Hi. My name is Dan Fox and I write a blog called Dan Agonistes where I struggle with various baseball questions ("Agonistes" is a Greek word that means "struggler"). I was delighted when Marc asked me to contribute to Beyond the Box Score. I've been perusing the contributions and been very impressed with the level of discussion and so am hoping that my thoughts will stir the pot a bit.
Speaking of which, my first article on The Hardball Times, Pythagoras and White Sox caused a bit of a stir last week on The Baseball Think Factory. My thesis was simply that the success of the "Smart Ball" Sox cannot be attribute to a new offensive style of play, but rather good pitching and good fortune in one-run contests. This led to an analysis of how well Bill James' Pythagorean method works during a season in predicting a team's final record.
While I don't want to recount the entire discussion here, what I found interesting in looking at Pythagorean projections were the Washington Nationals. Including today's 5-2 loss to the Mets, the Nationals are 50-32. The interesting thing is that they've scored 337 runs and given up 338. Using the standard Pythagorean method where:
Expected Winning % = (RS^2) / ((RS^2)+(RA^2))
the Nationals "should" have a winning percentage of .499 which equates to 41 wins rather than 50. In other words, they've won 9 more games than expected given their ratio of runs scored to runs allowed. Projected over the remainder of the season that means they'll win around 90 games. Meanwhile the Braves are sitting at 45-37 but have scored 392 runs and given up just 325 (best in the NL) which leads to a projection that they'll win around 92 games.
So why are the Nationals sitting at +9 and can they hold on to win the east? The answer to the first question can be found in their 22-7 record in one-run games for a winning percentage of .759. Since 1901 no team has played so well in one-run games and very few teams have played so many one-run games. Only the Orioles of 1981 got as high as .750 going 21-7 while few teams have played more than 58 one-run games in a season (only 7% through 2004).
And that gives us a clue to the answer to the second question. Clearly, they can't be expected to fare quite so well in one-run games can they? Well, record in one-run games can reasonably be linked to good relief pitching and in Chad Cordero, Luis Ayala, and Hector Carrasco, the Nationals have a trio of guys that can shut the door. Still, my opinion is that historically you can't expect to win 75% of your one-run games even with three or four good relievers. Given that, in order to stay competitive in the second half the Nationals are going to need to score more runs as their one-run winning percentage regresses towards the mean. Right now their infield (Vinny Castilla, Christian Guzman, Junior Spivey, and Carlos Baerga) and Marlon Byrd in left field are all sub-par offensively. Once Jose Vidro, Ryan Church, and Nick Johnson all come off the disabled list they'll look a bit better but they'll still struggle to score runs.
Should the Nats look to trade Vidro for some offensive help at third base or shortstop? And will Major League Baseball allow GM Jim Bowden to make the appropriate moves? What do you think?