Republicans kept 85% of Trump's share. That isn't evidence of fraud.
A post Trump shared says one formula fits 59 races. We rebuilt its 30 Senate races. They fit, and so did 2022's and 2018's.
![Trump reposted, without comment, a screenshot of this post by Truth Is Freedom (@TruthAnswersAll) on X. A second repost showed its chart.]
THE 15% BAND
Look at this chart.
59 races. Senate and governor.
States ordered from reddest to bluest in 2024.
The solid lines are 2024.
The dashed lines are 2026.
The shaded bands are what moved.
One formula fits all of it:
Republican 2026 = 0.85 × Trump 2024
Democrat 2026 = Harris 2024 + 0.15 × Trump 2024
For every 1,000 Trump points, 850 stayed Republican. 150 landed on the Democratic side.
The numbers:
Senate: Republicans kept ~85% of Trump’s share
Governor: ~86%
Variation explained: 88%
What makes it strange:
It didn’t matter how red the state was. Wyoming, Ohio, Illinois, Massachusetts: same 15% (p = 0.69). The Democratic gain was sized by Trump’s vote, not Harris’s. The Harris part is zero.
The 0.85 beat both standard election-swing models in every set of races.
Drop any single state and the multiplier stays between 0.846 and 0.853.
Why seats flipped:
At 0.85, any state where Trump got under about 58.5% falls below 50%. Texas, Kansas, Iowa, Ohio, South Carolina, North Carolina: all under that line. All flipped.
What this means:
Midterms are messy. This one isn’t. Two sets of races, different candidates, different campaigns, and the same 15% decrease almost everywhere.
A pattern this even demands answers at the ballot level: paper ballot hand counts, poll-book check-ins against votes counted, and voter files showing who actually turned out.
The numbers raise the question.
The ballots can answer it.](/posts/fifteen-percent-band.jpg)
The bottom line
- Missing context"One formula": 0.85 × Trump
- Misleading"Same 15%" everywhere
- MisleadingThe 58.5% flip line
- Wrong"Algorithm" (@WallStreetApes)
The post's math checks out for its 30 Senate races. Republicans kept about 85% of Trump's 2024 share, a rule that explains 89% of the variation. But races ran from 71% to 101%, and the same test explains 92% of 2022's results and 73% of 2018's. Nothing in the data points to switched votes.
The claims
What the post says or clearly implies, checked one by one. How verdicts work
One formula fits all of it: Republican 2026 = 0.85 × Trump 2024, and it explains 88% of the variation.
Missing contextThe math holds up: we get 0.85 and 89% for the post's 30 Senate races. Fits like this aren't rare. The same test explains 92% of 2022's Senate results and 73% of 2018's.
The post's chart uses each party's share of the two-party vote, in 30 Senate and 29 governor races1,2. With unofficial counts read Nov. 4–8 (sources in the chart below) and Trump's official 2024 results3, the best fit for those 30 Senate races is 0.848 times Trump's share. It explains 88.5% of the variation, by our calculation. The chart says its governor shares were estimated from NBC's vote shift, not counted1.
States vote much as they did two years before because most voters stick with a party. From 2012 through 2024, 91% of Senate races (222 of 245) went to the party that won the state's last presidential race4. On 2022, the post's test puts Republicans' Senate shares at 1.03 times Trump's 2020 share, explaining 92%5,6. For 2018 it's 0.89 times his 2016 share, explaining 73%7,8.
The Democratic formula isn't a second finding. In two-party shares, whatever Republicans lose, Democrats gain, so "Harris 2024 + 0.15 × Trump 2024" follows from the Republican formula by arithmetic. That's why "the Harris part is zero."
The 0.85 fit is real. By R², 2022 fit its line better, 2018 worse.
