Understanding sabermetrics
February 20, 2013
The Miguel Cabrera-Mike Trout battle royal for the American League’s Most Valuable Player award captivated the baseball world in 2012, with the argument boiling down to old-school versus new-school.
Cabrera dominated the old-school, winning the Triple Crown by leading the AL in batting average, home runs, and runs batted in (RBI), while Trout dominated in new-school statistics like wins above replacement (WAR).
Both sides of the debate were often arrogant, ignorant, and frustrating. To avoid such an annoying situation in the future, the public at large could use a better understanding of sabermetrics.
Sabermetrics is the name given to new statistics that were developed by people with a strong statistical background. The goal of sabermetrics is to figure out exactly how good a particular player really is. Those statistics are represented in runs, since the goal of every baseball team’s offense is to score runs.
If you break down every event that can happen in a baseball game – single, double, home run, walk, groundout, etc. – the approximate worth in runs of each play can be determined. Then, you can add together all the runs created by the player, thus figuring out how many runs a player is worth.
Luckily, turning runs created into wins is simple. Each 10 runs a player creates are equal to one win. So if Player A, creates 100 runs of offense, he is worth 10 wins. Easy!
Now, baseball is a sport in which even the worst team will win quite a few games – there is just too much luck involved. A team of average minor leaguers, or replacement-level players, will win around 45-52 games. That is terrible, but theoretically, no team can do worse than that. The goal for each team, then, is to find players who are better than replacement-level to push their win total above 52.
Back to Player A. If Player B is a replacement-level player and he would create 40 runs in the majors, then Player A is worth 60 runs more than Player B, or 6.0 wins above replacement.
This puts Player A’s contributions in context. He is worth 10 wins overall, but only adds six wins to that 52-win baseline for each team.
WAR also adjusts for the offensive level at each position. Shortstops do not hit as well as first basemen, so it has adjustments to put all positions on a level playing field. This way, the offensive contributions of each player are weighted equally.
After that, defense and baserunning are added. This allows WAR to come close to figuring out a player’s overall ability, since baserunning and defense are also big parts of the game. Sabermetric defensive statistics are not perfect, but they do a decent job of figuring out how good of a fielder a player is.
So WAR does a good job of evaluating players, but the most important thing to remember is that it is not blind gospel. If Player A is worth 6.0 WAR and Player C is worth 6.1 WAR, Player C might not be better in reality. WAR has its flaws, but despite the error involved, it does the best job possible of seeing how good a player is.
The key is to use WAR as one tool to aid in your own opinions. If Player A and Player C are close in WAR, then use your opinion from watching both of them play to decide who is better. Remember: WAR is a guide to help rank players, not an ironclad fact.
Now, if Player A has 6.0 WAR and Player B has 1.0 WAR, then you can be very sure that Player A is better. But if it is close, you have to use your own mind to decide who is best.
I know that WAR is more difficult to understand, but these sabermetric statistics really are better than many older stats. WAR has flaws, but they are not as big as others:
• Batting average ignores walks and treats a single the same as a home run,
• Slugging percentage counts home runs more, but it also ignores walks,
• On-base percentage counts walks, but it also treats a single the same as a home run,
• And RBI is completely dependent on if someone is on base, something completely out of his control.
Plus, none of those stats even try to count defense or baserunning.
So, WAR is not perfect. But as long as you combine it with what you see on the field, it is less flawed than things like batting average and RBI.
Contact Jim Piascik at [email protected].