Edit: It’s the next morning, and I feel like I lost my cool a bit when I was typing this up late last night. I apologize for that, but I’m going to leave the post up anyways. At the very least, you’ll be happy (or shudder) when you see how QB rating is calculated.
From my previous post, you saw that I had a heyday on blasting Patrick Reusse because he included Star Tribune sports colleague Joe Christensen on his annual list for Turkey of the Year simply because Joe C. uses OPS as a statistic. Reusse called it a “make-believe statistic,” which greatly annoyed me. Maybe he was alluding to the fact that OBP and SLG have different denominators, thus it makes no sense to add the two of them together (I’ll explain in a second). Maybe he just thinks it’s a complicated statistic, which is a bit ridiculous considering how easy it is to calculate it. Or maybe (like I believe why many people dislike new or sabermetric statistics) it’s because he didn’t grow up with OPS, so he doesn’t see why he should believe in it. Whatever the case, I just want to show that while OPS and other sabermetrics maybe be complicated to calculate, baseball is not alone in using stats with extensive calculations.
Example 1: OPS (MLB)
Rather simple, huh? I suppose to make this fair though, I’ll include how you get OBP and SLG as well.
Well I learned something new. In calculating OBP, you do NOT use plate appearances as the denominator. Plate appearances, in addition to what is included above, also includes sacrifice hits and times that the batter reaches base due to defensive interference.
Now do you see what I meant when I said that the denominators aren’t the same? So if we were super anal-retentive about the laws of mathematics, we’d say that OPS shouldn’t exist. It would be similar to trying to combine inches and ounces, or oil and water: They don’t mix.
This is it for the baseball examples. I could easily pull up something more difficult like FIP or UZR or tOPS+ like I have used in the past, but I’m sticking with OPS because that was what bugged Reusse.
Example 2: QB Rating (NFL)
Here’s the kicker though. A, B, C, and D must be between 0 and 2.375. If the letter’s value is below 0, it gets rounded to 0, and if it’s above 2.375, it gets rounded to 2.375. You have probably assumed correctly that if all four are 0, you get a 0.0 QB rating, and if they’re all 2.375, you get the perfect 158.3 rating.
Now it’s time to blow our minds.
Example 3: Efficiency Rating (NBA)
Back when Kevin Garnett was still with the Timberwolves, this statistic was cited quite often because KG usually was either the leader or among the leaders. Are you ready for this? I am not liable if your jaw drops and knocks your F, G, and H keys out of your keyboard. You probably should click on the image so you can clearly read everything.
Did I mention that KG led the league in that? Way down in the lower-left corner is PER, the final product from this hellish calculation. This thing is so long, its Wikipedia page includes a horizontal scroll bar, because it’s about 3 1/3 times the width of a normal Wikipedia column. Thank god we have computer programs that can calculate that in seconds.
Oops, forgot some things.
I included the average because PER is normalized such that an average rating is 15.0. Wikipedia adds that the highest PER ever achieved in a season is 31.84. I don’t know why the average is closer to the lower limit than the upper limit, but hey, if I knew, I would have actually read the entire calculation instead of taking one glance and thinking, “My brain hurts.”
I’m sure if I continued searching, I could find other non-baseball statistics that could easily qualify as being in that sport’s realm of sabermetrics. However, I used QB rating and player efficiency rating because although they are quite lengthy to calculate by hand, they are still used on a semi-regular to regular basis. My point is, Patrick Reusse included colleague Joe Christensen as a candidate for Turkey of the Year simply because Joe C. referenced a statistic that has now become so commonplace that ESPN regularly shows it on their telecasts now. I don’t know why Reusse feels that OPS is so asinine, other than the fact that he doesn’t think it qualifies as a real baseball statistic. If he has such a problem with it, then I feel he should either complain about QB rating and PER as well, complain about the fact that another colleague, La Velle E. Neal III, was guilty as well of using OPS, or just keep his mouth shut.*
* I was holding back on my true feelings of how I wanted to end that thought. But, I want to remain at least somewhat professional, even if this blog incorporates humor every now and then.
One last thing.* The title of Reusse’s Turkey of the Year article was, “Voters saw flight risk, so comeuppance was expedited.” Reusse doesn’t understand what the hell OPS means. Similarly, I don’t understand what the hell that title means.
* Edit: I lied. This is the last thing. The reason I’m so annoyed at Reusse is simply because of his ignorance. I wouldn’t have had a problem if he had said something like, “I don’t believe OPS is useful because of this and that and…” But instead, he simply called it a make-believe stat. No other evidence to support his opinion. It’s stuff like this that makes me really want to crank out a post on the battle between the sabermetric mindset and traditional scouting, because just like politics, if you get a bunch of outspoken extremists from both sides arguing with each other that the issues are black and white, you end up pissing off the people like me that believe in the gray areas. I know it appears that I lean more towards sabermetrics here at OTM, but I did also get excited to see Joe Mauer hit .400 in his first month of ’09 or that Justin Morneau cracked 100 RBI right before his season ended in mid-September. It is possible to believe in the gray areas, and I want to show that.