off the top of my head isnt this something to the diffuse nature of wave packets? and picking up the leading edge of the packet at the receiving end when you started the timer at the outgoing end on the peak of the neutrino wave packet? remember something like this happening at the tevatron a while ago. cox'll fix it though.
No. I've skimmed their paper and they've got that one covered. They seem to have checked all the obvious stuff by multiple different methods.
They generate a pulse of protons that's mostly a 10us long rectangle, but with some structure like a 2us saw-tooth and a smaller 5ns sine wave added to it by the accelerators that provide the protons. When the experiment is running, every 6s they fire two of these pulses (spaced 50us apart) into the neutrino beam generator which converts approximately all the protons into muon neutrinos. These head off towards Gran Sasso in a fairly tightly focused beam that spreads out to about 2km diameter on that 730km journey. When they get there usually nothing happens - the neutrinos go straight through or past the detectors without interacting. But on about 16000 occasions over the 3 years they've been running the experiment, a neutrino has interacted either with the detector or with the rock in front of it and produced a characteristic muon that has been picked up by the detector.
By putting together the time delays of all those detections and comparing the resulting frequency distribution with the shape of the proton pulses that generated those neutrinos, they work out a time-of-flight for the neutrinos.
If the pulses were purely rectangular, their result would depend very sensitively on the few interactions from neutrinos right at the beginning or end of the pulse. They do suffer from this a bit, but thanks to the saw-tooth structure of the pulse, the interactions from neutrinos in the middle still contribute something to the evidence. It's still a pretty small effect - a 60ns shift of a 10000ns pulse, and it won't take much of an unaccounted timing error to explain it away.
@moth
They generate a pulse of protons that's mostly a 10us long rectangle, but with some structure like a 2us saw-tooth and a smaller 5ns sine wave added to it by the accelerators that provide the protons. When the experiment is running, every 6s they fire two of these pulses (spaced 50us apart) into the neutrino beam generator which converts approximately all the protons into muon neutrinos. These head off towards Gran Sasso in a fairly tightly focused beam that spreads out to about 2km diameter on that 730km journey. When they get there usually nothing happens - the neutrinos go straight through or past the detectors without interacting. But on about 16000 occasions over the 3 years they've been running the experiment, a neutrino has interacted either with the detector or with the rock in front of it and produced a characteristic muon that has been picked up by the detector.
By putting together the time delays of all those detections and comparing the resulting frequency distribution with the shape of the proton pulses that generated those neutrinos, they work out a time-of-flight for the neutrinos.
If the pulses were purely rectangular, their result would depend very sensitively on the few interactions from neutrinos right at the beginning or end of the pulse. They do suffer from this a bit, but thanks to the saw-tooth structure of the pulse, the interactions from neutrinos in the middle still contribute something to the evidence. It's still a pretty small effect - a 60ns shift of a 10000ns pulse, and it won't take much of an unaccounted timing error to explain it away.