I'd be very curious about that analysis should it be publicized. Specifically: how many other people does that stylometric analysis match "with a high degree of probability"? And what measures were taken to rule out things like adversarial stylometry?

My point: stylometry alone doesn't strike me as particularly strong evidence - especially in this day and age wherein people influence each others' writing styles bidirectionally all the time.

>Stylometric analysis shows this to be the same person with a high degree of probability.

Sounds par for the course with all the kookery and bullshit that somehow passes for “expert” evidence in US courts.

Yeah they'll have him take a polygraph next.

> Stylometric analysis shows this to be the same person with a high degree of probability.

Sounds like this is a new technique unproven to the high standards really needed for legal evidence.

The "anything plausible sounding goes" approach is already a problem for other widely used types of "evidence".

Those need to be stamped out too, not have new unproven crap added, making the problem worse.

> An acquaintance of mine worked on this analysis for the plaintiffs.

On that note, does your acquaintance have published work that can be fact checked?

It sounds like they're an expert, so surely they must have public info about this technique, and its accuracy when they're utilising it.

>"high degree of probability"

what degree? what's the false positive rate? what's the false negative rate? what's the true negative rate? what other evidence have you got?

even better, why are you talking about this at all? isn't it best practice to shut up about legal cases?

Sounds just like the FBI's hair comparison analysis technique.

Regardless, that is what the motion to dismiss is based on.