The graphs at the end (in the PDF) show that the Yankees were well above average at fielding ground balls up the middle in 2006. I'd really like to see that explained as this has always been believed to be one of the Yankees' biggest weaknesses. The only places where they are significantly below average are the zones on either side of third base.

The graphs at the end (in the PDF) show that the Yankees were well above average at fielding ground balls up the middle in 2006. I'd really like to see that explained as this has always been believed to be one of the Yankees' biggest weaknesses. The only places where they are significantly below average are the zones on either side of third base.

Wang inducing lots of weak grounders that even Jeter can reach?

I expected to see two humps around the shortstop and second base positions, but I am surprised to see that double-peak structure on the left side.

John - I think this might be an artifact caused by the way hit locations are assigned by mlb. I have taken mlb hit location for HRs and matched them with HitTracker data for 2006. The angular measurements match very well in general but much better down the lines and directly up the middle. This is to be expected since these are easy reference points for the data recorder. In the alleys the uncertainty level rises from about +-2 degrees to +-5 degrees. My interpretation of the double humps is that the mlb data recorder is being influenced by which fielder fields the ball. If the left fielder or third baseman makes the play I think the recorder inadvertantly places the ball a little closer to 3d base and if the center fielder or shortstop makes the play I think he records the location a little closer to second base. On the right side I think the 2nd baseman makes the majority of plays in the gray area so there is no corresponding double hump. Just a theory.

This seemed to fly under the radar around here, as it was posted on a weekend, but I tried a "back to the basics" approach to fielding using retrosheet data. I don't have detailed hit location in mine.

Part One


Part Two

This is to be expected since these are easy reference points for the data recorder. In the alleys the uncertainty level rises from about +-2 degrees to +-5 degrees.

This is almost certainly true as well for the zone scoring that STATS and BIS do. The difference is their data is not made public so nobody can check it.

This is cool stuff. Other than batted ball speed, you should have all the ingredients here to do something similar to UZR. I hope John follows up on this, batter handedness makes a big difference, as does bases occupied for everyone but 3B.

Wang inducing lots of weak grounders that even Jeter can reach?

There is something to this, MGL has found that groundball pitchers have groundballs turned into outs at a higher than average rate. I've wondered how much of this is due to player switches, like playing Todd Walker when Pedro is on the mound and Pokey Reese when Derek Lowe pitches, or playing HOJO as a shortstop because Sid Fernandez only gives up 3 groundballs per year.

If the left fielder or third baseman makes the play I think the recorder inadvertantly places the ball a little closer to 3d base and if the center fielder or shortstop makes the play I think he records the location a little closer to second base.

Given that the dip in the distribution happens in the spot where neither the 3B nor the SS tends to field the ball, could it be that there are a portion of infield grounders that are not recorded as "infield grounders" simply by virtue of the ball having gone through to the outfield? IOW, the classification "infield grounder" might be used for balls fielded by the infield, plus balls judged to have been fieldable by the infield? I don't know how the data was segregated to just infield grounders, but it could be a classification error.

Aren't all grounders infield grounders? Do you think some grounders that make it though the infield are being recorded as line drives?

There is something to this, MGL has found that groundball pitchers have groundballs turned into outs at a higher than average rate. I've wondered how much of this is due to player switches, like playing Todd Walker when Pedro is on the mound and Pokey Reese when Derek Lowe pitches, or playing HOJO as a shortstop because Sid Fernandez only gives up 3 groundballs per year.

Hmmm, never thought about that. Have to check on that. Although it makes sense that ground ball pitchers yield easier ground balls than fly ball pitchers.

About the intro, to whomever wrote that, the author, Walsh, was not trying to "rate" the defenses, per se, only to look at fielding rates for the various positions and in the various sections of the field. Of course, when a third or first baseman misses a ball, it is more costly.

Maybe the reason there are so few balls hit up the middle, is that pitchers who yield lots of balls up the middle tend to have short life spans?

Could the Yankees up-the-middle performance be influenced by Wang (who led all pitchers in assists) and to a lesser extent Mussina and Rivera, who both seem to field a larger than normal number of ground balls.

Although it makes sense that ground ball pitchers yield easier ground balls than fly ball pitchers.

Since the mid-1980s, when Elias was attempting to compete against Bill James and proposed the idea of platooning by FB/GB tendencies, it's been well known (although not well documented, because Elias wouldn't release its proprietary data) that GB pitchers have a higher percentage of GB turned into outs and a lower percentage of FB turned into outs (vice versa for FB pitchers). I did several studies when Retrosheet first released complete seasons from the Baseball Workshop that contained ball type data for all BIP and confirmed that hypothesis.

The reasons why this happens aren't especially complicated to understand (although not necessarily easy to document). When a pitcher with a given ball type tendency yields a ball type of the opposite tendency, he's made a mistake - and mistakes are almost always hit harder.

-- MWE