clock menu more-arrow no yes mobile

Filed under:

Preface: BTB Power 30

A simple estimation for crediting runs scored is the following:

Runs = AB * SLG * OBP

That's the simplest way, in my opinion, to estimate the number of runs that you're producing. There are some flaws. For example, if a guy has 1 at bat and hits a homer, it's worth 4 runs under this system. It's a fairly safe estimate, however, for team runs scored; back in 2004, it was, on average, 25 runs away from the team's total runs scored, and its worst estimate was only 60.4 runs less than the D'Backs total. It also tended to be a little less than total runs, but I would attribute that, at this point, to the lack of "smallball" included in it.

What else could skew the estimate, in my opinion?

  • Lots of stolen bases, and an abnormally high or abnormally low steals rate.
  • Good situational hitting.
  • Luck.
Various run estimation methods incorporate stolen bases and other factors, making them more accurate than AB * OBP * SLG.

None are simpler than AB * OBP * SLG, however. That's the premise for my ranks system in the Power 30 that I put out every week. I don't exactly use AB * OBP * SLG and Pythagorize it; I scale them equally to the average and then rank the teams that way.

For my ranks, I don't want something as deep as Pythagenport's third order records, which goes into strength of opponent players and adjusts the equivalent runs and converts those into wins. I'm not quite looking for that... I want a different approach. I don't want the ranks to resemble Pythagenport. I want to rank the teams who seem to be getting lucky lower than the others.

But I would like to incorporate speed into the system. While situational hitting is largely variable and luck-dependent, and, if not that, largely inconsequential, stolen bases can add runs.

I played around with XR and a few other run estimation methods, but then it occurred to me that they were all far too similar to the Pythagenport third-order ranks, with Equivalent Runs.

Long story short: I didn't make any changes this week, after much evaluation and experimenting. The Scotto scores are pretty consistent with the overall ranks, interestingly enough, and I looked at the small-ball attributes, and they seem largely neglible. Maybe the stats are missing something. We'll see at the end of the year, I guess.

The ranks are done, but I haven't written anything next to the teams. I will complete the brief captions tomorrow.

The brief spoiler is that the Orioles maintain the #1 spot for another week.