Allen St. John in this WSJ ($) article today analyzes Major League Baseball’s top teams from last season with baseball statistician Bill James’s Pythagorean Theory, which concluded that a team’s run differential (runs scored vs. runs allowed) correlates closely with its winning percentage. Unfortunately, the 162 game MLB schedule does not provide enough games to even out all the bad bounces and bloop singles (i.e., bad luck). So, Mr. St. John calculates each team’s Pythagorean winning percentage — which largely eliminates the effects of luck — by taking the square of the total runs they score in a season and dividing it by the sum of runs scored squared and runs allowed squared. Some of his observations:

Luck, it turns out, was a big factor in pennant races last year. In the world of P-Wins, four playoff teams — including the World Champions — would have watched October baseball from their Barcaloungers. The American League Central would have been taken by the White Sox (who had three fewer wins than their P-Win total would indicate, resulting in a P-Win differential of -3), not the Twins (whose real wins topped their P-Win total by five — a differential of +5). The Mariners (-6) should have edged the As (+1) in the AL West. In the National League, the Phillies (-5) should have blown past the Marlins (+3) for the wild card. In the NL Central, the Cubs (+2) would have finished third, behind the Astros (-8) and Cardinals (-4).
This year, it’s a new ballgame. That’s because of a corollary to Mr. James’s theory — luck evens out. Teams that fall short of their P-Win total one season tend to bounce back the next year. Teams that exceed their P-Win total often slip back the following summer
What do P-Wins tell us about the coming season? Looking past the personnel changes, the numbers still tell an interesting story. In the AL East, the margin between the Yankees (+4) and the Red Sox (0) should wither. The Twins should find the White Sox in hot pursuit. In the NL West, Seattle seems primed for a playoff run, likely at Oakland’s expense. And in the NL East, look for the gap between the Braves (+4) and Phillies to get very small.
The NL Central poses an interesting problem. While most of the P-Win differential is due to luck, that isn’t the entire story. Cubs manager Dusty Baker has a career P-Diff of +18, i.e. his teams have won 18 more games than the numbers would suggest. Astros skipper Jimy Williams has a career P-Diff of -24, with just one season in which his team exceeded expectations. The NL Central this year may prove Branch Rickey’s adage: Sometimes luck is the residue of design.

Meanwhile, this interesting MLB.com article describes how many MLB clubs are finally adapting sabermetrician principles to evaluation of baseball players.