Buy The Book
As Homer says of donuts, I say of Linear Weights: “Is there anything they can’t do?” Dave Allen gives us Field Blobs:
My comment:
Dave, fantastic. One request please: show those maps based on the handedness of the batter (or even batter/pitcher). Ideally, we should see the red blobs shifted over, with patches of green more clearly separating them.
Can you also put in the foul-lines in a bold black line?
What’s cool is that you can figure out the OF fence based strictly on the darkest blue color code. And from that, we can also see where all the ground rule doubles go.
There’s something odd there, isn’t it? If I use the HR locations (the blue), I can infer the fair/foul line based on the left-most and right-most spots.
The RHB chart has a ton of HR to pull-side (left field), but they also have a ton if hits there. There’s a ton of outs to RF (opposite field). It almost looks as if you’ve got the OF data mixed up, but the HR data is correct. Can you confirm?
Also, do the white spots mean there were no balls hit to that location at all?
Tango, what is odd about that?
I would expect that balls that are pulled are more likely hit harder and/or a line drive than opposite field balls. That expectation seems to match up with the graphic.
Split BAbip SLGbip
Pulled-RHB 0.387 0.591
Up Mdle-RHB 0.263 0.354
Opp Fld-RHB 0.273 0.374
Pulled-LHB 0.358 0.566
Up Mdle-LHB 0.278 0.375
Opp Fld-LHB 0.320 0.433
David - Thanks for the kind words about my articles. I think I mentioned in Part 1 that an additional adjustment needed to be made to the outfield distances to make them correct. I just finished today calculating the Outfield Park Factor Adjustment for each park and year. I can send them to you if you wish with an explanation of how to use them. Won’t change your graphics much, but if you intend to do any quantitative analysis they will improve the accuracy.
SirKodiak/#5, yes, I agree. Did you pull those numbers from Baseball Reference?
The work I did on the Gameday fielding locations, which is the same data set that Dave is using here, showed exactly what you say.
http://www.hardballtimes.com/main/article/confessions-of-a-dips-apostate/
I agree on the pull. But the opposite field red blob down the line is just too red. It just looks like the opposite field OF is covering alot more ground than the pull OF.
Tango, I will post a link to images with foul lines soon (for some reason I am having a hard time getting them where they should be).
Regarding the size of the red blob in the opposite field, I think it comes from the LD%. In Mike’s article that he linked the figure ‘line drive fraction of outfield balls by direction,’ shows it very well I think.
Peter, I would really appreciate the Outfield Park Factors. Can you send them to dnallen -at- baseballanalysts -dot- com. Thank you.
Tom, the opp-field OF is in essence covering more ground than the pull OF because he has more time to get to the softer-hit balls. The pull field sees many more line drives than the opp field.
If you look at the line drive fraction graph in my article, you can see that the line drive fraction is around 50% in the pull field and only 30% in the opp field.
Mike,
Yeah, I pulled them from B-R, but forgot to credit them. The numbers are for MLB in 2008. The SLGbip I calculated since they don’t have it. I can’t believe I missed your only article of the year so far on THT, if I hadn’t missed it, I would have pulled the numbers from there. Thanks for the link.
Peter, can you send me those numbers as well? Thanks. What are they exactly and how did you calculate them?
Looking at those left and right batter pics shouldn’t there be a slight shift to the left for all the outfielders to optimize for right handed batters, and the reverse for lefties, in a very general sense??
Hamzen, this is what I was thinking. I understand that pull hitters hit harder on the pull side. It seems that there is way too much falling in for hits on the pull side. I was expecting to see gaps in the alleys, not what I see here.
Shea may be a bad example. The Mets led MLB in LD% (according to FanGraphs) and do have a lot of left handed and switch hitters.
I’d be greatful for those figures as well, Peter. You can e-mail me at:
pontifexexmachina at hotmail.com
re #15 SirKodiak
Re all of mlb, I’d really like to see all the stadium splits, and by home team and away team to see how universal that situation is and how aware teams were of their stadium specifics if it wasn’t universal as a general rule.
re #14 Tangotiger
Exactly, I too was expecting the same. And looking further at the graphic for left handed batters, it also hints at pulling the right fielder toward infield as well, with the cf covering his back a bit, as well as a slight shift to the right. I’d love to see whether the numbers would confirm that if the changes were made, and by how much, because to my eyes it looks noticeable
Several people seem to be assuming that the outfielder is positioned roughly at the middle of the red blob. I do not believe that automatically follows from the construction of the graphs, for at least two reasons. One, all batted balls to the various sides of a given outfielder are not likely to have the same difficulty level. Two, all hits that fall on the various sides of a given outfielder will not necessarily have a similar distribution of run values. This is particularly the case for the pull field, where the outfielder sees many more hard hit balls toward the line than he does toward the gap
One other thing that’s hard to pick up from these graphs is exactly where the middle of the field and foul lines are on the graphs split by handedness.
Another issue with repositioning the outfielders based on these images is that these show the average run value by location, but not densities of where the balls are hit. So pulling outfielders in might cut down on high run value balls hit in front of them, but maybe there are not that many balls hit to that location relative to balls that they would now have a higher run value behind them.
I added foul lines and did it for Petco, thinking maybe there would be more discernible gaps in the huge outfield. You definitely do see some gaps, but they are still pretty non-distinct and do not have very high run values. It must be lazy flies that outfielders can track down off set the hits in the gaps.
I am getting some positive values outside the foul lines in the infield. I don’t know if this is caused by how I am averaging the data to make the image or from mistakes in the underlying Gameday data.
Big thanks again to Peter for publishing the conversion factors.
The Mets also had the lowest FB% in MLB last year according to FanGraphs. OF fielding might be a factor as well:
Name Pos Inn UZR UZR/150
Carlos Beltran CF 1407 7.0 7.0
Ryan Church RF 724 2.8 4.8
Endy Chavez RF 400 6.6 19.8
Fernando Tatis RF 292 0.2 2.2
Fernando Tatis LF 284 -0.7 -3.6
Daniel Murphy LF 249 1.2 7.5
Endy Chavez LF 197 6.8 38.2
Nicholas Evans LF 186 1.8 17.5
Angel Pagan LF 169 0.3 4.1
Marlon Anderson LF 165 1.8 12.6
Moises Alou LF 92 0.1 0.9
Trot Nixon LF 44 2.4 69.6
Endy Chavez CF 38 1.1 55.3
Also, is the data where the ball first landed or where the fielder picked it up (for non-HR)? The latter would explain the positive values outside the foul lines.
I think that Gameday reports where the ball is fielded.
The would also explain the lack of “gaps”. Line drives to the gap would most often be fielded near the wall, I expect.
Feb 12 05:18
Reader Mail of the Day: Why do we need X years of fielding data? And what about outliers?
Feb 12 04:55
Who is Jeremy Lin?
Feb 12 03:15
New PECOTA
Feb 12 02:42
Whitney Houston
Feb 12 02:23
Psst… wanna intern in Canada?
Feb 12 00:40
Clutch analogy
Feb 11 20:11
Fighting leads to goals?
Feb 11 19:55
Why do players get crappy caps?
Feb 11 19:12
Hero of the month: Brittney Baxter
Feb 11 17:59
MGL: Today on Clubhouse Confidential
Good idea Tango.
http://baseballanalysts.com/shea_rl.png