Skip to main content

Reply to "Mlb testing rule changes in independent Atlantic league"

Hammer823 posted:

Radar technology for calling balls and strikes still has an average margin of error of about .5 inch per reading.  This is for horizontal location, vertical location, and vertical movement (each individually) based on the Trackman and Pitchfx data from 2015-2017.  This error rate fluctuates, and can be much higher at times (well over 1 inch, per reading) just like with a human. When taking the readings as a whole, the combination of error shows that a robo ump is not going to be perfect, and can make erroneous calls, just like a human ump.

MLB only uses TrackMan (Statcast) now, so the PITCHf/x data isn't really relevant. Having said that, I don't know which system is more accurate at calling balls/strikes. TrackMan is also tracking the ball all over the park, not just over the plate. The sensors have to be checked for calibration regularly. I hope MLB is looking at a more specialized technology for calling balls/strikes. I have to believe that a dedicated system could basically do the job with 100% accuracy. This is similar to self-driving cars. Overall, the current technology is probably better than humans, but it won't even be close in the near-future.

×
×
×
×