The Baseball Stat to End All Baseball Stats

Batting average, earned run average, on-base percentage, strikeouts per nine innings, wins above replacement, walks plus hits per innings pitched…

Baseball is the sport for statisticians. Statisticians design each new measure of baseball prowess to prove that good players are actually good, and it wasn’t all just luck.

The holy grail for baseball stats has always been to find the single number that sums up the entirety of a player’s abilities. First, it was the “Triple Crown” stats of batting average (AVG), home runs (HR), and runs batted in (RBI) for hitters, and earned run average (ERA), wins (W), and strikeouts (SO) for pitchers. The basic stats evolved to more advanced statistics like on-base percentage (OBP), slugging percentage (SLG), and on-base plus slugging (OPS) for hitters, and walks plus hits per inning pitched (WHIP) for pitchers. The current emphasis on Sabermetrics such as Wins Above Replacement (WAR) attempts to further condense a player’s value to the team into one statistic.

Sabermetrics tend to focus on the number of runs or wins that a player creates. That is where my thinking starts to deviate. When you think of the core essence of baseball, it’s all about the bases (hence the name “base” ball). A batter’s goal each time he goes to bat is to touch as many bases as he can. A pitcher’s goal each time he pitches is to deny hitters as many bases as he can. At the most basic level, a universal stat for baseball would show how many bases a hitter reached. A comprehensive stat for pitchers would show how many bases the pitcher denied hitters.

Baseball has many nuances, so such a statistic can become cumbersome to calculate. For a hitter, it could include total bases, stolen bases, bases reached on error, walks, hit by pitch, and others. You can also take into account more subtle things like sacrifice flies, hitting into double plays, being caught stealing, committing fielding errors, and others. But I think we can ignore many of these nuances in all but the most extreme cases. For my purposes, I keep it simple. For hitters:

TBavg = (TB + BB + SB) / (AB + BB)

For pitchers, I look at how opposing batters fare against them (little “o” stands for “opposing”):

oTBavg = (oTB + BB + oSB) / (oAB + oBB)

A more accurate statistic would take things like fielding, wild pitches, and balks into account, but unless the player is an outlier, these stats tend to be negligible.

Now, a problem we still have is that we can’t compare a batter’s TBavg with a pitcher’s oTBavg. Good hitters have a high TBavg, while good pitchers have a low oTBavg. To allow for comparisons, we can adjust these numbers against the league average to form a “combined” TBavg. Let’s call the combined TBavg something cool like “Dominance”. So for hitters:

Dominance = (TBavg) – (league TBavg)

For pitchers:

Dominance = (league TBavg) – (oTBavg)

With the Dominance statistic, we can compare batters with pitchers. The higher the Dominance, the better the player is compared to average players. A player with a negative Dominance is weaker than the average player in the league.

Dominance is measured by averages. It shows which players are excelling above other players on average. But it doesn’t measure total contribution. A player who hits four home runs in four at-bats before having a career-ending injury would have a much higher Dominance than a player who hits 50 home runs over the course of a whole season and leads his team to the World Series.

One tweak that can fix this problem is to measure the hard numbers. Don’t average them out. Let’s call this “TB+”.

TB+ = (TB + BB + SB)

TB+ is a measure of total offense from the hitter. It is similar to the Sabermetric WAR, which measures total “wins”.

Determining an equivalent TB+ is not so simple for pitchers, though. We would probably have to award the pitcher with a certain number of points for each batter faced. Then we subtract points if the result is anything other than an out. Then we would likely have to normalize the results somehow so we can compare pitchers with batters. I haven’t finished working this out yet, but you get the idea of how it would be done.