Fascinating New Fielding Metric

by windyfelix on Mar 17, 2009 7:41 AM PDT

Over on hardballtimes.com today-- based on actual GameDay reporting from 2008. I'll leave it to others to judge its merits-- these attempts to quantify fielding skill are all crapshoots, but here are the most pertinent results. Data is based on per 150 chances.

1B Berkman led-- Daric Barton 5th. We are clearly losing something defensively with Giambi or whomever instead of Barton;

2B Kennedy led-- Ellis 4th.

SS Aviles led-- Cabrera 15th, Crosby not so far behind at 19th. The move was made out of frustration at Crosby's bat.

3B Rolen led-- Hannahan 5th!! That was a surprise to me, and may speak volumes about who the final backup infielder will be

LF-- Jay Payton (!) led-- Holliday 11th, Cust dead last at 34th

CF-- Jody Gerut led-- no A's qualified Kotsay 29th-- how the mighty have fallen

RF-- Gutierrez led (great sleeper fantasy pick, by the way) Sweeney was 10th.

If Hannahan is this good, then I don't see why we'd keep Crosby around, given Cabrera's durability. And we have to have Cust DH as much as possible, which does sacrifice defense at 1B. It's either 1B or RF-- pick your poison. Barton could still make the team if they decide to go with just 4OFers and 6 relievers. But it seems doubtful.

Tweet Comment 16 comments | 0 recs |

Do you like this story?

Comments

Crosby?

Money.

I could prove God statistically. Take the human body alone - the chances that all the functions of an individual would just happen is a statistical monstrosity.
~George Gallup

by UncleLeo on Mar 17, 2009 8:42 AM PDT reply actions

Any metric that doesn't have Juan Rivera last

still needs some refining. That guy has a bright future ahead of him as a butcher.

I like Cindi. A. She never pretends to know more than she does. B. She has unbridled enthusiasm for her "Hotties," and isn't afraid to show it. -IM4Oakgal

by Nico on Mar 17, 2009 9:26 AM PDT reply actions

scutaro #2 SS

David Wright dead last

by Future Ed on Mar 17, 2009 11:28 AM PDT up reply actions

What is the actual methodology?

What is being measured? Where is the link explaining the metric? This needs to be edited.

"To this day and dating back 25 years, before every game he plays, Henderson stands completely naked in front of a full length locker room mirror and says, "Ricky’s the best," for several minutes."

by VORP is too nerdy on Mar 17, 2009 10:00 AM PDT reply actions

link

here

by Future Ed on Mar 17, 2009 11:20 AM PDT up reply actions

What's it called?

Seriously. If you’re going to make a fanpost about something being discussed elsewhere, provide a link.

m*****f***ing c***s***ing peanut butter and jelly!! f*** f*** f***!!!

by JediLeroy on Mar 17, 2009 4:13 PM PDT up reply actions

Gutierrez is a terrible fantasy player...

all his value is in defense. He’s not a good hitter, although he’s tolerable against lefty starters.

Many years from now, when his name's recalled
Everyone will say, "He should have passed the ball"
-- Al Stewart, "Football Hero"

by PaulThomas on Mar 17, 2009 10:55 AM PDT reply actions

I'm not sure I buy some of this

Ryan Howard slightly better than Kevin Youkilis?
Michael Young better than Troy Tulowitzki? Omar Vizquel? Jimmy Rollins?
David Wright and Ryan Zimmerman are among the worst 3Bs in baseball?

I'm starting to like our bullpen......Wuertz and all.

by scatterbrian on Mar 17, 2009 11:35 AM PDT reply actions

Cust being the worst LF?

I do agree though, not looking at the numbers but just looking at the rankings suggest a problem. THe lists do agree with CW for the most part, but the outliers provoke more questions.

by Future Ed on Mar 17, 2009 11:55 AM PDT up reply actions

I think this analysis proves one thing, and one thing only

We really don’t have much of an idea how to measure defense.

Also, in a shout out to PT, it certainly suggests that Ellis isn’t the best defender relative to his position in baseball

"There's never enough time to do all the nothing you want" -Bill Watterson

by nevermoor on Mar 17, 2009 12:51 PM PDT reply actions

I disagree with your premise 1 billion per cent

We absolutely have an idea how to measure defense. There’s really no debating that. The only problem at this point is that batted ball data isn’t perfect, but it’s advanced to the point where the good metrics are fairly reliable in reporting what happened (though defense needs to be regressed a lot, so what happened isn’t always indicative of true talent).

This metric is quite obviously not good because they use Gameday data. THT even admits that Gameday data sucks. It’s better than nothing, but not as good as other options.

Are the Gameday data as good as STATS or BIS data? No. I have already discussed the problem of Gameday giving the location of where the ball is picked up instead of where the ball lands on base hits. But in addition some fielding locations are missing and others seem to be clearly in error. The Gameday data had to be normalized because of differences in the hit recording diagrams and the normalization process will create errors.

That doesn’t mean that the information is useless. It’s just not as good as UZR or a couple others.

For anybody really looking to analyze defense, the best bet is to take all the metrics worth anything (and this appears to be worth something) and figure out a way to weight them and come up with an aggregate result. With that, we can figure out who is good defensively about as well as we can figure out anything else. I don’t really understand why people assume defense is so hard to analyze. It’s really not anymore.

by thejd44 on Mar 17, 2009 1:53 PM PDT up reply actions

The problem with defense analysis

Is that we try to “average out” so many things that what we are left with is far less helpful than hitting data. For example, does a player’s zone change when their alignment is altered? Who determines these zones, and based on what (this stat’s description made the selection sound basically random/hand drawn)? How is a 1B’s ability to corral bad throws measured? How can we measure C defense at all?

I think that until we can focus in on every play with the sophistication with which we can focus in on every at bat (and despite your assertions to the contrary I am not convinced we are anywhere close), defensive stats have a LONG way to go.

"There's never enough time to do all the nothing you want" -Bill Watterson

by nevermoor on Mar 18, 2009 9:06 AM PDT up reply actions

ball speed off the bat

height and trajectory, where the ball hits the ground, spin, cut of the grass, field shape, weather, dew, hardness of the ground.

with a large enough sample size for everyone maybe, but each event is so particularized ,I doubt we can get to a large enough sample.

by Future Ed on Mar 18, 2009 12:00 PM PDT up reply actions

Just because the metrics aren't perfect does NOT mean we have no idea how to measure it

You’re making two separate arguments here.

You can be not convinced all day long, but you’d still be wrong. Even without Hit f/x, we’re close. With hit f/x, there’s not really even an argument to be made.

by thejd44 on Mar 19, 2009 12:35 PM PDT up reply actions

Thanks for answering zero of my questions

Strong argument.

Although you are on to something about separate arguments. I will concede that “no idea,” or at least “don’t have much of an idea,” is too strong. I’ll replace “are nowhere close to as accurate” if that makes you happier.

"There's never enough time to do all the nothing you want" -Bill Watterson

by nevermoor on Mar 19, 2009 1:49 PM PDT up reply actions

I'm not an expert on advanced fielding systems

but it’s tough to get excited about a system that’s basically, as far as I can tell, very similar to several other fielding systems with bigger zones and inferior source material. Still, the more coherent fielding systems out there, the easier it is to see each system’s biases, so that helps.

by Josh Deletchi on Mar 17, 2009 8:29 PM PDT reply actions