A Nevada poll can show a close race in the morning and a different result on Election Day. That is not necessarily a contradiction. Understanding how Nevada polling works helps voters separate useful information from campaign chatter and focus on what matters most: whether every lawful vote is properly counted.
Polls are measurements of public opinion at a particular moment. They are not ballots, predictions, or a substitute for participating. A well-run poll can identify trends, test which issues matter to voters, and show where a campaign needs to earn trust. It cannot tell us with certainty what millions of individual Nevadans will do when they cast a ballot.
A poll starts with a sample, not the whole state
Nevada has far too many voters for a pollster to contact everyone. Instead, a polling organization selects a smaller group intended to represent the broader electorate. That group is called a sample.
The quality of the sample matters more than the size of a headline. A poll of 800 carefully selected likely voters may be more useful than a larger survey built from people who volunteered to answer an online questionnaire. Reputable pollsters use voter files, random phone dialing, text invitations, online panels, or a combination of these methods to reach people with different backgrounds and voting histories.
A good sample must account for the Nevada electorate as it actually exists, not as anyone wishes it existed. That means reaching voters in Clark and Washoe counties, where most Nevadans live, while also including rural communities from Elko to Esmeralda County. It means hearing from Republicans, Democrats, nonpartisan voters, Independent American Party members, and voters registered with other parties.
It also means making an effort to reach people who answer calls, people who respond by text, younger voters, seniors, shift workers, military families, and residents whose first language is not English. Every survey method has limits. Phone polls may miss people who do not answer unfamiliar numbers. Online polls may underrepresent people without regular internet access. Strong polling recognizes those limits instead of pretending they do not exist.
How Nevada polling works after responses come in
Once responses are collected, pollsters typically weight the results. Weighting is a statistical adjustment designed to make the completed survey better reflect the likely voting population.
For example, imagine a poll receives too many responses from voters over 65 and too few from voters under 35. The pollster may reduce the influence of the older group and increase the influence of the younger group based on known demographic and voter-registration data. Pollsters may also weight by gender, race, education level, geographic region, party registration, and past voting behavior.
Weighting is necessary, but it is not magic. It depends on sound assumptions. If a pollster guesses wrong about who will vote, the final result can be off even if the math was performed correctly. Nevada makes this especially challenging because turnout can change significantly from one election to another, particularly among nonpartisan voters and voters who have participated less consistently in the past.
That is why the phrase “likely voter” deserves attention. A registered voter is someone eligible and registered to vote. A likely voter is someone the pollster believes will actually participate in the coming election. Pollsters make that judgment using questions about voting plans, interest in the race, prior voting history, and other indicators.
There is no single perfect likely-voter model. One pollster may include a broader group of registered voters. Another may apply stricter screens. Both may be acting in good faith, yet reach different conclusions. When reading a poll, ask who was surveyed, when they were surveyed, and whether the results reflect registered voters or likely voters.
Why Nevada can be difficult to poll
Nevada is a competitive state with an electorate that does not always fit simple national assumptions. Registration numbers provide useful context, but they do not automatically determine turnout or candidate support. Voters can split their tickets, change their minds late, or decide not to vote in a particular race.
The state also has a large number of nonpartisan voters. These citizens should never be treated as an afterthought. They are Nevada voters, and they often decide close contests. Their views are not identical simply because they share a registration category. Some lean conservative, some lean liberal, and many are focused less on party labels than on competence, affordability, public safety, election administration, and whether government follows through.
Nevada’s voting calendar adds another layer. With mail voting and early in-person voting, many ballots are cast before Election Day. A poll taken before ballots begin may capture a different electorate than one conducted after early voting is underway. Events, candidate debates, economic news, and voter outreach can still influence people who have not yet voted.
Language and access also matter. A poll that is available only in English, or only through one communication method, may miss part of the electorate. The same principle applies to election administration: confidence grows when every lawful voter has clear information and a practical way to participate.
