Not directly. They would have to bring an expert witness to generate expert testimony.
The thing with the legal angle is that time is severely limited. The legal strategy has to be super-tight, so, there may be limited use of those things unless the expert testimony is likely to greatly influence the judge.
It(Benfords) was used in federal court in 2015 in US v Channom and the judge upheld it’s use because it had been accepted in the particular software community as indicating fraud. It was not upheld in Fed Court in NY in 2020 because they provided no support that “legitamtely created financial data sets generally follow Benfords law”. So it seems like it would be up to the lawyers/parties to establish Benfords law was relevant to the elections. Not sure about Doc Shivas because I don’t know what it’s called.
Look at a logarithmic plot from near 0 to 9. The length between numbers is not equal: its larger from near 0 to 1, then from 1 to 2, etc. You can calculate the percentage for each segment and generate a hypothetical distribution.
People have observed that for certain processes, the distribution of the first digit, second digit, etc. of a count of some thing (e.g., precinct vote tallies) follows the hypothetical distribution described above. You can do statistical tests to determine if differencws from the hypothetical are significant or not.
Not directly. They would have to bring an expert witness to generate expert testimony.
The thing with the legal angle is that time is severely limited. The legal strategy has to be super-tight, so, there may be limited use of those things unless the expert testimony is likely to greatly influence the judge.
It's what took down enron and bbc said it can be used to spot electoral fraud in 2016
It(Benfords) was used in federal court in 2015 in US v Channom and the judge upheld it’s use because it had been accepted in the particular software community as indicating fraud. It was not upheld in Fed Court in NY in 2020 because they provided no support that “legitamtely created financial data sets generally follow Benfords law”. So it seems like it would be up to the lawyers/parties to establish Benfords law was relevant to the elections. Not sure about Doc Shivas because I don’t know what it’s called.
What's Benford's law, in a nutshell?
Look at a logarithmic plot from near 0 to 9. The length between numbers is not equal: its larger from near 0 to 1, then from 1 to 2, etc. You can calculate the percentage for each segment and generate a hypothetical distribution.
People have observed that for certain processes, the distribution of the first digit, second digit, etc. of a count of some thing (e.g., precinct vote tallies) follows the hypothetical distribution described above. You can do statistical tests to determine if differencws from the hypothetical are significant or not.
cool. so its testing the observed curve generated from voting versus a know curve. thx.