The interesting part if you are not interested in the business as usual part of particle physics is:

> 10 March 2020: B0 → K μ+ μ-: more data confirm old puzzle.*

> Compared with the previous LHCb results, the overall tension with the Standard Model (SM) is observed to mildly increase.

> However, the global fit to several angular observables shows that the overall tension with the SM increases from 3.0 to 3.3σ.

[Note that the "increase" in the second quote and the "increase" in the third quote are in different comparisons.]

The idea is that the transformation in the experiment is actually B0 → <something> → K* μ+ μ- , where there are many known versions of <something>. But they got a weird result, so there "must" be some unknown versions of <something>.

This is 3σ, that means that a random fluke has less than .3% of probability of producing a similar looking weird data. The problem is that they (and other groups) are running hundred of experiments, so there is a chance that this is just good/bad luck instead of a unknown <something>.

A 3σ is unusual enough to remember, and to try to explain it with new theories, but don't get to attached to a 3σ resonance because they may fade away.