Run Differential Underselling the Yankees

May 7, 2014; Anaheim, CA, USA; New York Yankees manager Joe Girardi (28) walks to argue a call in the fifth inning against the Los Angeles Angels at Angel Stadium of Anaheim. Mandatory Credit: Andrew Fielding-USA TODAY Sports

The 2nd month of the baseball season is underway and some teams appear to have been underrated (Marlins and Brewers) or overrated (Phillies and Diamondbacks) in the preseason based on the current standings. However, it is still very early and win-loss record can be deceiving and not totally indicative of a team’s talent level despite the fact that those wins and losses are “banked” and no one can take them away. Dave Cameron of Fangraphs devised a method for evaluating true talent of teams this early in the season that attempts to shield from small sample noise. The link contains the complete methodology.

In summary, he tries to strip out the sequencing that greatly influences runs scored/runs allowed and wins/losses. He wants to only look at the wOBA differential (offense wOBA minus defense wOBA) to get a better sense of the quality of each team and how they would be expected to perform over the remainder of the season when the sequencing (essentially luck) evens out.

Mr. Cameron goes a step further than simply wOBA differential. He adds in base-running components because teams that are better on the basepaths should score more runs than their batting lines (offense wOBA) indicate. He also adds in catcher caught stealings, double plays turned, and outfielder throwing (both base-runner kills and base-runner advancement deterrence) because these are important aspects of defense that wOBA against doesn’t account for.

A chart is devised that runs this calculation for each team. As of May 2nd the Yankees have an expected run scored of 122 and expected runs allowed of 121 for an expected run differential of +1 (good for T-14th in the league). In actuality, they have scored 112 runs and allowed 125 for an actual run differential of -13. The gap between their expected run differential and actual differential, -14 runs, suggests they haven’t had a lot of luck with sequencing and they are better than the actual run differential figure.

Since sequencing is random, the Yankees would be expected to have a better actual run differential in the future. The data is a few games old, but the greater point is that this methodology is a better measure of a team’s talent level (especially this early in the season) than simply looking at win-loss records and actual run differentials. To date, the Yankees have an 18-15 record and a -11 run differential but their adjusted wOBA differential suggests they have gotten slightly unlucky on sequencing so far and should be a legitimate contender in the tight AL East.