How to Predict Election Results Using Registration Data

Did you know that the numbers of voters registered for each party is a strong predictor of a final election result? In this article you’ll learn how to use these party registration numbers to predict the election results for a state or county. We’ll also compare this prediction with actual 2020 election results to assess the likely validity of those results, and derive a formula for an estimate of “excess” votes.

Seth Keshel used this method to correctly predict the outcome of all 50 states at the 2016 General Election, as well as his heat maps of suspicious counties in 2020. This section is heavily inspired from his work.

Let’s jump right in with a case study of Westmoreland County, Pennsylvania.

Introductory Walkthrough — Westmoreland County, PA

In order to perform this analysis we start by collating the total votes by party and total voter registrations by party for the last several years, for the particular county we’re interested in.

This data can usually be downloaded directly from your Secretary of State’s website, or from US Election Atlas which has a lot of free information online or detailed spreadsheets available for purchase.

Here’s our data for Westmoreland County, from 2000–2020:

We’ve abbreviated the column names, but here’s what they mean:

TR	Total Registrations
RR	Republican Registrations
DR	Democrat Registrations
OR	Other Registrations
TV	Total Votes Counted in Election
RV	Republican Votes
DV	Democrat Votes
OV	Other Votes

From this information we can then calculate a few other numbers:

• The increase/decrease in votes and registrations between elections for each party
• The percentage share of votes and registrations each party has
• The “Votes / Registrations” ratio for each party (sometimes referred to as the “turnout”)

Here’s our updated spreadsheet:

Here are the abbreviations for each additional column. (These abbreviations are used throughout the article, so refer back to this table as needed.)

RRP	Republican Registrations as Percentage of Total
DRP	Democrat Registered as Percentage of Total
ORP	Unaffiliated (Other) Voters as Percentage of Total
PGap	Percentage Gap between Republican and Democrat Registration Percentages (RRP – DRP)
TRI	Total Registration Increase (the change from the previous election)
RRI	Republican Registration Increase (the change from previous election)
DRI	Democrat Registration Increase (the change from previous election)
ORI	Unaffiliated (Other) Registration Increase (the change from previous election)
RVP	Republican Votes as Percentage of Total
DVP	Democrat Votes as Percentage of Total
OVP	Unaffiliated (Other) Votes as Percentage of Total
TVI	Total Vote Increase (the change from previous election)
RVI	Republican Vote Increase (the change from previous election)
DVI	Democrat Vote Increase (the change from previous election)
OVI	Unaffiliated (Other) Vote Increase (the change from previous election)
T/TR	Total Turnout (Votes) Divided by Total Registrations (TV ÷ TR)
TI/TRI	Turnout Increase Divided by Registration Increase (TVI ÷ TRI)
T/RR	Republican Turnout (Votes) Divided by Republican Registrations (RV ÷ RR)
T/DR	Democrat Turnout (Votes) Divided by Democrat Registrations (DV ÷ DR)

Some of these columns are important to understand, so let’s explain them.

PGap

This column is very important and features pre-eminently in a lot of Seth Keshel’s videos.

PGap is calculated by subtracting the Democrat Registration Percentage from the Republican Registration Percentage (RRP — DRP).

By looking at the PGap values over the different years, we begin to see a trend. From 2000 to 2020, we can see the Democrat Registration Percentage is decreasing and the Republican Registration Percentage is increasing at each election.

The PGap went from a Democrat lead of 26.9% in 2000 to a Republican lead of 6.4% in 2020.

Registration Numbers

If we look at the raw registration numbers by party, we can see that there has been an exodus from the Democrat party to the Republican party.

Keep these 2020 numbers in mind:

• 15,880 new people registered with the Republican party
• 10,167 people de-registered from the Democrat party

Vote Increase

Now let’s look at both the Registration Increase as well as the Vote Increase for each party:

What do you notice when you look at the trend?

Here we begin to see some interesting things:

• The RRI and RVI move mostly in the same direction (either positive or negative) and have roughly similar values for every election since 2000.
• The DRI and DVI values move mostly in the same direction also for every election since 2000, except for 2020.

Here we find a strong exception to the prevailing trend: In 2020, the Democrats lost 10,167 registrants, but increased their vote count by 12,523.

Since the population in Westmoreland has been decreasing over the last 10 years, where did these extra votes come from?

“Votes / Registrations” Ratio

Finally, the T/RR and T/DR columns represent the “turnout” — which is votes divided by registrations — for the Republican and Democrat party, respectively.

What trend do you notice in the yellow highlighted numbers?

We see the T/RR value (Republican turnout) increased slightly up until 2008, after which it slowly but consistently decreased. The T/DR (Democrat turnout) remained fairly constant around 0.53 since 2000 … except in 2020 where it jumped to 0.70. That is a 32% increase from the previous election — a very large shift!

💡 TIP: From all our analysis across many counties and states we have found the “Votes / Registrations” ratio to be very stable and reliable. Big jumps in value are uncharacteristic and warrant a closer look.

