Measuring Consistency

[Irrelevant Dude]

After seeing may of the Cubs bats go cold in the playoffs, I started wondering whether there are any statistics that measure consistency in performance. It sure seems that players like Rizzo and Bryant have had huge up and down swings throughout the season. I would imagine that most players follow a similar pattern, but are there any stats that would measure the variance in a player's performance over the course of seasons? In other words, can we identify a guy whose rate stats were very stable over the entire season versus someone who fluctuated up and down to arrive at the same result?

 

I don't know how much value, if any, there would be in such a measurement, since there are many other factors like the park, pitcher, weather, etc. that could impact results. I am really just curious whether anything exists, or has been attempted, to measure hitter consistency.

[Cubswin11]

I've always been interested in this too, trying to find out if there was a way to identify low variance vs high variance players, but I think it would be incredibly hard to do. It's almost something you'd have to wait to identify after years a player was in the league or even retired as you'd need multiple seasons of data.

[Backtobanks]

I don't have any scientific proof, but after watching over 60 years of baseball, it sure seems like the streakiest hitters are the sluggers with high strikeout rates. When they're hot they'll hit 4-5 HR in a week and then strikeout 10 times the next week.

[Guest]

I mean asset managers measure volatility everyday. Just have to figure out how you treat statistics more like a portfolio. I'm sure there's a way to do it and now I know what I'm gonna waste my afternoon on.

[Guest]

Would it be possible/useful to chart a player's accrued WAR for each day of the season on a line graph? A lot of peaks and valleys would mean inconsistency, whereas a relatively straight line would mean consistency.

[benzona79]

On a slightly different topic, here is a very interesting article on the "science of hot streaks" http://www.thedailybeast.com/articles/2015/10/25/the-science-of-an-insane-hitting-streak.html

 

Basically it argues that contrary to the stat-enthusiast viewpoint, hot streaks are not simply the coincidental stringing together of excellent at bats or whatever, but a hot streak can have predictive value as to how a player will perform in subsequent at bats. I would imagine that the opposite is true as well, that cold streaks have predictive value as to subsequent performance as well.

 

This is nice, because it sort of confirms what watchers would swear up and down - that of course Daniel-Freaking-Murphy was going to hit bullets, just like Rizzo was definitely destined to weakly dribble one to the second baseman this post season.