Go back to my homepage

Just what is WAR?

This page assumes you know the basics of baseball. If you don't know the rules, this page is pretty good at explaining.

Sabermetrics made easy

The goal in baseball is to score more runs than the other team. Thus the premise behind calculating player value is to judge how many runs they can positively contribute. Somewhat arbitrarily, every 10 runs you produce is considered one "win".

How do we produce runs? Of course, the most obvious method is to do so by hitting singles, doubles, walking, etc. Each of those actions will create an increase in the amount of runs you produce. (For further reading, check out wOBA)

How else do we produce runs? Defense, while in reality cannot produce runs, is treated as being able to produce runs by taking away runs. So if a player is fast and can reach balls hit far away from him and thus a huge zone that player produces more runs. Keep in mind this is different from just looking at errors. This is because if you had no legs and were incredibly solid fielding from where you were you'd have no errors at all, but you'd be a very bad fielder. Errors in general don't take into account how far a fielder moved for the ball, so sabermetrics accounts for your zone.

Finally we can produce runs one more way- baserunning. Baserunners who steal a lot and get caught stealing very little are obviously valuable. But it's not only about stealing- runners who go from 1st to 3rd on a single are more valuable than runners who only move 1st to 2nd. You lose value if you are only able to go from 2nd to 3rd on a single instead of scoring, etc.

Finally there are things you can do to gain negative WAR (in other words, lose runs). You could make errors or strikeout, for example.

So that's WAR in a nutshell.

There are some neat conclusions from studying baseball though. For example, two singles in 2 PA are much more valuable than a double in 2 PA. The reason for this is because 2 singles produce no outs, whereas you have to subtract 1 out from the double (outs have negative value).

What's wrong with traditional stats?

First off- none of these stats are "bad" they just don't give you the whole picture.

RBI: The reason why this is bad is because it's mostly a product of opportunity. If you have a bunch of good players in front of you, you'll have more RBI chances. This right off the bat lets you know this stat isn't fair. Playing on the Astros would affect you negatively when it comes to RBIs, for example.
The 2nd reason why it's bad is because it's a counting stat and doesn't really give that much insight on how good a player is. A guy with 30 RBIs in 600 PA is obviously worse than a guy with 15RBIs in 30 PA, yet the "30" is bigger than "15".
When RBI can be good: RBIs are good at the margins. For example, if you played a full season and had a bunch of RBIs (say... 200), then it's obvious you're good. There's just a limit to how many chances you can possibly have, so being able to have 200 RBI is obviously an accomlishment (one that would take a lot of luck too).
BA: The whole reason why BA sucks is because walks are almost as valuable as singles. Remember, as long as you don't produce an out, you produce positive value, and if you are out you produce negative value.
When BA can be good: BA is also good at the margins. If you have a .400 BA, EVEN IF YOU DON'T WALK AT ALL, your on base percentage is good. Basically, 40% of the time, you're not hurting your team by creating an out. And that's pretty good.
HR: Home runs are somewhat good. But they are far from telling you the whole story. Remember, 2 doubles are worth more than 1 HR. HRs can be good as long as you know the number of walks, doubles, etc. and the number of PA.