A while back, Vladimir Kramnik drew attention to some extraordinary winning streaks by Hikaru Nakamura. Kramnik is a former world chess champion. Hikaru has long been among the top few chess players in the world, and has made a career as a streamer, playing hours and hours of blitz online.
Chess.com, which is the online platform Hikaru plays on, engaged Jeffrey Rosenthal to do an independent review of the data. Rosenthal published his results in the Harvard Data Science Review.
The math in this paper is over my head, but I get the general idea: create a model, use it to run Monte Carlo simulations, compare the results to what actually happened, and along the way address valid criticisms and corrections regarding the methodology or the data. Rosenthal's bottom line: Hikaru's streaks were not really implausible, given the massive number of games he has played, and given how much stronger he is than his average opponent.
I'm a little annoyed that Rosenthal didn't consider nuances that were pointed out by Kramnik, like the fact that players have different performance expectations under different time controls. He did the right thing by incorporating this feedback and re-running his numbers, but I wish he had thought to do the work in the first place. As it happens, he didn't find much difference in the outcomes.
Kramnik posted a lot of scathing comments on the paper, claiming to have data and methods that refute Rosenthal's work. I can't speak to whether Rosenthal did good science or bad science. All I know is, Kramnik has been claiming for a long time to have damning evidence, but has refused to share it, meanwhile making nasty, thinly veiled accusations against other players under the guise of "just asking questions". That's a pattern of behavior I have no stomach for, especially these days, and especially from a person of his stature. I wish he would either put up, by opening his own work to the same academic scrutiny as Rosenthal did, or shut up.