Jump to content
Umpire-Empire locks topics which have not been active in the last year. The thread you are viewing hasn't been active in 4515 days so you will not be able to post. We do recommend you starting a new topic to find out what's new in the world of umpiring.

Recommended Posts

Posted

Worthless article. But thanks for sharing. 5 minutes of my life I will never get back!  :wave:

Posted

I just want to hear these pro-machine guys the first time the machine calls a breaking pitch a strike that hits the hollow of the front knee on the outside corner and breaks into the dirt 5" off the plate in the dirt in front of F2

  • Like 3
Posted

The methodology of this study is flawed, failing amongst other things to take into account:

  • Pitch f/x margin of error (no less than 1/2")
  • Baseball radius considerations (e.g., a ball whose center is located one iota of an inch off the plate will be judged a ball by pitch f/x, even though the physical existence of the ball's radius means the ball in some fraction actually passed over the plate, making it a technical strike)
  • Three-Dimensional strike zone (Pitch f/x plots output a 2D zone, captured as the ball crosses the plane located at the front edge of home plate)
  • Batter height variability (Pitch f/x often fails to account for height variability mid-AB and is dependent on a third party human operator's calibration)

But detecting these errors would require a grasp of not just the PFX technology, but the rule book as well, specifically 2.00 (Strike) and 2.00 (Strike Zone).

 

We've already seen Pitch f/x ball a pitch thrown right down broadway and strike a pitch that bounced in front of home plate. Also, this (and YES, this and similar errors are PART OF THE DATA that the linked article is using for its "analysis"):

jay_jon_g_mp_576.jpg

  • Like 1
Posted

The methodology of this study is flawed, failing amongst other things to take into account:

  • Pitch f/x margin of error (no less than 1/2")
  • Baseball radius considerations (e.g., a ball whose center is located one iota of an inch off the plate will be judged a ball by pitch f/x, even though the physical existence of the ball's radius means the ball in some fraction actually passed over the plate, making it a technical strike)
  • Three-Dimensional strike zone (Pitch f/x plots output a 2D zone, captured as the ball crosses the plane located at the front edge of home plate)
  • Batter height variability (Pitch f/x often fails to account for height variability mid-AB and is dependent on a third party human operator's calibration)

But detecting these errors would require a grasp of not just the PFX technology, but the rule book as well, specifically 2.00 (Strike) and 2.00 (Strike Zone).

 

We've already seen Pitch f/x ball a pitch thrown right down broadway and strike a pitch that bounced in front of home plate. Also, this (and YES, this and similar errors are PART OF THE DATA that the linked article is using for its "analysis"):

jay_jon_g_mp_576.jpg

I thought pitch/fx definitions state that the dots are the exact centerpoint of the baseball and therefor a ball that hits the white barely outside the black square of their chart is also a strike and is counted as such until it gets to a certain point outside of the square in which case the whole ball would have missed any part of the square (plate as well as high/low) and therefor have totally missed the zone.

Posted

I thought pitch/fx definitions state that the dots are the exact centerpoint of the baseball and therefor a ball that hits the white barely outside the black square of their chart is also a strike and is counted as such until it gets to a certain point outside of the square in which case the whole ball would have missed any part of the square (plate as well as high/low) and therefor have totally missed the zone.

That's what ZE and (for me) UEFL rules are for. PFX is very black and white since it is raw data. If the center point is out, the pitch is out.

×
×
  • Create New...