Source: John Smith

The November 3, 2020, Presidential election data from California should set off alarm bells.

Election data generated by the natural voting process will closely match Benford’s Law predictions. (Benford’s Law, or the rule of leading digits, is used as an indicator of fraud in numerical data.) Of course, there can be exceptions, but Benford’s Law points to areas to investigate.

Benford’s Law predicts that for real, natural data, the frequency of leading digits (1 through 9) will occur at or near the following percentages of the transactions or tallies. Deviations from these predicted percentages should be investigated.                                                                

Here are the official presidential totals for California, as released by Dr. Shirley N. Weber, Secretary of State (also released by Alex Padilla, Secretary of State through January 28, 2021):

Precinct Tallies:15,946Percent
Vote count, Biden11,110,63964.9%
Vote Count, Trump6,006,51835.1%
Total17,117,157 

The difference in votes tallied is 5,104,121 in favor of Biden. The average precinct size is 1073 votes.

Biden had a nearly 2-1 official victory margin over Trump across the state. Here are detailed statistics for the statewide election, again, from Dr. Weber.

VBM is Votes by Mail.

Now, let’s apply Benford’s Law to identify possible fraud. Here is the statewide Benford chart, displaying the frequency of precinct tallies’ leading digits for Biden and Trump, compared to Benford targets. The official data have separate tallies for votes cast In-Person and by Vote-by-Mail for each precinct. Given this, there are roughly twice the tallies than the number of active precincts in the state.

This chart looks almost normal. The Benford targets (i.e., the expected leading digit percentages) are shown by gray vertical bars. Trump’s precinct leading digit percentages are shown by the orange line; Biden’s are the blue line. If we look closely, however, the Biden line is low in leading digits for 2, 3 and 4, and high in leading digits for 1, 6, 7, 8, and 9.

The Chi-Squared test is a statistical tool that measures the correlation of two curves (two sets of sequential data) and provides a numerical value for the correlation, with zero percent indicating little or no correlation, and 100% indicating high correlation. The Chi-Squared test is calculated and provided to show the candidates’ precinct tally data’s closeness to the Benford target curve.

Now let’s isolate the data for sources of fraud. Here is the equivalent state-wide chart for In-Person votes only.

These In-Person precinct tallies’ leading digits for both Trump and Biden correlate very closely to the Benford targets. The average precinct in-person vote counts are 103 Biden, and 97 Trump. The 15,226 precinct tallies come in all sizes, and the tallies are distributed across the leading digits as Benford predicts. These In-Person votes represent 17.8% of all official votes tallied. There is little or no fraud indicated by Benford’s Law here.

Now, here is the Vote-by-Mail chart:

The trend we saw on the first statewide Benford chart is more pronounced here, with Biden low in leading digits 2, 3, and 4, and high in 5, 6, 7, 8, and 9. The average official vote-by-mail count per precinct here is 600 Biden, 284 Trump. Notwithstanding these averages, the Biden precinct tallies are low in the 200s, 300s, and 400s, and high for the 500s, 600s, 700s, 800s, and 900s. The Trump tallies look high in the 200s through 500s and low in the 700s through 1,000s. Across 15,787 statewide precincts (a large sample size), this information, and the resulting “X-Pattern” of the candidates’ lines above, indicates likely manipulation of naturally generated precinct tallies.

We are faced with the following statements and questions:

  1. Based on Benford’s Law, Vote-by-Mail data are suspect.
  2. With suspect data, it can be difficult to draw conclusions. (Averages and standard deviations for suspect data are easy to calculate but difficult to assess. There is also the possibility of circular arguments.)
  3. What points us to believe data are suspect?
  4. Given 3, above, how suspect are the data?

As we saw little or no likely fraud in the In-Person chart, let’s look at the candidates’ precinct VBM tallies to answer questions 3 and 4. The following chart, showing voting data by candidates’ precinct tally size, is from official VBM tally data:

I ask the reader, does this chart, for which the official statewide ballot total is nearly 2:1 in favor of Biden, look right to you? The blue, Biden, bars are much higher than 2:1 for the higher candidate precinct totals. What happened here? Did someone, somewhere, stuff the Vote-By-Mail ballot tallies with Biden votes generating the much higher Biden VBM totals?

Officially, Biden received 185% of the Trump votes (64.9/35.1%.) Here are the data for the chart above. (Pardon the detail, but this is key.) I ask the reader to compare the RH column of the table below with the 185% Biden/Trump ratio for the overall race and draw your own conclusion.

Here is a summary of the last four rows of the table above:

Does the reader believe that the sum of each candidates’ precinct tallies greater than 4,000 votes by mail shows Biden received over 16 times Trump’s votes?

Adding the portion of the large table above greater than 500 votes per candidate per precinct (rows six and below), this totals as follows:

While many of these Biden votes may be valid, the difference, 5.7 million votes, does not include the potential additional Trump votes represented by the earlier statement that the Trump VBM tallies look low for the 700s through 1,000s.

For those who would like answers to the questions above, perhaps the county Registrars of Voters and Secretary Weber can provide voter registrations and ballots beginning with 4,500 precincts for which Vote-By-Mail tallies may be suspect. Here is an initial list. Perhaps we should begin an audit of the November 3, 2020, election with the following three-dozen precincts.

Note 1: All data are directly from complete, official, and released county or state-wide election reports.

Note 2: The zero values for Trump at the bottom three rows of the VBM table indicate he did not have any precinct tallies in the range shown in the first column. The largest VBM precinct tally for Biden is 7,384; for Trump, 4,352.

Note 3: The numbers in all but the last table are grouped not by VBM votes cast, but by VBM votes cast for each candidate.

Note 4: Links to selected county reports follow:

Link to Los Angeles County Votes: Past Election Results (lavote.net)

Link to San Diego County votes: https://rov.sandiegocounty.gov/rov/Eng/Past.html

Link to San Francisco County votes: November 3, 2020, Election Results – Detailed Reports | Department of Elections (sfgov.org)

Link to Orange County votes: Election Results Archives | OC Vote

Images: California Secretary of State and Author.

To comment, you can find the MeWe post for this article here.