Visual Baseball: Introducing the Paintomatic
Here’s an attempt at visualizing a pitcher’s arsenal (what they throw, how often they throw it, and how well they throw it.) How well they “paint” so to speak. For each pitcher you’ll see a visual expression of the pitches they throw, with size representing how often they throw it and color indicating the quality of the pitch (measured by Runs Above Average / 100 pitches thrown). I’ve included Paintomatics for a variety of pitchers – ranging from Cy Young contenders like Zack Grienke and Tim Lincecum to duds like Sidney Ponson and Rich Hill. And a few odd-balls like Tim Wakefield. Let me know what you think.
Kevin Dame is a writer and visual designer who brings sports information to life in new and meaningful ways. Visit his website and follow him on Twitter @kevintdame.
Yeah, Wakefield was one of the more interesting ones because I was surprised by (a) how average his knuckleball was (I would have expected it to be tougher) and (b) how effective his fastball was. It goes to show you pitch value can’t be looked at in a vacuum (ie. his fastball is that good because of the kuckleball).
Checking Tim Wakefield’s pitch value data on Fangraphs.com (so indirectly from Baseball Info Solutions), his fastball has been filthy good for three years in a row but before that it was quite bad for three consecutive years. There are a couple of Fangraphs posts during 2009 about Wakefield’s fastball remarking on its effective too.
It doesn’t seem like random small sample noise. It sure seems as if Wakefield did something to improve his fastball entering 2007 and it’s now been an excellent pitch (when used sparingly one assumes) for the past three years.
Circular areas are notoriously difficult for people to perceive. Though this is “interesting”, a more effective visualization may be a simple stacked bar, histogram, or (though I’m usually loathe to suggest it…) pie chart. You can still use the color scheme, but the pitch distribution would be easier to see that way.
This awesome, more good work. I would love to see these for Rich Harden and Roy Halladay.
I love it! Where’s the Paintomatic for Felix?!
Would be great to see these generated for the entire league.
Sort of a context-dependent filthy, isn’t it? If you took Wakefield’s fastball to anyone else’s game it would be destroyed. That it’s thrown off his knuckleball is really the only way it survives in pro ball.
Though, really, I guess this point can really be applied to any guy who depends on offspeed stuff.
These look really cool, but I’m wondering about a couple things…
-What did you use to determine “quality”?
-Is there any reasoning behind the placement of each circle?
Dan, thanks for reminding me – I just updated the post to explain that pitch quality is measured by Runs Above Average / 100 pitches thrown. As for the placement of each circle, no reasoning. I wonder if that’s something that could be used to communicate something else about the pitcher.
Kevin, I like what you did here. Especially Chris Carpenter for some reason. I think it would be sweet if you found a way to incorporate velocity of the pitch types, too. You could either do something with placement or shapes.
This is great. I’m confused about Derek Lowe. Is his sinker considered a fastball?
These look really great, and I think there’s a lot of value to something that at a glance gives you a feel for the type of pitcher you’re looking at. On the placement of the circles, maybe there is some way to use that to aid in the comparison of pitchers. If there was a consistent standard, certain “orbits” of circles would then indicate similar pitchers. The coloration would then take on more meaning as you could see two pitchers whose repertoires would indicate shared traits, but whose stuff then separates them.
The coloring scheme is somewhat difficult to discern. I know you’re going for a hot-cold style, but maybe alter the colors a bit more?
Kevin, great stuff. My question is sort of inspired by the Wakefield data and discussion. Wakefield’s fb is “filthier” because of its relation to his knuckler. But in the case of someone like Lincecum (or other elite pitchers) do you think this indicates they should be throwing their filthier pitches more (like Lincecum’s change)? Or is Lincecum’s change filthy because he sets it up by throwing that many fastballs?
This is a long-winded way of asking if you think that there would be value in a pitcher altering the proportion of his pitch makeup based on these charts?
Needs more King Felix
This visualization is superb. Here are two suggestions for tinkering:
1.Velocity on the vertical scale.
You might set the Fastball velocity on a vertical scale, higher is faster and lower is slower. (You could introduce a dotted line from the center of the Fastball circle that goes upward, or downward, to the league average point. If it’s not too cluttered, you could put the average Fastball velocity in whole numbers at the Fastball center.)
Then you could array the Curve, Slider, CHangeup, and so on with the velocity set on the same vertical scale as the Fastball. The arrangement would give a sense of separation.
2. Direction on the horizontal scale.
Looking from the pitcher’s viewpoint, see the Fastball on the Right for a righthander and on the Left for a southpaw. Array the Curve and Slider to the left for a righthander and on the right for a southpaw—as would be the normal location for well thrown breaking balls. The CHangeup would be on the opposite side of the Fastball. The Knuckler might be below (or animated to jump around?) the Fastball.
Great thinking here.
First, I want to make clear that I think that this is fantastic.
If this had been my brilliant idea, here’s how I’d change it:
[1] Use pitch movement to roughly generate the relative locations of the pitches. I’m not talking about anything to scale, but seeing a rough picture of the movement that the batter sees would lead to a more intuitive understanding of the pitchers’ repertoires.
[2] Add text with average MPH for each pitch. If you have an idea how to do this visually, I’d be on board with that too, but I really like the current feel of your images so I am not eager to change them much.
Great stuff.
obviously wakefields knuckler is average because its all he throws and hes an average pitcher overall. no real suggestions but how about one for nick swisher:)
My favorite part of this is how it reveals just how nasty Tim Wakefield’s fastball is …. usually at about 73 miles per hour. xD
This is awesome. I just was having a debate with someone about whether Lincecum threw a lot of sliders. And this illustrates it perfectly!!
Just a note on Wakefield’s “nasty” fastball…
Joe Castiglione, the excellent long-time radio announcer for Red Sox games here in New England, calls it a “straight ball” when he’s doing play-by-play for Wakefield’s games. Because, you know, it’s not really accurate to call it “fast.”
I like this too. In order to make the placement of the objects useful, I would suggest a quadrant approach. The two axis could be x) how freqently the pitch is thrown and y) how often the pitch is thrown when the pitcher is ahead in the count/behind in the count.
Hey Kevin, this is really cool. And I love the Chris Carpenter one
.
One thing I would suggest is to make the location of the circles meaningful. To do this I would employ the relatively simple method of plotting the mode pitch location inside of a strike zone. So say that Carpenter threw his fastball on the low outer half most often, that’s where you would plot the circle.
Let me know if you are interested in this and I can send you a spreadsheet with each pitcher’s pitches, their rv100 and the mode location.
Nick, that sounds like an interesting direction to explore. Thanks! Please send your spreadsheet to
.
I saw this written in a different comment, but I think this would be really interesting if you did it with the entire league and made it easy to find. Mariano Rivera would be a really interesting graph too.
I’ll get that to you in a couple of days when I get on my computer with my data.
Sweet! Great idea…would love to see the lefty/righty batter split if you have the data.
Love it, love it, love it.
With all the commentary about Lester’s 5 quality pitches, can we see one of these for him?
Speaking of Mo, it would be nice to correlate the placement of the pitches with the average location of the pitches to the black. We all know Mo’s cutter eats up lefties better than righties. But, I’m not aware of other pitchers’ strengths or habits. Maybe a split page to show pitches locations versus left handed and right handed batters. A Hit-o-matic for batters would be just as awesome.