Originally published on DRaysBay on August 2, 2013
The Rays grabbed first place from the Red Sox Monday night in a dramatic game that featured a rain delay, a close call at home plate, and a late rally by the Red Sox. The Rays (63-43) lead the Sox (63-44) by a ½ game going into Tuesday’s night game against the Arizona Diamondbacks. Of course, many analysts were skeptical of both teams’ ability to contend this season and were bearish on the AL East as a whole. While the Red Sox had a disastrous season in 2012, the Rays also fell short of expectations by missing the playoffs (mostly due to the Orioles shocking resurgence and Evan Longoria’s injury). Both teams tried to address their needs in a variety of ways. The Rays traded ace pitcher James Shields for Wil Myers (shades of the Matt Garza deal) in hopes that they could use their pitching depth to replace Shields’s production in the rotation and that Myers, when called up, could bolster an anemic offense. The Red Sox were “quiet” in free agency, spending over $126 million on 8 players but giving no player a deal worth more than $40 million. Aside from the blockbuster Shields trade, the Rays did their annual “dumpster diving” in free agency, signing Red Sox castoff James Loney (-1.1 WAR/73 OPS+ in 2012), Yunel Escobar (75 OPS+ in 2012 and well known as a clubhouse cancer), Kelly Johnson, and the pitching act Fausto Carmona now known as Roberto Hernandez. Some Rays fans believed that this was going to be the year that budget constraints finally caught up with the Rays.
The national media generally considered the Blue Jays, who were aggressive in the pursuit of big names in free agency and through trades, the favorites in the weakened AL East. The Orioles were going to come back to earth after a historically lucky season in 2012, the Yankees were too old and hurt to contend, the Red Sox won only 69 games the year before and the Rays were coming off a season in which they missed the playoffs and, as always, were operating on a shoestring budget. The reasons for the Blue Jays failure to contend and the Red Sox’s “sudden” resurgence have already been discussed at length. The Yankees have received more than their fair share of coverage (Vernon Wells and Travis Hafner are awesome; they’re awful! A-Rod is gone; he’s back maybe).
More interesting is the tight race between the Red Sox and Rays. It seems like just yesterday that the Red Sox were in a tight race with the Yankees for first place, the Orioles were lurking in the middle of the pack, and the surging Blue Jays had passed the Rays, who despite a winning record were in last in the AL East. A 21-4 July by the Rays and the Yankees unbelievable injury problems have completely changed the landscape.
A significant amount of coverage has been dedicated to the Red Sox’s signing of Shane Victorino to a 3 year, $39 million contract and more recently, Cliff Lee’s monster contract and how these are team friendly if they continue to produce Wins Above Replacement (WAR) at their present rate. But how accurately does WAR predict team success? Bill James famously developed the Pythagorean Expectation to predict how successful a baseball team will perform based only on the amount of runs they score and allow over the course of a season. It is not unreasonable, using today’s advanced Sabermetrics, to predict a team’s future performance by adding potential signings estimated runs created (through batting and base running) and estimated runs saved (through pitching and fielding) by replacing current players with potential signees or callups. However, there is another metric that has become the Holy Grail to modern baseball statheads: WAR. Again, how accurately does WAR predict team success? Or, how accurately does WAR predict a team’s projected success based on Bill James’s tried and true equation?
First a look at how Pythagorean expectation predicted American League records in 2012:
This could very well be too small a sample size, but the Pythagorean holds up quite well here, with the outliers labeled.
Next question: How well does a team’s total WAR predict their win total? After all, the name Wins Above Replacement suggests that there should be a strong correlation. Again, the 2012 AL is used.
The correlational coefficient suggests a slightly better correlation between actual wins and WAR than actual wins and expected wins. But, what happens when expected wining percentage is compared to total team WAR.
The correlation coefficient approaches 1 when WAR and expected wins based on expected winning percentage are graphed, again with the 2012 AL as the guinea pig.
Since the Pythagorean winning percentage is based solely on runs allowed and runs scored, would summing a team’s offensive WAR and pitching WAR lead produce similarly high correlational coefficients? Again using the 2012 AL:
Although WAR was never designed to be treated like this, it is surprising that although total team WAR is highly correlated with expected winning percentage, the individual components of WAR do not correlate strongly with the individual components used to calculate expected winning percentage. This can be explored in future analyses.
How can we use this data to predict the division race in the AL East? Historically speaking, a high correlation between expected wins and team WAR has held. Therefore, the team whose expected wins matches actual wins with a strongly correlating WAR can be assumed to be performing where it will continue to perform. If -of course- you put your faith in WAR.
Through Monday the division leading Rays have a record of 63-43, an expected record of 61-45 and a team WAR of 28.1.
The second place Red Sox have a record of 63-44, an expected record to match and a team WAR of 35.2.
Should be an interesting finish to the season.