A last look at the run differential for the 2012 Mets

Last night’s loss to the Braves dropped the Mets’ record to 73-85. The team’s actual record matches exactly its Pythagorean Record. Earlier in the year the talk was how the Mets were exceeding their Pythagorean Record by an insane amount. Those of us who pointed this out were scoffed at and while I looked for reasons why Pythagoras might not apply to the Mets back in late May, run differential once again has done a good job of predicting the team’s fortune.

In case you’ve forgotten, here’s how Pythagoras has done the last five years with the Mets:

2011: 79-83 (actual: 77-85)
2010: 81-81 (actual: 79-83)
2009: 72-90 (actual: 70-92)
2008: 89-73 (actual: 89-73)
2007: 86-76 (actual: 88-74)

After May 14th, the Mets’ actual record was 20-15 while Pythagoras had them earning a 16-19 mark. At that pace, the Mets were on target for a 93-win season while their run differential predicted 74 wins. Clearly they are going to finish the 2012 season with a ledger much, much closer to what their mid-May run differential predicted.

The point of this is not to suggest that you can tell a team’s ultimate record by its May run differential. Instead, while a team can exceed expectations at the one-quarter mark, by the end of the 162-game season that Pythagoras will be pretty darn close to being right. Even the Orioles have a positive run differential right now. No team in baseball with a winning record has a negative run differential.

Even the 2012 Mets, who were a (-35) in run differential after May 24, climbed to a positive run differential. By July 3rd, the Mets had a +32 run differential and their Pythagoras and actual records both projected an 88-win season.

Since then the Mets have been outscored by 87 runs as the offense has posted two or fewer runs 33 times in 77 games. They have a 3-30 mark in those contests.

In September alone, the Mets have scored 88 runs and allowed 110 in 26 games. The pitchers have allowed 4.23 runs per game, which is exactly league average for the month. The problem remains the offense.

Over the last 15 games, the Mets are 8-2 when they score three or more runs.

To date there have been 2,528 games played with a team from the National League. Here are the league averages for low scoring games and what the Mets have done in those same instances:

0 Runs 174 7% 12 8%
1 Run 302 12% 17 11%
2 Runs 338 13% 20 13%

Overall, the Mets are about average in these low-scoring games. A team in the NL will score two or fewer runs about one-third of the time (32%). But earlier we saw the Mets had a low-scoring game 33 times in their past 77 games or 43% of the time.

Perhaps the offense would not seem so dismal if these low scoring games were evenly distributed throughout the year. I’m shocked the Mets have only been shut out 12 times this year but four of those have come in September and eight of them since August 8th. Meanwhile, the Mets were blanked just three times the first three months of the season.

Regardless of the distribution of runs, there simply have not been enough of them here in 2012. If the Mets hope to improve their record going forward, they will have to address the lineup in the offseason, including ways to perform better versus LHP, who held Mets batters to an OPS 43 points lower than RHP.

2 comments for “A last look at the run differential for the 2012 Mets

  1. Name
    September 30, 2012 at 11:41 am

    I guess the Pythogrean theorem is sorta like fielding metrics, the longer the sample size, the more credible it gets. In the short term, you can have odd splits. It’s hard to continue winning in all low scoring games and then get blown out in all losing games. Eventually, good teams will blow the other team out and also lose close games.

    I don’t think that low scoring chart really tells the whole story this season. In the first half in had 4/6/7 in 0/1/2 games respectively. Those are percentages of 4.7%/7.0%/8.1% , which are almost half the league averages. In the 2nd half, they are 8/13/10 which translates to a 11.1%/18.1%/13.9%. Those numbers are WELL above the average. I’m pretty sure any fan could have deduced this because we all know that in the 2nd half the offense has virtually been non-existant. But i think we’re not giving them enough credit for what they did in the first half.

    • September 30, 2012 at 11:59 am

      So the question remains: How much did the Mets overachieve in the first half compared to how much they underachieved here in the 2nd half? Do you think the 73-85 record right now is indicative of the talent on hand?

      My completely biased view is that the underachievement here in the second half has outweighed the overachievement earlier in the year. I think the talent as currently distributed is closer to .500 than 12 games under.

      Of course, that talent needs to be redistributed in the offseason.

Leave a Reply

Your email address will not be published. Required fields are marked *