On-base percentage has long been underrated by a large portion of the baseball world, from fans to writers, all the way up to players and coaches. Most are much quicker to cite batting average, and then completely jump OBP on their way to home runs or RBIs. This is fine, for those who are fine with a cursory perspective on the game, but many have begun to evolve beyond that, particularly those running things in MLB front offices.
In reality, the majority of these GMs actually have their own detailed metrics and projection models based on data and formulas not available to the public, but on a more basic level, they value OBP more than they do straight batting average. Some, it seems, also value the ability to get on base more than they do raw power. This would include new Mariners General Manager Jerry Dipoto, who has not been shy in making it known he values the on-base and speed profile highly, particularly in Safeco Field. But this idea can also be traced around to others, most famously Billy Beane who made patience and the ability to draw a walk a key piece of his Moneyball ideology.
But how advantageous is this idea? Is the ability to get on base actually more important than the ability to hit for extra bases, home runs in particular. Of course we know that you want as much of both as possible in an ideal world, but breaking it down, which one matters more? To study this, I first ran a regression analysis between OBP and runs scored, and then ISO (isolated power, which measures raw power by subtracting batting average from slugging percentage) from every team from 2011 through 2015).
Above is a trendline comparing OBP to total runs scored for all 30 teams since 2011. For those unfamiliar, r^2 is, in basic terms, the correlation between the two metrics. A value of .6646 means there is roughly a 66% correlation between OBP and runs scored, or in other words, getting on base explains 66% of the variance in runs scored. This is fairly significant, and you can see from the chart above there is a pretty clear relationship between the two. The more you get on base, the more runs you score. But is it more significant than power, in the form of ISO?
Above we see that ISO is slightly less predictive of runs scored, with the r^2 value down at .5837. That is certainly still meaningful, but it means it only explains about 58% of the variance in runs scored, compared to 66% for OBP. This more or less confirms the idea that OBP is more important to runs scored than power, but this is something we have actually been pretty sure of for a while. It is actually one of the reasons some of us statheads don’t like OPS as much as something like wOBA or wRC+. OPS isn’t bad, and certainly gives a more complete picture than just batting average would, for example, but it also treats OBP and SLG% as equals, when in reality they aren’t. OBP is more important.
We can extend this thinking beyond math, though. As Dipoto said recently during an interview with ESPN Seattle, OBP plays “whether you are on the moon, or in the depths of hell.” There aren’t many, if any, parks in which having an OBP skillset is a detriment. There are those that are simply less conducive to scoring runs, but there isn’t really an OBP equivalent to what a place like AT&T Park in San Francisco or Safeco Field in Seattle do to home runs. It may not be as exciting as the long ball, but getting on base is the best way to consistently put up runs.