I've linked a 5 minute interview by Lou Dobbs of Russ Ramsland, from the Allied Security Operations Group. Ramsland has been digging into the issue of election fraud--computerized fraud--for several years. He began his research--as I understand it--following the 2018 midterm elections in Texas and Kentucky and has since expanded to other states.
In the video below Ramsland discusses the situation in Michigan. He points out that to do a correct analysis you need to know exactly what version of the machines were used in any given precinct, because different version of the machines have different characteristics and capabilities. He obtained the serial numbers for the machines that were used in Michigan, then focused on four counties. Those counties had voter turnout that appeared to be prima facie impossible and also saw very large spikes in votes--mostly for Biden. According to Ramsland these spikes are "physically impossible", because the machines in use were not capable of processing the given number of votes in the time period indicated.
If true, this should obviously prove to be a very significant finding. Ramsland estimates that the number of votes involved is between 225K and 400K. That's important, because courts will not generally look at such anomalies unless the numbers involved could lead to reversing the outcome of the election. These numbers, even at the low end, are well past that threshhold.
Ramsland also stated that his team has been concentrating on Michigan but is now "turning our sights" on Georgia.
What I find so fascinating about this interview is that it shows how forensic data analysis can build a pretty convincing case for fraud, one that goes beyond statistical analysis alone. I'm not knocking statistical analysis. However, to be able to tell a court--We targeted what appeared to be statistical anomalies in four counties in Michigan, and subsequent analysis confirmed that those anomalies were actually physically impossible--builds a strong basis for arguing that similar statistical anomalies in other states should be presumed to, at a minimum, require closer examination. If not a presumption of fraud.