For experts: Uncertainties

Measurements are always subject to statistical fluctuations, i.e. random changes in the measured value. These become smaller the more data one has. In addition, there are systematic uncertainties that can bias a result. In experiments, these are determined with control measurements.
In physics it has established best practice to set the threshold of a discovery very high in order to exclude random fluctuations as far as possible. The convention is to speak of a "hint" if the probability for a statistical fluctuation is less than 0.15 percent and of a "discovery" if it is less than 0.00003 percent (or 1 in 3.3 million).

The different measurements of LHCb correspond to a probability of 0.1 - 20 % for a random statistical fluctuation, from the theoretical prediction that muons and electrons behave identically and hence corresponds to a hint.