Question about Spray Charts
by NoNameOnCard on Dec 16, 2009 11:51 PM EST
Email
Print
Inspired by iamawesomer's original post about Fly Ball Charts (http://www.beyondtheboxscore.com/2009/2/2/744622/from-my-latest-post-on-rus), I've added some spray charts to TexasLeaguers.com PITCHf/x Tool, but I have some questions for those interested in providing some feedback.
Jump ahead to help me out.
I've got a very simple set up right now. The spray charts plot all batted balls for the selected criteria, but right now I'm categorizing them simply as "Hit" or "Out".
My question to you all: If you could have anything you wanted in a spray chart like this, what would you want?
I'll take any feedback you guys have, but if you could prioritize your wants, that would be helpful.
22 comments
|
0 recs |
Do you like this story?
Comments
BTW...
I haven’t brought in the ballpark images yet – which was the inspiring part of iamawesomer’s chart. I plan to do that, but they’ll probably trickle in rather than show up all at once.
Just added Rangers Ballpark.
It is the default ballpark. I will set up the others as I have time.
by NoNameOnCard on Dec 17, 2009 1:33 AM EST up reply actions
I would add borders around the field
It makes it easier to visualize.
I would also split it up into LD/FB/PU/GB – I don’t think people would care so much about the actual outcomes. Or you could make two different charts. One with the batted ball types, and one by 1b/2b/3b/Hr/Out.
At any rate, I think this is totally awesome.
this article seems relevant to your interests
by iwasawesomer :( on Dec 18, 2009 8:03 PM EST reply actions
His homeplate y coordinates...
Don’t seem accurate in any conceivable way. The y coordinate I use is 211. The only adjustment I made was changing it from 210 so that my “base box” wasn’t jagged on one or more edges.
I don’t even see how it’s possible for the data to suggest that homeplate was sometimes nearly 15 feet closer to the mound than the data says? My brain is rejecting his findings.
by NoNameOnCard on Dec 18, 2009 11:30 PM EST up reply actions
Wait...
It’s not even 15 feet. It’s 15 times his distance multiplier…
That means that for Kansas City’s 2008 y coordinate… 210 – 195.1 = 14.9… 14.9 * 2.86 = 42.6 ft. Homeplate was found to be 42.6 feet closer to the mound than the raw data says?
by NoNameOnCard on Dec 18, 2009 11:34 PM EST up reply actions
and our pitchers still get lit up?
Crap!
Glad I came, just wish I hadn't stayed so long.
by Warden11 on Dec 19, 2009 1:46 AM EST up reply actions 1 recs
I'm not measuring it.
It’s a given coordinate (125,210) on the 250×250 grid.
by NoNameOnCard on Dec 19, 2009 9:32 PM EST up reply actions
You said "what the data says."
So I thought you were validating it against the data somehow.
The 125,210 location holds true for most parks in most years, and should be correct for all parks post-2007. But for instance, in 2005-2007, I get a home plate location of 206 for Arlington. (That’s not estimated – I took the 250×250 graphical representation from Gameday into Photoshop and measured it.) Observe the difference:
There are also some subtle changes in fence differences you should notice.
Yeah.
I took Adler’s “word” as given. I’ve been talking to Peter Jensen today about this very topic. If I can get all 30 ballpark images, I’m going to play around with the coordinates of each home plate.
by NoNameOnCard on Dec 19, 2009 10:10 PM EST up reply actions
There's more to it than that.
Go ahead and measure a straight line from the front of home plate to the centerfield wall, which is a distance of 400 feet. That’s 178 pixels, or .445 pixels per foot. Go ahead and then measure out to the left field corner, which is 332 feet. You get 143 pixels, or .431 pixels for foot.
In other words, distances out to the corners are "shorter’ than distances to straightaway center, as measured by pixels.
(And you don’t need all 30 ballpark images – you need at least 60, one set for 2005-2007 and one for 2008-2009. They changed all the graphics before the ‘08 season.) That’s why Jensen did what he did.
Now, I tend to think that some of what Peter is measuring is “noise” and not a real difference in home plate location, but unforunately one can’t assume a constant home plate location for all parks.
Do you happen to know where I can get these grid images?
Also, I’m struggling to understand why my straight-up mappings look correct despite these problems that have been outlined.
by NoNameOnCard on Dec 19, 2009 10:36 PM EST up reply actions
Nevermind.
It took me a while, but I figured out a way to pull it all in. I’m going to continue talking with Peter about the rest of it.
by NoNameOnCard on Dec 19, 2009 11:42 PM EST up reply actions
Hi Trip
Great work, one possible suggestion – though I am not sure if this is doable:
On the spray charts, it would be neat if the user can change the park on which the data points are plotted on. It might help a lot to see the effects of home parks on a hitter, particularly hitters with quirky home parks like Fenway. For example, the Rangers are trying to trade for Mike Lowell right now, and since he is a pull hitter who is believed to benefit from the Green Monster, one might wonder if he will lose those benefits if he comes to Texas. Therefore, one would like to plot data points for Lowell’s batted balls on the TBiA background – instead of Fenway ( I realize that you can get this info right now since the Fenway is not up in the database yet, but once it’s up, I assume that I won’t be able to see Lowell’s spray chart superimposed on TBiA anymore)
I know spray charts are not directly transferable from park to park since weather effects, but it would still be very helpful especially when we are speculating on FA acquisitions or trades.
I might eventually include that as a separate piece of the site.
It gets pretty complicated to handle all of that on the profile pages as they are set up now.
I think it might confuse some people to map a player’s entire spray chart onto a stadium where he only put 5 or 6 balls in play. The difference being “what would this look like in Houston” vs “what did his spray chart in Houston look like”.
I’ve been thinking about it, but I’m not sure when/where/if it will get worked in.
by NoNameOnCard on Dec 19, 2009 6:30 PM EST up reply actions
A few more data pieces that would be nice:
If you could break down the whiffs and foul ball data to in and out of zone %
Follow Me on Twitter @FreeZorilla
spray statistics
It would be helpful to see the number of balls hit on the infield to the right, and left side, as well as balls hit to the outfield on the right and left side. Using this data, we could statistically analyze that data, to help more accurately predict a batters future BABIP.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recent FanPosts
Follow us on Twitter!
BTB: @BtBScore
Justin: @justinbopp
Satchel: @SatchelPrice
J-Doug: @RationalPastime
Daniel: @CamdenCrazies
Peter: @CapitolAvenue
Josh: @garik16
Julian: @splashinpumpkin
Matt: @devil_fingers
Bill: @BillPetti
JD: @JDSussman
Lucas: @DBITLefty
Jeff: @jeffwzimmerman
Lewis: @LewsOnFirst
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
