Respeck. We’re back to our chat about stats and our first real topic: linear weights. As a refresher, I finished up the intro post with the some goals:
- Figure out which baseball player is responsible for which baseball thing
- Figure out how much that baseball thing helps that baseball player’s team win more baseball games
Lets focus first on batters, as they will be easier to deal with the first issue above.
We’ll start with things batters aren’t responsible for. They’re not responsible for the ballpark where they hit half of their games, so we’ll have to deal with that later. They’re also not responsible for the skill of the defenders or pitchers against them, though those defenders and pitchers rotate throughout the league as hitters play against different teams. So while not perfect, we can feel comfortable saying defender skill isn’t a big enough issue to deal with here. Finally, batters aren’t responsible for the number of baserunners when they come up to bat, so we won’t need to worry about them.
Batters are responsible for the types of hits and other outcomes of their plate appearance.
Towards the end of his 2013 campaign, Padre second basemen Jedd Gyorko saw a marketing push focusing on his 23 home runs, which lead rookies across all of baseball. A nice piece of trivia, but if you’re the kind of fan who wants your players helping the team win, the fact alone becomes less helpful considering all the other things a hitter must do successfully. Fans, including media analysts and guys getting in arguments in bars often face the difficult challenge of comparing one thing a hitter can do with all of the other things a hitter should also do.
I’m excited as I’m typing this, because I get to tell you about such a simple and beautiful method of measuring different outcomes of plate appearances. Say what you will about things that excite me (TWSS), but holy crap it’s awesome we’ve figured out the value for everything a batter does so specifically.
Baseball actions must lead to either scoring or preventing runs, so we’ll convert each action into the equal amount of “runs” it is worth. This allows us to compare the value of different things consistently, and in the form something that directly helps the team win. Here’s the run value of a few things thing a batter can do:
Home Run: 1.39 runs
Triple: 1.09 runs
Double: 0.77 runs
Single: 0.47 runs
Walk: 0.32 runs
Stolen Base: 0.16 runs
Pretty incredible huh? We call them linear weights, and they make some intuitive sense if you think about them. A home run, for example is guaranteed to be worth at least one run since the batter will always score. But there’s also a chance someone could be on base for the home run. So a home run is worth a little more than that, based on how often guys are on base in front of batters.
We can now say the Gyork Store’s 23 home runs were worth about 32 (23 x 1.39) runs. Yasiel Puig didn’t hit as many home runs as our boy Jedd, but he did enough other things to help the Dodgers on offense more than the Padres. Puig’s 19 home runs were worth about 26 runs, but his additional singles, triples, walks, and (oddly) hit by pitches unfortunately helped his team more.
This is all possible through the magic of the run expectancy matrix. It takes the 24 possible situations of men on base and the number of outs (bases empty no outs, man on second 2 outs, etc.) to figure out how many runs a team can expect to score. The run value of many other events can be found here. Then you can get really dirty with baseball legend Tom Tango if you’re into that sort of thing.
While calculating each of these items individually for every player ever sounds like a raging good time, we’re lucky enough to live in a world where someone has done the work for us. We’ll get into that in a future post.