See also the main article on Maine, for documented election integrity issues and further reports.
Former US military intelligence officer and statistical analyst Seth Keshel analyzed the trends of voter registrations versus actual votes. The following report indicates which counties’ vote counts align with the trend in voter registrations and which have small or large divergences. Each county is flagged as red, yellow, or green, based on whether the 2020 vote totals aligned with the expected trends (as past decades demonstrate they typically do), or if they diverged in statistically unlikely ways.
|Trump votes||Biden votes||Other votes|
|Officially reported results||360737 (44.0%)||435072 (53.1%)||23761 (2.9%)|
|Estimate of potential fraud||31000 (3.8%)|
|Estimate of actual result
(with fraud removed)
|360737 (45.7%)||404072 (51.2%)||23761 (3.0%)|
Statewide, Maine’s election numbers trended as registration indicates, with the state moving 2.3% left in registration, with most third party voters coming back to support Biden, at least on paper. New Hampshire’s main suspicious issue was in mass volume of registrations dwarfing those from 2012-16. The one that stood out this year was Cumberland, a 27k to 3k Dem registration advantage.
Trump is up 25k from 2016, Biden up over Clinton by 77k, and above Obama’s 422k from 2008’s 17.3% blowout. Trump took the single electoral vote from ME-2 on both campaigns.
Cumberland and York, in southern area of the state are about 15k and 6k over expected outcomes.
If Biden is 31k heavy, an accurate victory margin for him would be 5.6%, or about 51.6% to 46.0%. If there are votes being flipped, it could be a different story.
Best audit target – Androscoggin (3k heavy), Penobscot (2k heavy)
Top 100 Worst Counties
Seth has listed Maine’s Cumberland County as being among the Nation’s Top 100 Worst Counties in terms of abnormal trends during the 2020 election.
Seth Keshel’s estimates are based on the percentage of voters who are registered to each party (where it’s possible to obtain this information) compared to the actual votes for each party. He examined these party trends over the last two decades, as well as population growth, which brought to light the strange and statistically unlikely outliers and anomalies that occurred in 2020. We explain this process, step by step, with visual examples, in our guide How to Predict Election Results Using Registration Data, so that you can investigate the numbers for yourself.
This video from Telegram also explains some of Seth’s approach to analyzing precincts.
We aim to publish links to both the raw election data and voter registration data for Maine so that citizens and researchers can analyze this information for themselves.
|Certified Election Results||
In most states, the certified election results are available from the Secretary of State or State Board of Elections. Check their website for details. States are also required by HAVA law to inform the public of how many absentee ballots were both sent and received to uniformed services and overseas voters.
|Voter Registration Rolls/Database||
In some states, these are freely available from the Secretary of State or State Board of Elections. Check their website. In other states, voter rolls must be purchased and/or accessed via a signed legal agreement. Some officials are also obstructing access to the rolls, to make auditing difficult. Let us know via Telegram or via the comment section below if you experience issues.
|Cast Vote Records||Ordros Analytics has collated a repository of Cast Vote Records (CVRs) which list everyone who voted in the Nov 2020 election. Only some counties in some states are represented, but the list is growing.|
|The New York Times 2020 Election Results||Results for all states, with several maps and charts.|
|Data Explorer Tool||Our own tool for inspecting the 2020 New York Times data, including the time-series data of how the counting progressed. Also provides download links for raw JSON or CSV data, including counts for every precinct and county.|
|US Election Atlas||
Recommended when doing trend analysis, as shown on this page
Detailed results for 2020 and previous years. Some data is freely accessible on their website, while some, such as detailed historic results in CSV format, are purchasable for a fee.
This appears to be the source commonly used by Seth Keshel for his analysis, although we have not officially confirmed this.
|2020 General Election Data & Research||A broad collection of national stats, vote and registration counts, time-series data, voting machine information and manuals, PDF reports, and other research collated by citizen auditors.|
|Election Night Time-Series Data from Edison||
Provides more detail than is available from the New York Times, and includes numerous interactive charts.
Download ZIP of Raw Data
Published by Jeff O’Donnell, MagaRaccoon.com
|Weekly HAVV SSN Reports||
Social Security Administration (SSA) Weekly Data for Help America Vote Verification (HAVV) Transactions by State.
Learn more about this data
Published by Jeff O’Donnell, MagaRaccoon.com
- A team of scientists from Election-Integrity.info produced the report “2020 Presidential Election Startling Vote Spikes” which includes a section on Maine
Volunteers are needed to help verify the irregularities found. One key way this is done is through voter canvassing, with teams analyzing the county and state records and voter rolls, and others going door-to-door to identify whether the records match the actual residents living at the address.