skyking162

Do baseball stat analyses consider run variance?

This is one I?ve been rolling around in my head since the NCAA tournament. (bear with me)

I was thinking about this last year as West Virginia went on a 3 point tear in the tournament and Vermont upset Syracuse. A team that relies on the 3 pointer is likely to be one that has a high variance in their average score, they could go on a tear and score 100, they could be cold and score 50. Their average score may be 80 and playing against a team with a scoring average of 60, but in a ?one and done? situation, if . My buddy and I discussed graphing these scoring bell curves for each team to decide the likelihood that an upset could occur (the area where they overlap). Clearly, being the type of guys we are, this idea fell by the wayside and we picked based on conventional wisdom (and failed miserably in out pools)

Ok, so back to baseball.

So I?ve seen the numbers that say teams that employ a ?Wait for a 3 run HR/Earl Weaver? strategy have higher average scoring than ?small ball? teams, but I offer that teams that can ?manufacture? a run have a lower variance in their scoring, thus giving them a much greater chance of winning that it would initially seem.

Again, I have no actual proof of any kind that ?small ball? teams have a lower run variance, just something I thought you might know. Also, the likelihood that I have missed something in my 5 minute analysis is moderate to certain. please feel free to tear this to shreds.

A few thoughts…

Yes, worse teams with greater variance are better than similar scoring teams with lower variance (and the opposite is true for good teams). It’s probably not considered as much as it could be.

Baseball scores are weird because you can’t really treat them as continuous like you could basketball or fantasy football. That discreteness plays havoc with using traditional statistical distributions.

I would guess that on average, as run scoring goes up, so does variance. So if you wanted to do a controlled study of history, you’d need to find teams with similar run-scoring averages but different styles. Small ball teams that score 900 runs per year and 3-run homer teams that score 600 runs per year would be in the minority. (Which is an interesting observation itself.)

Instead of taking a large-scale view of the question, you can take a small-scale view and use win-expectancy tables. With 2 on and 1 out in the 5th inning, you can compare the expected value of swinging away with Joe Shmoe to putting down a sacrifice bunt. Sometimes it’s the right thing to do, sometimes it’s not. Win-expectancy tables are better than run-expectancy tables becaue they take game-situation into play. For example, the run expectancy after an intentional walk is almost always larger than it was before the walk. But in the context of the game, many of those same intentional walks pretty much maintain the same probability. There are times to bunt (1st and 2nd no out being the best) or not. I’d guess that small-ball teams make more errors in judgment than 3-run homer teams (because my personal guess means so much.)

Teams that play smallball throughout the game tend to play more close games than 3-run homer teams. Thus they’ll have more dramatic wins and seem to be more clutch. They might even have a small edge in close games. Playing a 3-run homer style might decrease the close wins, but lead to more big wins relative to the big losses.

This psychological affinity for small ball style is related to the hatred of the strikeout, I think. Even though Adam Dunn hits the ball 400 feet 40 times a year, people don’t want to think about that when he strikes out YET AGAIN. If only Dunn hadn’t struck out, he could have put the ball in play and had a shot at a hit. But you can’t have both the homers and the contact. (One of my favorite statistical observations is that strikeouts are POSITIVELY correlated to player value.) And in team style, you can’t have both the ability to score big and play lots of close games that you win. The second is sexier, but the first probably more successful. Of course, it’s really the pitching that wins championships.

Popularity: 2% [?]

This entry was posted on Sunday, October 30th, 2005 at 9:10 am and is filed under Baseball. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.