Show the numbers
| Race | Trump 2024 | Republican 2026 | Kept |
|---|---|---|---|
| Rhode Island | 42.9% | 30.5% | 71% |
| New Jersey | 47.0% | 36.0% | 77% |
| Texas | 56.9% | 43.8% | 77% |
| North Carolina | 51.6% | 40.0% | 77% |
| New Mexico | 46.9% | 36.7% | 78% |
| Georgia | 51.1% | 41.0% | 80% |
| South Carolina | 59.1% | 47.7% | 81% |
| Mississippi | 61.6% | 49.9% | 81% |
| Kansas | 58.2% | 47.4% | 81% |
| Virginia | 47.1% | 38.5% | 82% |
| Colorado | 44.4% | 36.7% | 83% |
| Louisiana | 61.2% | 50.9% | 83% |
| Alabama | 65.4% | 54.8% | 84% |
| Iowa | 56.7% | 47.9% | 84% |
| West Virginia | 71.3% | 60.6% | 85% |
| Illinois | 44.4% | 37.8% | 85% |
| Ohio | 55.7% | 47.5% | 85% |
| Arkansas | 65.7% | 56.1% | 85% |
| Delaware | 42.5% | 36.4% | 86% |
| Kentucky | 65.5% | 56.2% | 86% |
| Tennessee | 65.1% | 56.2% | 86% |
| Oklahoma | 67.5% | 59.3% | 88% |
| New Hampshire | 48.6% | 43.0% | 89% |
| Wyoming | 73.5% | 65.2% | 89% |
| Massachusetts | 37.0% | 33.7% | 91% |
| Florida | 56.6% | 51.4% | 91% |
| Michigan | 50.7% | 46.4% | 91% |
| Minnesota | 47.8% | 44.8% | 94% |
| Oregon | 42.6% | 40.0% | 94% |
| Maine | 46.5% | 46.9% | 101% |
It didn't matter how red the state was: the same 15% decrease almost everywhere.
MisleadingHow red a state was didn't predict the drop. It ran from 71% of Trump's share (Rhode Island) to 101% (Maine). That spread was only a little tighter than 2022's or 2018's: 6.1 points, against 7.4 and 8.7.
By our calculation, Ken Paxton kept 77% of Trump's two-party share in Texas, and Susan Collins 101% in Maine9,10,3. Jack Reed's opponent in Rhode Island kept 71%11. The post's four examples ran from 85% in Ohio and Illinois to 91% in Massachusetts, with Wyoming at 89%12,13,14,15.
That evenness is ordinary. Nineteen of the 30 races came within 5 points of the fitted 85%. In 2022, 21 of 32 Senate races came within 5 points of that year's fit, and in 2018, 20 of 326,8. The ratios' standard deviation was 6.1 points in 2026, against 7.4 in 2022 and 8.7 in 2018, by our calculation.
In NBC's national exit poll, 7% of Trump's 2024 voters backed Democrats for the House, and Trump voters were 44% of the electorate to Harris voters' 49%16. If a similar share of Trump voters switches or stays home in every state, the Republican share shrinks by a similar fraction everywhere.
Any state where Trump got under about 58.5% fell below 50%: Texas, Kansas, Iowa, Ohio, SC and NC all flipped.
MisleadingThe six named Senate seats did flip, but the line isn't a rule. It's the average applied in reverse, and it misses races: Republicans held Florida's Senate seat and governorship below it, and lost South Carolina's Senate seat above it.
At 0.85, the break-even is 50 ÷ 0.85, or 58.8% of the two-party vote. Democrats won the Senate races in all six states the post names17,18,19,20,21,22. On the chart's own two-party basis, Trump had 59.1% in South Carolina, above the line3.
Trump had 56.6% in Florida, and Republicans won its Senate and governor's races23,24. In Mississippi, where Trump had 61.6%, the Democrat led an uncalled Senate race by 2,696 votes as of Nov. 625. South Carolina's governorship stayed Republican26. The chart left out seven governor's races and named three. One was New Hampshire's, where Republican Kelly Ayotte won about 107% of Trump's share, by our calculation27,28.
As @WallStreetApes spread it: "For every 1000 Trump votes, 150 are algorithmically switched to Dems."
WrongThe pattern is the only evidence offered, and a similar fit shows up in 2022 and 2018. Races ranged from 71% to 101%, not a fixed 15%. No audit or court has reported switched votes.
