I think this will give them credibility.

http://sportsillustrated.cnn.com/mlb/news/20140521/mlb-advanced-media-defense-tracking-new-york-mets/

The example in the article would seem to neuter any ability to generate an advanced individual defensive statistic.

In the Quintanella play, the only data that can be taken away is that he was out of position to make the play; if he was in position, he might have made the play, but we don't know that he would have.  I would assume that across all data points the fault for him being out of position would be placed on Quintanella himself... but a major point of the article is that the captain of the IF (Murphy) had him out-of-position.

Maybe I'm not looking at this the right way, but it seems to me that this just points out how many outside factors come in to fielding that can't really be captured in an equation to determine who is adding more or less value in the field.

That thought occurred to me but you should be able to tell how quickly a player can react and how quickly he moves once he does react.    Beyond that, it will measure how hard the ball was hit, how fast the runner ran and how hard the fielder threw the ball.    Using this data, you should be able to generate averages.   After doing that, you should be able to know which players are above/below average. 

But, yeah, it throws speculation out the door.   Rather than saying "75% of the SS make that play", you'll have to say "His poor jump, which was well below his average, and poor positioning is why he failed to make that play." 

As I said, defensive stats will have some credibility now. 

statnerds just getting home, removing the pocket protectors and will start the retard-slap fight @ 5-6pm

I'd be interested in knowing where the human inputs into the system are.

It seems like there are just too many variables in a given play for the system to automatically know what information to store, track, calculate, look-up, etc. for each and every play.

I wouldn't think there are any.   MPH is measured by a radar gun.   Distance and time by camera.  Seems pretty cut and dry.   Without the human inputting ".04370 because that's our best estimate to make this stat work."

Posted by MikeT23 on 5/27/2014 4:19:00 PM (view original):
I wouldn't think there are any.   MPH is measured by a radar gun.   Distance and time by camera.  Seems pretty cut and dry.   Without the human inputting ".04370 because that's our best estimate to make this stat work."

Not quite what I mean.
I mean, is all of the meaningful data (i.e. the player's reaction time, speed, etc.) automatically calculated and stored for each play, or does a human have to pull up a play to analyze.  

I don't see why human input would have to be involved.

It seems like it's going to be pretty cool.  I've heard about it for a while now, just waiting to see it in action.  I think it'll be fun for the OF especially - see how much ground different guys can cover, analyze routes to ball, etc.

That was pretty amazing. It would be a cool add to the every day broadcast!

I saw this and think it is awesome...now add runner info...how fast was the guy running, what type of a jump did they get out of the box...etc. Way more involved than one can simply analyzing through the naked eye.