A Data Scientist's Deep-Dive That Could Change History

Article just assumes the anomalies are due to fraud and then says this anomaly is fraud because earlier we said it might be cAused by fraud.

Someone pushing their Agenda not a proper analysis

I'm really not following the logic here. Are they pointing at the jump in D vote share, and then pointing at the influx of Milwaukees ballots? Milwaukee is excessively blue (86% to Biden). Why wouldn't the vote share change drastically as that county came in? This is basic math. How does this point to fraud?

Also hilariously, they're basing their whole argument off of this apparent "trend" that they haven't supported of mail-in ballot share shifting from D to R... But then they provide an example of Georgia, where... that isn't at all the case... and then gloss over it. Laughably bad analysis.

Took 20 minutes to confirm the "trend" and immediately none of the following states follow the "trend": Alaska, Florida, Hawaii, Illinois, Indiana, Maryland --- aaaand I've stopped looking because this is silly.

Full disclosure, this was just a quick visual inspection of the data presented, plotting x='vote_share_dem' against 'timestamp'.