Friday, January 21, 2011

Simplifying Sabermetrics: OPS+

Installment Two: OPS+ (Adjusted On-Base Plus Slugging)

A few weeks ago we dove headfirst into the world of sabermetrics and discussed the ins and outs of OPS, or on-base plus slugging. We learned that OPS is the sum of a hitter’s on-base percentage and slugging percentage. On-Base Percentage (OBP) is essentially a players productivity – how often does he reach base out of X opportunities – and slugging percentage (SLG%) measures how far he typically goes when he gets on base – single, double, triple, or home run. Adding those two figures together give you OPS, which gives a better way to gauge a player’s relatively complete hitting profile.

We started with OPS because it was the stepping stone to more advanced statistics. Not too long ago, I was watching MLB Network and overheard Peter Gammons use the OPS+ stat while he was discussing Roberto Alomar’s Baseball Hall of Fame nomination. I had seen this stat before and knew it’s meaning, but it was one of the first times I could remember hearing a TV analyst use it as if it were just as commonly thrown around as batting average or runs batted in.

So what is OPS+ and what exactly is being “adjusted?” Before we go any further, let’s take a look at the formula:
Let’s break down the math a little bit first. A 100 OPS+ is considered league average, and each point above or below that represents a percentage point. We already know how to calculate a player’s OBP and SLG, but we have two new parts to the equation, *lgOBP and *lgSLG. These are adjusted OBP and SLG for each league, meaning it takes the pitchers out of the equation.


The biggest draw of OPS+ is that it takes into consideration the park (Citizens Bank Park, Coors Field, etc) and the league the player hit in (American or National) and adjusts his OPS according. Why is it necessary to consider these variables? Well, there are a few reasons. One, every ballpark is different. You are probably used to hearing the terms “pitcher’s ballpark” and “hitter’s ballpark” thrown around a lot. Each park’s dimensions are different in that the walls of the outfield are at different angles and distances from home plate. In addition to park dimensions, there are natural variations to consider as well, such as the low air pressure at Coors Field. Because of all of these variables, OPS+ uses the league adjusted average OBP and SLG in its equation.

Why does OPS+ worry about taking out the pitcher’s hitting stats? Well, mostly because a pitcher only hits every fifth game if he’s in the National League, and maybe only a handful of times if they’re an American League pitcher during interleague play. There is not a large enough sample size or enough consistency across the board to justify using their numbers when ranking hitters when they are not, but definition, hitters themselves. Sure, some of them love to hit and do it well (Cliff Lee), some of them will surprise you with a homer (Joe Blanton), and some of them will smack the catcher in the face with his bat while just standing there awkwardly, praying that he can head back to the dugout sooner rather than later (Cole Hamels). But for the most part, it’s a nice picture to look at when a player is ranked against those he has the most in common with, so we adjust.

Now that we know the “why,” we can look at the “how.” Here is the formula one more time:
Let’s use frequently debated Chase Utley vs. Robinson Cano as our example. Because the whole point of OPS+ is to maintain neutrality, I will use Cano’s 2010 statistics and Utley’s 2008 numbers, both stand-out years for these players with tons of similarities.
OBP SLG *lgOBP *lgSLG
Cano (2010) .381 .534 .328 .408
Utley (2008) .380 .535 .340 .427
Now that we have the data, we can plug it into the formula. I’ll use Cano’s numbers to show my work. For the sake of simplicity, we round the final result to the nearest whole number:
Using the same formula for Utley, we get a 2008 OPS+ of 137. But is that all? Is the great debate settled? Is Robinson Cano actually better than Chase Utley?

Well, this is going to be debated over and over again, but what we can learn from the calculations is that in 2010, Cano was 47% better than the average batter in the American League not including pitchers, and that in 2008 Chase Utley was 37% better than the average batter in the National League not including pitchers. Although their batting statistics may have been virtually identical these years, you get a much different picture of how they compare to the rest of their leagues when you take the outliers out of the equation.

To be perfectly honest, that’s the long and short of it: how much better is someone than the average bear? As the example above indicates, a player’s OPS+ rarely mirrors their OPS. We can clearly see this in practice considering Utley and Cano had near identical OPS and SLG in the data we used, however they are 10 points apart when we adjust for parks and pitchers.

For more information on OPS+, check out Baseball-Reference.com. Next time, we’ll look at BB% and K% if you want to get a head start! Pin It Now!

No comments:

Post a Comment