Here's a 2019 article with the Zebra CEO saying the accuracy is only good down to 6 inches. That's not enough.
Also, if the ball is covered up or blocked by bodies it isn't going to register, especially not as accurately as would be needed to spot the ball.
Thanks, I had not looked up specifics of the system. From using similar systems (over a much smaller space), I assumed the accuracy was better. You know what happens when you assume.
The ball being blocked by bodies should not matter (I am making assumptions again, this is not addressed in the article you linked), the radio signals should be able to pass through any intervening body.
If they really are +/- 6", how do you think that compares to manual spots? Do you think most of those are within +/- 6" of actual? Not sure, it would be interesting to see how well the ref-spot and the zebra data correlate.
I am always amazed at how well refs spot the ball. Hundreds of times I have immediately derided a spot in real-time, only to see it in slow-mo and see that they got it pretty much exactly. But those head-scratchers do happen where they are off by a yard or better, and it would be great if there was an easy way to eliminate those.