This is an interesting problem that somebody here might have a clue, it involves interference with a 635nm laser range finder by a white Cree XP-G2 emitter in an art exhibit.
Here are some sentences about the problem, additional details are described here, range interruptus
I would guess that it has something to do with the spectrum of the white LED including 635nm. A green LED won’t disrupt a sensor calibrated to 635nm but even a little light in that frequency from the white LED might disrupt it, especially as the ball gets higher and thus closer to the LED, increasing intensity.
Just a thought…