Seven days of conversation about Susan Collins and Troy Jackson, split between posts that can be placed in Maine and posts from the rest of the country. A debate on Wednesday and an Obama endorsement on Thursday carried the week, and the two audiences reached opposite conclusions about who had the better of it.
TOTAL VOLUMEPosts, articles and comments across all geographies. Only 5% of the U.S. portion is original writing; the rest is reposts and replies.
+23.1
COLLINS, NATIONALNet sentiment outside Maine. +34.0 on posts naming her alone.
+45.2
JACKSON, NATIONALNet sentiment outside Maine. +38.2 on posts naming him alone.
120K
DEBATE DAYWednesday, Oct 7. The Collins and Jackson debate produced 120K posts, more than four times any other day, and is the only day Collins finishes above Jackson.
What The Polls Say
Published polling, shown to set context. A poll asks a sample of voters which candidate they intend to support. Everything else on this page measures something different: the balance of positive and negative labeled posts about each candidate in the captured conversation. A candidate can lead the vote and still take the worse of the argument online, and where the two diverge this page explains the conversation, not the electorate.
Washington Post-Schar School · field date Oct 9, 2026 · 800 LV · margin of error 3.6 points
The closest race on this page by polling and the furthest apart by conversation. Jackson leads the average by about 2 points, inside the margin of error of every poll in it, while Collins reads 34 points better inside Maine. Both in-state figures rest on about 30 stored documents, so read the direction rather than the size.
The Week In One Line
Net sentiment runs on the left axis from −100 to +100. Gold bars carry total daily
volume for both candidates combined; hover any bar for the narratives driving that day. Jackson leads Collins on five of seven days. The exception is Wednesday, debate day, which is also the largest day of the week by a factor of four.
Collins net sentimentJackson net sentimentTotal posts, articles and comments
Inside Maine
Maine's in-state cells are the thinnest in this race by a wide margin: 30 stored documents behind Collins and 32 behind Jackson, against thousands nationally.
MAINE
Collins
Maine
2,340posts
210 positive
90 negative
Hover a segment
722 unique authors · 200 original posts
(9% of volume; the rest are reposts and replies).
Sentiment from 300 labeled posts, 12.8% of this slice;
stored sample behind the labels 30 documents.
Top 3 narratives driving sentiment
1
She Answers The Allegations On Camera
POSITIVE DRIVER
Collins's strongest in-state material is her own response to the campaign finance allegations, carried by Portland broadcast and drawing the highest engagement of any Maine post in the week.
Sen. Collins further addressed the pay-to-play scheme allegations on Saturday, saying those who know her know she serves the people of Maine with integrity.
The in-state argument for Collins is not about this week at all. It is that the polling was wrong about her once before, and the people making it are largely Maine political operatives rather than voters.
The polls said Susan Collins would lose in 2020. She won by 9 in a state Joe Biden carried by 9, because she doesn't listen to national pollsters.
Collins's in-state negative cell holds nine stored documents in total, and the sharpest concerns her campaign's finances rather than her record. That is a very small base for a figure as strong as +40.0.
Susan Collins's campaign finance director and her husband's boss both conspired with a felon to receive money that he couldn't legally give.
884 unique authors · 190 original posts
(8% of volume; the rest are reposts and replies).
Sentiment from 320 labeled posts, 12.7% of this slice;
stored sample behind the labels 32 documents.
Top 3 narratives driving sentiment
1
Mortgage And Residency Questions
NEGATIVE DRIVER
Maine's own conversation about Jackson is dominated by questions about an FHA loan and where he actually lives. It is pressed by Maine political writers rather than by the Collins campaign, and it is the main reason his in-state figure falls below zero under the stricter test.
Troy Jackson totally fails to defend himself on charges he committed mortgage fraud when he got an FHA loan for an Augusta home.
The counter-argument is that federal investigators found no fault, and the dispute is about whether that is an exoneration. Both sides of it sit inside a cell of 32 stored documents.
Here's the problem: Troy Jackson is lying. The FBI found no fault. The DOJ found no fault. The investigation was done under the Biden administration.
Where Maine argues about policy rather than conduct, the subject is energy prices and specifically solar subsidies. It is the only in-state thread in either candidate's cells that is about what the state pays for something.
