- FanGraphs Baseball - http://www.fangraphs.com/blogs -

How Should We Measure Power?

What exactly is “power”? Is it the ability to hit home runs? Doubles? Triples? Should we consider how far a player hits a ball, or are we just concerned with the outcome? How would you define it?

If we were to try and define power from the ground up, obviously you’d have to start with home runs. Power hitters are guys that mash lots of home runs, right? When I think power, I think of players like Jose Bautista, Babe Ruth, Hank Aaron, and Barry Bonds. Home runs are so flashy, they steal the show.

But there’s more to power than a player’s raw home run total. You can’t completely ignore other extra base hits, which is why there are statistics like Slugging Percentage and Isolated Power. Slugging Percentage measures a player’s total bases and Isolated Power measures a player’s extra bases*, so both statistics count doubles and triples as well as home runs.

*Quick refresher course for everyone. Slugging Percentage = Total Bases / At Bats ; Isolated Power = Extra Bases / At Bats

Or if you prefer to think about it another way, Jose Bautista has a .330 ISO this season. That means he averages nearly one extra base every three at bats. 

Both these stats have the same problem, though: not all bases are created equal. If a player has accumulated 30 extra bases in 100 at bats, isn’t there a big difference if those extra bases were accumulated through 10 home runs versus 30 doubles ? Both players have the same Isolated Power, but which one has provided their team with more value through their power production?

Good question, I’m glad you asked.

When trying to answer this question over at DRaysBay yesterday, I decided to bust out the weights included in wOBA. After all, Weighted On-Base Average weighs each type of hit in proportion to their actual importance for run scoring, so we should be able to use those coefficients to isolate power production. Here’s what I came up with:

wXB = (1.268 * 2B) + (1.610 * 3B) + (2.086 * HR)

I call it Weighted Extra Bases. It’s very simplistic (and not translated into runs form), but it essentially shows us how much value a player produces through their power. It’s using the same exact concept as both Slugging and ISO, except in this instance using coefficients that relate to value, not bases. Here’s a quick leaderboard for 2011:

This is a counting statistic, though, since players with more playing time have more chances to get extra base hits. Jose Bautista may look like he comes in second fiddle to Curtis Granderson, but once you adjust for playing time and opportunities, the difference vanishes. To adjust, let’s divide wXB by at bats:

For those of you following along at home, you may notice that those values look very close to ISO…and that’s because the formula’s for the two stats are very similar. Consider:

wXB/AB = [ (1.268 * 2B) + (1.610 * 3B) + (2.086 * HR) ] / AB
ISO = [ (1 * 2B) + (2 * 3B) + (3 * HR) ] / AB

When you actually weigh extra base hits in proportion to their value, it turns out that ISO is undervaluing doubles, while overvaluing both triples and home runs. This is why a player like Matt Holliday (33 doubles, 19 home runs) can have the 9th highest wXB/AB in the majors, yet only the 20th highest ISO and 17th highest SLG.

In all honesty, this experimental stat makes me appreciate ISO all the more. I never used to like ISO that much, due to the reasons I stated above, but this shows me that for all its simplicity, it’s a very effective shorthand way of conveying a player’s true power. It does undervalue players with extreme doubles totals — and overvalue players with lots of homers and few doubles — but for the most part it’s rather accurate. The largest disagreements between wXB/AB and ISO only move a player 15-20 spots in the rankings, while the largest differences between wXB/AB and SLG amount to something closer to a 40 spot adjustment.

If nothing else, I hope this helps people remember to look closer at a player’s double total as well when talking about power. Doubles are a lot more valuable than we typically assume, and they are undervalued by even the best publicly-available power statistic (ISO). I’ll have to mess around with wXB a bit more, but maybe this is a concept we can build upon.

Here’s a Google Doc workbook on wXB. Enjoy!