Umpire Called Balls and Strikes Heat Map
Umpire

Compare to another umpire. *Optional*

Scale
StrikesBalls
Compare to league averages
Regression Amount

Select a valid start and stop dates with all fields filled in or no data will be generated.
Start Date
End Date
Pitch Type (Leave blank for all)
CH, Change-up
CU, Curveball
EP, Ephuus
FA, Fastball
FC, Cut Fastball
FF, 4-Seam Fastball
FS, Split-finger Fastball
FT, 2-Seam Fastball
KC, Knuckle Curve
KN, Knuckleball
SC, Screwball
SI, Sinker
SL, Slider
UN, Unknown


Definitions:

Compare to Another Umpire - This function looks at the two umpires over the same time frame and set of conditions. Then it subtracts the percentage of called strikes divided by the total of the called balls and strikes by the second umpire from the first umpire. For example, if the Umpire called a pitch in a zone a strike 40% of the time and the Compare to Another Umpire called pitches in the same zone a strike 50% of the time, the output for that zone would be -10% (40%-50%) or 0.10.

Scale - Scale may need to be adjusted for a cleaner output. The data is put into subsets and there may not be enough data to have smooth and/or accurate output. The scales are based on the percentage of called strikes vs called balls and strikes. The 2.4 inch scale is 15x15 (225 total) outputted data zones and the 6 inch scale is 6 by 6 (36 total) outputted data zones.

Compare to League Average - This function works the same as Compare to Another Umpire, but instead it compares the selected umpire to the league average. Creating the league average is extremely time consuming (hours of processing), so as of right now there is a league average for both right and left handed hitters for both the 2.4 inch and 6 inch scales.

Regression Amount - This number is the amount of league average values that are added to the players total. At least some value (10 to 30) should be added to each player to make sure that each outputted data point has some values associated with that zone and that a few extreme values do not incorrectly dominate the outputted image.