Maine's solar subsidies are costing every Maine household an extra $250 per year.
The same week reads differently depending on where you stand. Net sentiment is plotted
from Maine on the left to the rest of the country on the right; the solid line counts every post
naming a candidate, the dashed line counts only posts that name that candidate and not the other.
All posts naming the candidate
Posts naming that candidate only
+16.9Susan CollinsMaine minus national · +37.4 on single-subject posts
Collins reads better in Maine than outside it, +40.0 against +23.1, and the gap widens to +71.4 once posts naming Jackson are removed. Maine is arguing about her record and her 2020 result; the rest of the country is arguing about a ProPublica report. The in-state figure rests on 30 stored documents, so treat the direction as the finding and not the size.
-39.0Troy JacksonMaine minus national · -42.2 on single-subject posts
Jackson is the inverse: +45.2 nationally against +6.3 at home, and he falls below zero in Maine once Collins's name is removed. An Obama endorsement plays nationally; questions about an FHA loan and where he lives play locally. Of the two, he is the one whose home state is least persuaded.
What each audience actually argues about
Geography, not party. Both columns cover the same week and the same two candidates; what separates them is which arguments actually travel.
IN MAINE
A Mortgage, A Contract And The Price Of Power
Maine argues about conduct and cost. Whether Jackson's FHA loan and residence are a problem, whether Collins's campaign finances are, and what solar subsidies add to a household bill. Both in-state cells are very small.
The FHA loanPressed by Maine political writers; the main drag on Jackson's in-state figure.
The pay-to-play allegationsCollins answered them on camera, and the answer is her best-performing in-state post.
Solar subsidiesThe only in-state thread about what something costs rather than who did what.
The 2020 precedentMaine's argument for Collins is that the polling was wrong about her before.
REST OF THE U.S.
An Endorsement And A Report
Nationally the race is two documents: a ProPublica investigation into Collins and an Obama endorsement of Jackson. Neither originated in Maine, and together they account for most of the national volume.
The Obama endorsementThe single furthest-travelling item in the race.
The ProPublica reportCarried by national committees as a pattern rather than an incident.
Energy and taxesParty research clips of Jackson's own words on fossil fuels.
Three debates, one storyWire coverage grouped Maine with Georgia and Michigan.
Both columns watched the same debate and drew opposite conclusions. Maine gave the night to Collins, the only day all week she finishes ahead; the rest of the country moved on to the endorsement the following morning and gave the week to Jackson.
The Rest Of The Country
Nationally Jackson reads twice as well as Collins, and an endorsement rather than a debate is the reason.
NATIONAL
Collins
Rest of United States
134,129posts
12,172 positive
7,611 negative
Hover a segment
63,237 unique authors · 6,600 original posts
(5% of volume; the rest are reposts and replies).
Sentiment from 19,783 labeled posts, 14.7% of this slice;
stored sample behind the labels 3,257 documents.
Top 3 narratives driving sentiment
1
The Pay-To-Play Allegations Travel
NEGATIVE DRIVER
A ProPublica report alleging improper contributions is the organizing national story against Collins, pushed by national committees and super PACs. It is framed as a pattern rather than an incident, and it is the most repeated claim in her national cells.
Susan Collins took thousands of dollars from a medical company before giving them a $4.5 million contract. This is now her third allegation.
Much of what the classifier reads as positive toward Collins is national commentary about her prospects. It names her, discusses her 2020 survival, and concludes she will not repeat it, which is approval of Jackson rather than of her.
Susan Collins didn't collapse in ME-1 in 2020. Polls right now show she is. Troy Jackson is gonna win.
Georgia, Michigan and Maine debated in the same week, and national coverage treated them as one story. Several items in Maine's national cells are wire pieces about all three, which inflates volume without adding anything about this race.
Israel emerges as flashpoint in key Senate races as Democrats face questions over the US alliance.
43,135 unique authors · 4,666 original posts
(5% of volume; the rest are reposts and replies).
Sentiment from 9,946 labeled posts, 11.3% of this slice;
stored sample behind the labels 1,539 documents.
Top 3 narratives driving sentiment
1
The Obama Endorsement Is The Week
POSITIVE DRIVER
Obama endorsed Jackson on Thursday and the post travelled further than anything else in this race. It is the reason his national figure sits at +45.2, roughly double Collins's, and the reason his best day is the day after the debate rather than the debate itself.
