Measurements of top-quark pair spin correlations in the \(e\mu\) channel at \(\sqrt s = 13\TeV\) using \(pp\) collisions in the ATLAS detectorYou should also see a 27-page-long presentation by Reinhild Peters.
To make the story short, the measured correlation between the top quark spins – in events with one \(e^\pm\) and one oppositely charged \(\mu^\mp\) at the end and a top quark pair in the middle – exceeds the theoretical prediction by 3.7 standard deviations if you pretend that the theoretical prediction is exact, or 3.2 sigma if you choose some sensible nonzero error margin for the theoretical prediction.
The chance that a deviation of this size appears by chance is comparable to 1 in 1,000.
It may be a fluke – after all, ATLAS and CMS have measured a thousand of similar numbers so one of them may deviate by a large deviation that seems like "one in one thousand cases". As always, there's some possibility that the top quarks' spin correlation is enhanced by some physics beyond the Standard Model. It could be many things, I have no idea what should be the default explanation. If the top quarks sometimes came from some new spinless or spin-one intermediate particles, you could move the spin correlation up or down, respectively.
The LHC (Les Horribles Cernettes) girls have sung a song about the spins of quarks. You are invited to listen to the song, measure the correlations yourself, and determine whether the deviation from the Standard Model is exciting enough.