Talking Sabermetrics: OPS vs. wOBA and Cubs vs. Astros

Two quick hits on the sabermetric front today. First, let's talk about OPS vs. wOBA and the relative effect of the two metrics on Astros' hitters. Second, a nugget regarding the difference between the 2012 Astros and 2012 Cubs.

OPS vs. wOBA

OPS (on base plus slugging) and wOBA (weighted on base average) are two offensive statistics we often use. Earlier this year, I discussed the difference between OPS and wOBA, pointing out that OPS is a quick and dirty measure of offense, but clearly inferior to wOBA in terms of accuracy. Both OPS and wOBA are correlated with runs, but wOBA is more closely related to scoring runs. Despite the fact that wOBA is the preferable metric, I admit that I sometimes use OPS. One of the main reasons is that wOBA, a fangraphs statistic, is not uniformly available on other baseball statistic web sites (like baseball-reference.com), which makes OPS easier to use in comparing data across multiple sources. The next link should make me think twice about using OPS when wOBA is available.

Hardball Times' recent article on the marginal effects of offensive events on OPS and wOBA is worth reading. The article shows that OPS may be pretty good at measuring offense at an aggregate team level, but it falls well short of wOBA is measuring the marginal effect of specific offensive events. That's not surprising since wOBA is based on linear weights, which measures the offensive value of specific events like walks, singles, doubles, HRs, etc. But the article demonstrates that the inaccuracy of OPS is surprisingly large if we focus on the change in OPS caused by each specific event. For example, on average: OPS over values a single, relative to a walk, by 52%; OPS over values a double, relative to a walk, by 98%; OPS over values a double, relative to a single, by 30%.

The article also finds that simply changing the weighting between OBP and SLG in the OPS formula is inadequate to correct the distortion. The reason is that the relationship between OPS and wOBA is variable (not uniform) at the individual player level.

This made me wonder which Astros' hitters are most over valued or under valued if you use OPS, rather than wOBA, to measure their offensive performance. In order to make this comparison, I created OPS and wOBA indices based on assumed averages of .717 for OPS and .315 for wOBA. (In other words, the index is a ratio of the player's OPS or wOBA to the assumed ML average.)

The following Astros' hitters are undervalued the most by the OPS metric, because their wOBA index is higher than their OPS index: Brian Bogusevic (1.05), Jordan Schafer (1.04), Chris Snyder (1.03), and Jose Altuve (1.01). The following Astros' hitters are overvalued the most by the OPS metric: Tyler Greene (.95), Matt Downs (.95), Fernando Martinez (.96), and Matt Dominguez (.96). Note that the ratio in parenthesis is the wOBA index as a percent of the OPS index.

The differences probably aren't a big deal. But the moral of the story? Don't just look at a hitter's OPS...the wOBA will tell a more accurate story.

Cubs vs. Astros Records

The Astros may have been lonely in the 100 loss club, and they took the direct route in bringing some company home for dinner. Monday's Astros win over the Cubs pulled the cubbies down into the 100 loss swamp. Currently the Astros have 106 losses, and the Cubs, 100 losses. Both teams are clearly the worst in the majors, with losses outdistancing other teams by at least three games.

What's the main difference between the Cubs and Astros? By that, I mean, what characteristic is most responsible for making the Cubs a better W-L team than the Astros?

I suggest looking at the defensive components of team WAR (fangraphs). The Cubs are +11 runs and the Astros are -45 runs on defense. That is a 56 run fielding differential. Assuming that 10 runs equals a win, that means that defense accounts for somewhere between a 5 and 6 wins difference between the two teams. Given that the two teams are separated by 6 games in the standings, defense appears to be a major cause of the difference in record.

You can also look at second base defense to recognize one of the largest discrepancies. Darwin Barney is +12 and Jose Altuve is -16, for a total of 28 runs difference in fielding by the two starters at that position---almost 3 wins difference. This kind of isolated comparison is cherry-picking to some extent. So, don't take it as "Jose Altuve's defense is the reason that the Astros have the worst record." That wouldn't be a valid conclusion. Moreover, viewing only the fielding component also ignores the fact that Altuve made substantial offensive contributions. But it does underscore the need for Jose to improve the defensive part of his game.