mfb
Mentor
- 37,391
- 14,223
I don't see any evidence backing up that claim. Any example in the history of particle physics? To have a realistic chance that two of them appear independently in the same way, history would have to be full of those things. I don't think I ever saw that.ohwilleke said:This profile of what tiny coding errors could look like corresponds pretty closely to the kind of anomaly we see with the 750 GeV bump if it is due to systemic error and is not real.
Most of the work goes into systematic uncertainties. Particle physicists are very good at analyzing systematic uncertainties. They do it all the time.ohwilleke said:Moreover, LHC investigators are very good at analyzing statistical issues in an experiment. They do it every single time. They have a protocol and a process for doing it. They have a lot of highly analogous computations to draw upon. So, the amount of statistical error is very likely to be correct.
Then "maybe a binning effect" should not be a guess in an unbinned analysis, for example.ohwilleke said:But, an ability to make an educated guess about the kind of error involved
Correct, but a large fraction (probably the majority) expected some time measurement issue. It was immediately obvious that timing is one of the three critical points, together with the length measurement and the profile fit. An issue with the profile fit was ruled out later by the shorter bunches. There is a huge difference between "your clock synchronization could have an issue somewhere" (which happens all the time, although not always with loose cables) and "let's claim a coding error could produce a peak in the diphoton spectrum in some unspecified way, without any example or description how".ohwilleke said:I'll bet that almost none of even the most savvy armchair physics analysts predicted that it was due to loose cable connections.