A bit of a random post

I’ve been doing a lot of work with fielding recently, and one thing I wanted to mention is just how well simple fielding stats can serve you. Fielding%, for example, is often criticized and with good reason. However, I’ve found that for infielders, if you remove putouts from the equation to create a modified fielding%, you can learn a lot. That is, A/(A+E) is a generally good measure of fielding talent at second, third, and shorstop, correlating with Zone Rating (which, of course, is much, much better) at about .5. Such a correlation is pretty solid; for example, Baseball Prospectus fielder ratings, which are much more complicated, correlate at about .6 with Zone Rating. Fielding Win Shares, also much more complicated, correlate at roughly .5. If you were to transfer this metric into the outfield and use putouts, it would not work well, mainly because errors in the outfield are too rare to tell you anything. But for infielders, there is nothing wrong with using A/(A+E) for a quick measure of their value. Generally speaking, it works.

Advertisements

6 Responses to A bit of a random post

  1. Kenny says:

    I’m not so sure that comparing correlations with ZR is the best way to compare the usefulness of a modified F% as comparied to BP’s more advanced metrics. It seems to assume that ZR is the best we have, and measures the others by how well they measure up to ZR. I think BP’s metrics are supposed to be better than ZR and deal with issues that ZR cannot (though, admittingly, I’m not an expert on defense stats by any stretch of the imagination). But to me, doing correlations with ZR like this is kind of like correlating various offensive measures with team OBP, say, instead of team runs scored, and then concluding (hypothetically), that batting average is as good as OPS for offense.

  2. David Gassko says:

    ZR is much, much better than what BP does. My Range metric is probably equal to ZR. BP’s stuff if inferior.

  3. Kenny says:

    Would you be able to briefly explain the advantages of ZR over FRAR/FRAA or Rate2? (If it’s too complicated, don’t worry, or if you have a link that would be great.) Thanks.

  4. David Gassko says:

    It’s simple:
    Zone rating uses actual play by play data. While it has some problems with how it treats balls out of zone, ZR tells us the actual amount of balls in play a player was responsible for. Any non-PBP metric can only make estimates, and BP’s estimates don’t happen to be very good. I like BP’s defensive numbers as a simple barometer of defensive skill, and the historical ratings are fun. But ZR is clearly better.

  5. Kenny says:

    Thanks, I had actually figured BP used PBP data, though I never thought much about who from BP would have the time to watch all the games or where they’d get the data from.

  6. Kenny says:

    Thanks. I guess I should look more into how BP calculates their fielding stats. Unfortunately I think much of it is proprietary. I realized ZR used PBP data, but wasn’t quite sure whether BP did or not.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: