Sunday, May 25, 2014

OPS at the Park

Saturday night I was at the Reds game against the Cardinals. On the big scoreboard they displayed BA/OPS/HR/RBI and on the ribbon boards they displayed BA/HR/RBI/OBA/SLG/R (I didn’t write this down, so I might be off a little on the categories, but I think the preceding is a pretty fair representation). In any event, there was a group of four people sitting in the row in front of me. It appeared to be a father, two adult daughters, and one of the daughter’s boyfriend/husband.

Late in the game, the gentleman asked the group what “OPS” was. The boyfriend said something to the effect of "It’s a stat that’s supposed to tell you how good of a hitter a player is", but didn't seem to know any details. This being 2014, he got out his phone and looked it up. It appeared that this took him to the Wikipedia page, which after explaining that OPS is the sum of OBA and SLG , says it can also be computed in one equation as follows:

OPS = (AB*(H + W + HB) + TB*(AB + W + HB + SF))/(AB*(AB + W + HB + SF))

He proceeded to show this to the rest of the group, and the reaction appeared to be "wow, that’s really complicated". It wasn’t "that stat must be bunk because it’s complicated", but seemed to be more of a “I’m not interested in learning more about this because it’s too complicated”.

In observing this I started thinking about what I might have said was I somehow involved in this conversation. Would I have thrown OPS (a metric about which I have written negative things about more times than I can count) under the bus by agreeing that it was too complicated, all the while concealing the fact that my preferred metrics might appear to be even more complicated to a layperson? Would I have said that the equation above was more complex than it needed to be, because it could be written as the equivalent and decomposed (H + W + HB)/(AB + W + HB + SF) + TB/AB--hand-waiving away the fact that the more complex version provides insight into how events are actually weighted by OPS? Would I, not wishing to contribute to any anti-sabermetric sentiment but also with no desire to teach a saber 101 course, have offered a lame defense of OPS?

Of course, I wasn’t involved in this conversation, and certainly wasn’t going to insert myself into it, so this was all hypothetical. But I think the thought exercise is worthwhile to consider for those who do have an interest in what I’ll call for lack of a better term sabermetric evangelization. I am not one of them, but if you are interested in getting more casual fans interested in sabermetrics, you should consider whether you could explain the metrics that you are advocating if asked to do so. I would draw a distinction between a metric that is complex in order to achieve precision or theoretical considerations, and one that has properties that defy explanation. If you think you could explain why total bases are weighted by plate appearances and times on base are weighted by at bats, then by all means, go ahead and point people to the Wikipedia page for OPS. Good luck.

1 comment:

  1. Not being a shy person, I would have spoken up and said, "On-base average is how often a player gets on base which is additionally reflective of how often he doesn't make outs. Slugging average is how much power a player has, i.e. how good he is at driving in runs per at bat. OPS is the combination of those two numbers, which as the one guy says - is an overall decent description of how well that player has been hitting. It is a better indication than either component stat alone - and either of those component stats are a better indication than batting average alone." I would have left it at that, unless they asked more questions.

    ReplyDelete

I reserve the right to reject any comment for any reason.