Either this is an extremely subtle parody or someone — nay, two people! — writing for the New York f@#$ing Times are trying to predict the World Series outcome based on starting pitchers’ win totals, batting average and fielding percentage, and presenting it all as analysis. Hat tip to James K.
Tedquarters
Ted Berg
Tedquarters 
This is the topper:
“This experiment resulted in a lackluster 1-3 record in the first round. The factors would have gone 0-2 in the League Championship Series, a testament to how surprising this year’s baseball playoffs have been. But these championship characteristics were identified by looking at World Series results, so let’s review the factors again in analyzing the World Series matchup between the San Francisco Giants and the Texas Rangers.”
Yes, I’m sure the problem with the Division Series and Championship Series predictions was that the “championship characteristics were identified by looking at World Series results,” and not the completely crappy analysis.
Oh and look at the predictions for the Division Series. Atlanta has a better 1-2 pitching punch than the Giants.
There is nothing in that story that isn’t based on faulty logic
Really fun fact. Over the same 20 year stretch (which is 19 series, because of 1994) the pitching staff with the higher walk rate has also won 68% (13 times) of world series. You read that right. The *higher* walk rate. By pitchers.
Even better, the staff with the *higher* team ERA has won 15 of 19, or 79%.
So, clearly, Jon Daniels’ big mistake was in trading for Cliff Lee and not Oliver Perez.
This blog post in the Times makes a lot more sense when you realize that the authors stopped following baseball in 1982.