05.17.07
Inter-league Dominance
Founded in 1900-1901, 25 years after the National League began play, the American League is sometimes called the “Junior Circuit”. Yet this name is half in jest; the two leagues share equally the label of “major league baseball”, draw from the same talent pool, and have similar resources available. Little distinction remains between the leagues since the separate offices were eliminated a few years back.
History provides scant evidence for either league to assert its primacy. The National League holds a 41-34-2 advantage in the All-Star Game — an argument that their top talent, at least, has historically been slightly superior. Against this we see that the American League has more World Series victories by a 60-41 margin. But even this seemingly lopsided comparison tells us little about the leagues as a whole, since the comparison reverses if you remove the 26 Yankees crowns from consideration.
Results from recent years have hinted that the American League might be emerging as the stronger. Over the last ten seasons, the AL has a 9-0-1 record in the All-Star game and a 33-20 advantage in World Series games. Admittedly these are poor indicators, reflecting only a small segment of the talent in each league, while inter-league play has supported the case for equality. The annual discrepancy in W-L has consistently been less than 22 games with the overall results close to .500. Neither league has held an advantage for more than two consecutive years.
In this context, the results we saw last year were shocking. Over 252 interleague games involving all 30 teams, the American League compiled a 154-98 record for a .611 winning percentage. For comparison, the top teams of 2006 each finished with 97 wins for a .599 winning percentage. The average AL team beat the average NL team by a greater margin than either the Mets or Yankees beat their opponents!
It is inconceivable that this outcome might occur in two evenly matched leagues. Estimating the odds using a binomial distribution [in Excel type =2*(1-BINOMDIST(154,252,0.5,TRUE)) ], we find there is only a 4-in-10,000 chance of seeing a split this large by random chance. Neither can this result be explained by excepting one or two teams. The top seven AL teams (based on the final 2006 standings) had a 79-47 record against the NL. The bottom seven AL teams “struggled” to a 75-51 record. We can only conclude that the AL last year was truly and uniformly superior.
Furthermore, 2006 marks the first time that one league has led for three straight seasons. Since the NL last led the AL by 22 games in 2003, the AL has posted advantages of 2, 20, and now 56 wins! The steady increase over a four-year span perhaps lends further credence to the argument that this is more than a fluke? This is a dramatic split, roughly equivalent to what you might expect if ten All-Star caliber players were to switch leagues with no talent coming the other direction. While the 2005-6 free agent market did seem to favor the AL, nobody anticipated an effect of this magnitude.
The 2007 results should offer us guidance. If the AL once again dominates by 25 or more games, even the skeptics might begin to believe. If the margin narrows more than that, then I’ll be left wondering how the 2006 blip arose — and whence it departed. The Baseball Prospectus Depth Charts imply a 14-game advantage for the AL, however this reflects a somewhat arbitrary choice by Nate Silver in setting the parameters of his PECOTA projection system.
Time will tell. Will Atlanta present a serious challenge to the Red Sox, as a team with a .625 winning percentage ought? Or will they remind us more of Oakland and Seattle, both teams with many strengths and some serious flaws? If the AL is once again the superior league, the middle of June should be a cakewalk for the Red Sox. If not, we’ll face some tough competition.