I have a mathematics background. I understand the potential for dataset sizes to set frequency of leading digits. I just haven't seen anyone offer evidence that that actually happened.
Lots of people saying "well if every district was exactly 1000 people then that would give a different benford curve" with nobody saying "Here is the list of district sizes"
Do you not think that based on geography, population density, and logistics that election planning commissions try to set up voter precincts to service a certain range of voters?
do you think that in cities some voter precincts are made to service 10 voters and others 1 million voters?
or perhaps within a region an election planning commission would plan and design them to be certain sizes say to cover 1000-10,000 people in the city, 500-5000 people outside of cities, and 200-2000 people in rural areas.
Obviously in cities you can have precincts that serve more people, and in rural areas drive distance and population density.
PLANNING, demographics, geography, logistics...all these things add structure to the dataset, whether a first digit vote counts do or don't follow Benford's law can therefore be due to a range of things besides fraud or no fraud.
I downloaded the milwaukee data, performed a histogram, and posted the results. This is a structured data set. vote precincts are planned sizes. This is obvious. citing benford's law for vote fraud makes us look like dumb shits.
It undercuts all the other arguments, especially when you post it at the top.
If you make a summary and an opponent can pull out a main point that is ridiculous and roast you over it then you convince NO ONE.
I have a mathematics background. I understand the potential for dataset sizes to set frequency of leading digits. I just haven't seen anyone offer evidence that that actually happened.
Lots of people saying "well if every district was exactly 1000 people then that would give a different benford curve" with nobody saying "Here is the list of district sizes"
Do you not think that based on geography, population density, and logistics that election planning commissions try to set up voter precincts to service a certain range of voters?
do you think that in cities some voter precincts are made to service 10 voters and others 1 million voters?
or perhaps within a region an election planning commission would plan and design them to be certain sizes say to cover 1000-10,000 people in the city, 500-5000 people outside of cities, and 200-2000 people in rural areas.
Obviously in cities you can have precincts that serve more people, and in rural areas drive distance and population density.
PLANNING, demographics, geography, logistics...all these things add structure to the dataset, whether a first digit vote counts do or don't follow Benford's law can therefore be due to a range of things besides fraud or no fraud.
"maybe this, maybe that, probably this" is all I've ever seen.
Hard numbers. Put up or shut up.
https://communities.win/c/TheDonald/p/12i48fTL50/
I downloaded the milwaukee data, performed a histogram, and posted the results. This is a structured data set. vote precincts are planned sizes. This is obvious. citing benford's law for vote fraud makes us look like dumb shits.
It undercuts all the other arguments, especially when you post it at the top.
If you make a summary and an opponent can pull out a main point that is ridiculous and roast you over it then you convince NO ONE.