These two turnout columns are extremely useful for:

• Highlighting potential fraudulent activity
• Estimating excess or shortfall votes for either party

Analysis

Now that you have a grasp of what the columns mean, let’s dig a bit deeper and see what else we can find. We’ll highlight any cells that stand out or diverge starkly from the trend.

Observations

• As we noted earlier, the Democrat Vote Increase (DVI) strongly contradicts the downward change in registrations (DRI). Democrats have been consistently losing votes in the thousands since 2008, but in 2020, with 10,167 less registrants, they managed to increase their vote count by 12,523. ⚠️ This completely violates the PGap and DRI trends.

• The RRI and RVI values align with the trend as we would expect

The increase of 12,523 Democrat votes is very suspicious and explains why the T/DR ratio is so high.

If the T/DR ratio remained on the same trend, it would be around 0.53. We can use this to calculate an estimate of the excess votes.

Estimate Democrat Votes (EDV):	DR * T/DR	=	103,415 * 0.53	=	54,810
Excess Democrat Votes:	DV — EDV	=	72,192 — 54,810	=	17,382

We could now say that in Westmoreland County, the Democrats appear to have 17,382 more votes than the trend would indicate.

Note that we have not accounted for a potential shortfall of Trump votes, but you could calculate that also if desired.

---

Now that we’ve explained the process of performing a “Registration Trend” analysis, watch the corresponding video from Seth Keshel that brings it all together:

or Watch on Rumble

Here is an image that shows how the table of data we have been using matches with the table Seth uses in his presentations:

Seth Keshel’s Videos

You can watch all of Seth’s county analysis videos on his archive Telegram channel or on our website.

Here are a handful to get you started (each is around 7 mins):

• Cochise County, AZ
• Henderson County, NC
• Mesa County, CO
• Nueces County, TX
• Erie County, PA
• Berks County, PA
• Westmoreland County, PA

Even though Seth places a lot of emphasis on the PGap trend, we have found that it is not always reliable for all states. Some states do a good job of keeping their voter rolls up to date. Others don’t.

When states keep their voter roll records up to date, the PGap trend is powerful. However for states that don’t, we have found the T/RR and T/DR trends to be more reliable.

For states that do not keep any party registration information, we can fall back on the RVI and DVI trends instead.

Case Studies

For further registration trend analysis for specific states and counties, see the reports below. (We will be adding more states and counties as time permits.):

Summary

Historical registration information can be used very effectively to identify counties that are “bucking” their own historical trends.

We explained the formula that can give us an estimate of excess (and shortfall) votes, based on the “Votes / Registrations” ratio (i.e. the T/DR and T/RR ratios).

From our own in-depth analysis that has been going on for many months now, we’ve observed this ratio to be remarkably stable over the years, except in 2020 where the ratio increased significantly for the Democrat party, in many counties, in nearly every state.

With so many counties bucking trends and setting new turnout records, combined with the reports of poor security and election mishandling, the results do look suspect. We believe contacting voters directly (canvassing) combined with forensic audits (where necessary) is crucial to conclusively determining the validity of the election.

Our next article reveals the counties that have the greatest registration and vote increase trend anomalies.

---

Other Articles In This Series

Overview	Identify Electoral Fraud Using Trend Analysis Diving deeper into the unusual trends and statistics discovered in the 2020 election.
Part 1	The Fall of the Bellwether Counties A look at the surprising failure of the bellwether counties in 2020, and what that tells us about the Presidential election outcome.
Part 2	The Battle for the Largest Counties The data shows the Democrats are winning less and less counties at each election, but are winning more and more of the largest counties. How is this possible?
Part 3	The Curious Case of the 2020 “Voting Rate” Blowouts Voter turnout rates shot up dramatically in many states in 2020. We look at the voting rates since 2000 to find out which states set a new record.
Part 4	When Winning Margins Go “Off the Charts” Learn how to "normalize" a county's winning margin to identify abnormalities. What does this reveal about the 2020 election?
Part 5	How to Predict Election Results Using Registration Data Learn how to use the party registration numbers for each county to predict the election results for a state and assess the likely validity of the results. This is a guide on the method Seth Keshel uses for his predictions and county heat maps, allowing you to dig into the trends of your own county and uncover anomalies.
Part 6	The Counties Where Votes and Party Registrations Don't Align We build on some of the previous techniques to scan 3,111 American counties, identifying those whose shift in vote totals moves unexpectedly against the shift in party registrations.
Part 7	Investigating the Large Democrat Vote Increases We compare the Democrat vote totals with previous elections which reveal some very large increases in unlikely places.
Part 8	Unlikely Z-Score Values We compare key parameters in the 2020 results against the previous five elections using the z-score, and find hundreds of counties breaking statistical norms.
Part 9	Down Ballot Election Analysis Interesting statistical findings on how the Presidential race results compared to the other races in the same election.
Part 10	Voter Roll Analysis Coming soon.
Part 11	The Art of the Steal Coming soon.

  Follow us on Telegram to be notified when we release the remaining articles.

Visitor Comments

Did you find this article helpful? Did you dig in and review the data for your county or state? What did you find? Let us know in the comments below.

Show Comments