red Sox 61 5. '49 Red Sox 61 5. '62 Giants 61 8. '31 A's 60 8. '42 Cardinals 60 8. '53 Dodgers 60 8. '69 Orioles 60 8. '77 Phillies 60Of the 12 teams listed, nine exceeded their Pythagorean records for the season, and they bettered those forecasts by a cumulative 47 games. All of this has at least some bearing on the Nationals' troubling run differential. Of Washington's 29 home wins in 2005, only two have come in extra innings or in walk-off fashion. That means in 27 games this season, the Nats haven't batted in the ninth inning. In other words, that's 81 outs they didn't get to use. Teams that are especially potent at home sometimes fall short of their expected records because they don't have as many opportunities to score runs. That's borne out in Washington's underwhelming Pythagorean record.
The Nats have certainly benefited from good fortune this season, but their run differential can partially be explained by some of the team's unique qualities. They have a strong bullpen that's properly deployed, and they excel at home. They're over-performing, to be sure. However, the extreme second-half regression you might expect based on their run differential probably won't come to pass. Summarily speaking, the Nats are worse than their record but much better than their run differential.
As for how the NL East will play out, it's far from decided. In the second half, the Braves will get back from the DL 60% of their Opening Day rotation and Chipper Jones. So it'll be an onerous task for the Nats to hold off the 13-time champs. However, if you're tempted to dismiss Washington because they've given up more runs than they've allowed, don't do it.