Troy Jackson is running to be your next senator, and I'm proud to be on board. In this economy we need a fighter like Troy.
The national case against Jackson is run through party research accounts and is almost entirely about fossil fuels and taxes, using his own words in clipped form.
Maine Democrat Troy Jackson wants to kill American oil and gas jobs and completely eliminate fossil fuels.
The single most circulated moment of the debate is an affordability exchange on electric costs. It is carried by neutral tracking accounts rather than by either campaign, and it reaches both candidates' national cells.
Sen. Susan Collins slammed Troy Jackson for voting against a bill to cut electric costs by $275 a year. Jackson pushed back.
This page is the product of two different kinds of work. One half is data that no public API makes available; the other half is judgment about what the data does and does not support.
What Quid made possible
None of the following could be assembled from public endpoints inside a week.
A 120K-post day isolated from six ordinary onesDebate night separated cleanly, with both candidates measured on the same aligned buckets.
Two audiences on one debateIn-state and national halves that reached opposite verdicts, which is only visible once they are split.
Sentiment scored toward the personEach candidate is defined once, centrally, with primary-term targeting.
Stored evidence per cellThirty documents inside Maine against thousands outside it, which is how the thinness became measurable rather than assumed.
What Claude was responsible for
The data does not interpret itself, and it arrives with traps in it.
Refusing to report Maine preciselyNine stored negative documents cannot support a decimal. The page says so in three places.
Reading Jackson's in-state figure correctlyHis +6.3 falls below zero under the single-subject test, and the page leads with that rather than the headline.
Separating the endorsement from the debateJackson's best day is the day after the debate, not the debate. The page attributes it to Obama rather than to his performance.
Keeping allegations as allegationsThe ProPublica and FHA material is presented as what was reported and disputed, not as established fact.
Neither half can establish cause. This page measures what was said, how much of it there was, and who said it. It is not a poll, it does not model turnout, and it does not predict the November result.
How to read this
Source. Quid, run against the full-resolution recent index. Each person is defined
by an ad-hoc boolean query with primary: sentiment targeting, so the sentiment is scored
toward that named figure rather than toward the post as a whole.
Window. Oct 3 - Oct 9, 2026, seven complete day buckets. Quid closes its day at 23:00 UTC and
ignores the timezone parameter, so each bucket covers roughly 23 hours of the labeled calendar day.
Net sentiment is (positive − negative) / (positive + negative) × 100, computed on
labeled posts only. Neutral posts are excluded from the meter but counted in the volume figures.
Sentiment coverage. Sentiment labels cover 13 to 15% of Collins's posts and 13 to 11% of Jackson's.
Both in-state cells are very thin. Collins rests on 21 stored positive and 9 negative documents inside Maine; Jackson on 17 and 15. Her +40.0 and his +6.3 are directions, not measurements, and a handful of posts would move either of them substantially.
Classifier attribution is imperfect. In a two-candidate race most posts name both people, and
the classifier sometimes assigns a post's sentiment to the wrong subject, and several posts attacking one
candidate are labeled positive for them. Every soundbite shown here was read and placed against the narrative
it actually supports, but the donut and meter totals are the classifier's unedited output.
Twitter/X counts are projected ×10 from a 10% sample, so volume figures exceed the number of
individually stored documents. Counts and samples are never mixed inside one ratio.
Sensitivity test. 31% of Collins's posts and 63% of Jackson's name both candidates, where the classifier can credit sentiment to the wrong one. Re-running every figure on posts that name one
candidate only pulls the two apart inside Maine. Collins rises from +40.0 to +71.4 while Jackson falls from +6.3 to -4.0, below zero. With 63 percent of Jackson's posts also naming Collins, his modest in-state positive is largely borrowed from a conversation about her.
Volume is amplification, not voices. Only 5% of posts are original; the rest are reposts and replies. Unique-author counts
are shown on every panel and are the better measure of how many people are actually talking.
Link integrity. Every source link on this page was checked for liveness at publication.
Exploratory. Figures may not match Quid's own dashboards exactly. Sentiment is a
measure of how people are talking, not a poll, and nothing here projects a November result.