The post Trump shared doesn't say "algorithm." The original said "Trump points," meaning shares, and that "a pattern this even demands answers at the ballot level"29. @WallStreetApes called it the "Democrats cheating algorithm"30. Democracy Docket said the post "excluded some races that didn't fit the pattern"31. Its Senate exclusions don't show that. Idaho, Montana, Nebraska and South Dakota had independents as the main challenger32,33,34,35, and Alaska, still counting, fits at 0.8836. But the governor races it left out include New Hampshire, where Ayotte ran ahead of Trump1.
Vote counts can't settle it. Midterm turnout usually falls, so Republicans lose raw votes even without switching. In the Texas exit poll, 11% of Trump's 2024 voters backed Talarico37. Exit polls are weighted to the count, though, so they fit it without confirming it. Checking the paper ballots can.
Paxton, Roger Marshall in Kansas and Michael Whatley in North Carolina conceded17,18,22. Newsweek reported that, in the posts it saw, Trump hadn't said the results were fraudulent38.
The real disagreement
Facts can't settle this part, so we don't rule on it.
When a result surprises one side, should a statistical pattern be enough to trigger full hand counts, or should checks follow rules set before the vote?
Every state runs some kind of post-election audit39. Georgia planned 20 risk-limiting audits this fall, up from two in May's primaries40. Kansas law adds a hand audit of 10% of precincts in any race within 1% on the election-night tally, as determined by the secretary of state41. Its governor's race was 0.7 points apart as of Nov. 7, inside that range if the election-night tally matched42. The post asks for more: hand counts, poll-book check-ins against votes counted, and voter files showing who turned out29.
Hand-count when doubt runs high
Paper ballots exist to be checked. Visible, ballot-level counts after a surprising result are the surest way to earn back trust.
Trust is shaky on both sides. In July, Pew found 55% of Republicans and 58% of Democrats confident that this year's midterms would be run fairly and accurately43. Most audits compare a sample of ballots with machine counts. The post asks for a different check: poll-book check-ins matched against votes counted29.
Georgia's 2020 hand count of about 5 million ballots confirmed the winner, and it caught real mistakes: four counties turned up a few thousand uncounted votes44. The conservative Heritage Foundation says thorough election audits "should be a routine practice in every state"45. The left-leaning Brennan Center says audits "help restore public confidence in elections"46. Make full checks routine and public, and finish them before certification. Then asking for one stops sounding like an accusation.
Follow rules set before the vote
Every state already audits. Extra full counts triggered by a pattern or a loss cost more and can count worse.
All 50 states and D.C. "conduct some type of post-election audit," NCSL reports, and they're "conducted regardless of the margins of victory"39. That treats both parties alike. Hand counts aren't the gold standard they sound like. In a study of New Hampshire recounts, machine-scanned ballots differed from the recount by 0.56%, hand-counted ones by 0.87% (0.83% historically)47. In a 2023 test in Mohave County, Ariz., counters made errors in 46 of 30,600 races, and a full hand count was priced at $1,108,48648.
Arizona's 2021 Senate-ordered review of Maricopa County found Biden's margin "slightly larger"49. Risk-limiting audits already widen toward a full hand count when a sample doesn't confirm the result50. Under those rules, the sample decides when to count everything.
Where they agree
- Both sides want paper ballots that can be checked, with the checks finished before results are certified.
- Neither side wants a check that only runs when its own candidates lose.
What would change minds
- Results of the audits already under way, such as Georgia's risk-limiting audits and any close-race audit in Kansas, compared with the machine counts.
- Precinct-level evidence, such as poll-book check-ins that don't match ballots counted, rather than patterns in statewide shares.
- Studies of whether full hand counts after a disputed result raise trust among the losing side's voters.
Sources
Every source was opened and read. Advocacy groups are labeled as such.
Show all 50 sources
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38
- 39
- 40
- 41
- 42
- 43
- 44
- 45
- 46
- 47
- 48
- 49
- 50
Corrections
- Nov 8, 2026: Shortly after publication, we changed our verdict on the post's 58.5% flip line from Missing context to Misleading. The post says any state below the line fell below 50%, and Florida, where Republicans held both statewide races below it, contradicts that. The facts in the piece did not change.
Think we got something wrong? Email editor@viewfromthegallery.com with a source. Corrections are made openly and logged on the page.
Get new breakdowns by email
New breakdowns in your inbox: what was true, what wasn't, and what the argument is really about. No ads, no spin.