Pennsylvania Election Data Analysis: Some Very Suspect Counties

By Stu Cvrk

Much has been written and commented on about voter fraud since Election Day, 3 November, from anecdotal evidence of poll watchers, to observations of concerned citizens, to various whistleblowers. Speculation has also been rampant about electronic ballot tampering, as well as the magical “discovery” of tens of thousands of ballots in key swing states like Michigan, Wisconsin, Georgia, and Pennsylvania. Facts are slowly leaking out, and President Trump’s team has filed lawsuits based on those facts in a number of states, as noted in this summary here. But what about statistical analysis of the votes counted?

At the request of the President’s team last weekend, a technical team was formed to conduct a statistical analysis of certain voter data in key states in order to determine whether there were any significant statistical anomalies evident, and if so, which counties were most suspect. The team's election objective was to help ascertain that all legal votes — and only legal votes, be counted. The team consisted of five statistics experts who were sympathetic to that objective. Based on the datasets they are being provided, the team is looking for “where the bodies were buried” in terms of possible fraudulent behavior.

In one part of the team’s analysis of Pennsylvania, four counties were identified as being suspect: Montgomery, Chester, Cumberland, and Pike. Here is one graph that plotted Democrat-to-Republican registration ratio by Biden/Other Democrat votes.

The key to understanding that graph is this: Biden’s vote totals in the four counties were 1.24 to 1.43 times greater than the totals for Hillary Clinton and Barack Hussein Obama (both elections). This is an absurd result, especially in the three Republican-leaning counties, given that Republican support for President Trump has hovered at 95% over the past year-plus. What is the probability that more Pennsylvanians in these counties voted for a thoroughly compromised candidate who did not even campaign down the stretch than voted for either Barack Hussein Obama or Hillary Clinton?

Another graph plotted the Biden versus Trump ratio of votes by the number of extra votes Biden got.

Dots in the below-the-zero-line reflect counties in which President Trump received more votes than Joe Biden. Blue dots indicate counties that have more registered Democrats than Republicans registered and red dots vice versa. Chester, Delaware, Allegheny, Montgomery, and Philadelphia counties added the most "excess votes" to Biden. Chester is an outlier because it is predominantly Republican. These counties are ripe for an audited recount.

Another aberration in Montgomery county is that some 90%± of registered voters voted. That is an extraordinarily high – and suspicious – percentage. For comparison purposes, the Pennsylvania state average turnout in the 2016 national election was 61%.

The next graph plotted aggregate county vote totals over 92 equal increments of time, as officially reported on 3 November.

The Biden count curve starts about 18% below the Trump count curve and ends up being about 1% above it. It is striking that the two time-series curves that start so far apart end up so very close together; this is NOT a very likely outcome. There was a time period in the middle of the night where counting stopped for several hours (the “Pause”) and then resumed. At that point, it was known how many votes were needed by Biden to win. The jump in the blue curve is the filling in of those needed votes. This cannot be a coincidence. [Note: it takes time to count paper ballots; otherwise, the up-tick in the above graph for Biden would have been more vertical – and more starkly anomalous!]

There are two anomalies related to that time-series graph above. The first is that President Trump “lost” just under 10,000 votes at 21:15 on 4 November, but curiously from three different counties simultaneously: -1,063 Allegheny; -2,972 Bucks; -7,135 Chester. Biden never lost any votes at all over the several days of counting through 7 November. The second is that Biden received a big increase of 27,396 votes at 08:53 on 6 November over one consecutive reporting period. That bump equates to the “jump” in the blue line. Both of those anomalies account for Biden’s 34,000 vote lead in the datasets that were analyzed by the team.

The team also examined Biden votes versus the number of Democratic registered voters. Here is a sampling showing excess votes for Biden over the number of registered Democrats in these counties:

In all these counties, especially in Montgomery and Chester counties, there are more Biden votes counted than registered Democrats. It is not statistically reasonable for Biden to have received more Democratic votes, by 38,500± in Montgomery county and 36,600± in Chester country, then there are registered Democratic voters. Researcher Corey's Digs has also been doing work into this anomaly which will be published later this week.

There were ten Pennsylvania counties in which Biden received excess votes above the Democrat totals in three prior presidential elections. Five counties alone had more Democrat excess votes added than the maximum number of votes added in the 2012 and 2016 elections. The team performed simple linear regressions to compare the number of registered Democrats who voted in those ten counties that delivered the most excess votes for Biden in 2020 versus the Democrat turnout in other Pennsylvania counties. They found that an average 70%± of registered Democratic voters voted in other than those ten counties. In comparison, some 76% of registered Democrat voters supposedly voted in those ten counties that delivered the excess votes for the Biden “win.” That is an unexpected statistical aberration.

Lastly, a Benford’s law analysis was conducted on votes tabulated in heavily-Democrat Allegheny county. In simplified terms, the law asserts that the number 1 should appear more frequently as the lead digit in a precinct voter count. The number 2 should be less frequent, etc. The distribution, with no fraud, should decline exponentially. Here are the Benford graphs from the voter datasets analyzed for Allegheny county:

Note that the shape of the voter count first digits follows Benford’s Law rather closely for President Trump but does not for Joe Biden. This anomalous result lends support for a complete audit of Allegheny county votes tabulated. A team member conducted a more detailed analysis of Pennsylvania’s results (linked here). [Note: additional datasets have been requested by the team in order to complete a more exhaustive Benford’s law analysis across multiple states, as the larger the datasets analyzed, the more accurate and conclusive the resulting Benford’s law analysis.]

Keep in mind that this team — which is doing this all for free, and they all have busy lives otherwise — have only examined Pennsylvania and Milwaukee datasets so far. In both cases, they were given just a few hours to wade through pages of data, graph multiple possibilities, and then write a report. More exhaustive analyses are in process, and more datasets from other states will be forthcoming. The information summarized in this article is just a window into the “anomalies” reported in many swing states over the past week. Still, it is becoming clearer that audits are required to achieve the team’s admirable objective of ascertaining that all legal votes — and only legal votes, be counted. This story is developing….


Stu Cvrk served 30 years in the US Navy in a variety of active and reserve capacities, with considerable operational experience in the Middle East and the Western Pacific. An oceanographer and systems analyst through education and experience, Stu is a graduate of the US Naval Academy, where he received a classical liberal education. This functions as the key foundation for his political commentary. He threads daily on Twitter on a wide range of political topics, such as the military, foreign policy, government, economics, and world affairs.

Twitter: @STUinSD

Get the latest news delivered daily!

We will send you breaking news right to your inbox

© 2024