Read the margin of error the right way
Most scientific polls report a margin of error. If a candidate receives 48 percent support with a margin of error of plus or minus 4 percentage points, the poll does not establish that the candidate will receive exactly 48 percent of the vote. It gives a range of plausible results, assuming the sample and methodology are sound.
If one candidate is at 48 percent and another is at 46 percent in that example, the race may be effectively tied. The two-point difference is smaller than the stated margin of error. Headlines often turn that into a “lead,” but voters should read it as a close contest.
There is another caution: the published margin of error usually measures sampling uncertainty, not every possible source of error. It does not fully capture a poor turnout model, misleading answers, people who cannot be reached, late shifts in opinion, or mistakes in weighting. That does not make polling useless. It means polling should be read with discipline.
A single poll is one photograph. An average of several credible polls, conducted over time with transparent methods, can offer a clearer view of the race. Even then, the result remains a snapshot, not a verdict.
Questions worth asking before you trust a poll
Before sharing a poll or letting it shape your opinion, look for basic details. Who conducted and paid for it? When was the fieldwork completed? How many people were surveyed? Were they adults, registered voters, or likely voters? How were they contacted? What was the full wording and order of the questions?
A poll paid for by a campaign, party committee, or advocacy group is not automatically wrong. But the sponsor should be disclosed, and the full methodology should be available for inspection. Transparency lets voters judge the work for themselves.
Be especially careful with online “polls” on social media, news websites, or email lists. Those are usually not scientific samples because participants choose themselves. They can show enthusiasm among a particular audience, but they cannot reliably describe Nevada as a whole.
Polls should inform voters, not discourage them
The greatest mistake is treating a poll as permission to stay home. A candidate ahead in a survey still has to earn every vote. A candidate behind can still make the case, organize supporters, and persuade voters who remain undecided. No poll has the authority to decide an election. Nevada voters do.
For the office that oversees elections, the real standard is not whether a pollster called a race correctly. The standard is whether registration records are accurate, voters receive clear instructions, ballot procedures are consistent, paper records are protected, systems are secure, and audits are transparent. Those are practical duties that build confidence regardless of who wins.
That’s why I took the Secretary of State to federal court over 9,607 Nevadans with duplicate active registrations. I’m not alleging fraud. I’m asking for an audit, because accurate rolls are part of counting every lawful vote correctly.
What the Secretary of State polling shows
The race for Secretary of State is a good test of everything above. A Noble Predictive Insights survey taken September 22–26 reported the race two ways. Among 800 registered voters, Cisco Aguilar had 41 percent and Jim Marchant had 32 percent. Among the 708 voters the pollster judged most likely to vote, the numbers were 44 and 37. Same poll, same week, two different pictures. That’s exactly why the likely-voter question matters.
The number I’d watch isn’t either candidate’s share. Among registered voters, 27 percent chose someone else or hadn’t decided. Even among likely voters, it was 18 percent. The published results don’t break out any other candidate by name. Everyone else on the ballot, including me, falls under “other.” A poll that measures a two-candidate race isn’t measuring the race you’ll actually vote in.
That’s not an accusation against the pollster. It’s a reason to read the numbers with discipline, and a reminder that more than one in four registered Nevadans haven’t settled on either party’s candidate.
There’s only one poll that counts
You’ll see more polls between now and November 3, and some will be paid for by people who want you to believe the race is already decided. Read them carefully, ask who paid for them, and then set them aside. The only poll that counts is taken at the ballot box. It has no margin of error, no weighting, and no likely-voter model. Every lawful voter is in the sample, and every lawful vote should be counted.
Nevada deserves civil debate about the direction of the state, followed by election administration that is accurate, open, and worthy of public trust. Study the candidates, ask hard questions, make a plan to vote, and let your ballot speak for you.
When the results are reported, the goal should be simple: every lawful voter should be able to say, with confidence, that the process respected their vote.
I’ve put my plans in writing and taken the fight for clean voter rolls to federal court. Is your candidate for Nevada Secretary of State fighting this hard to protect your vote? If not, it’s time to take a closer look.
Get to know me · Join the campaign · Donate · See all policy positions
I’m not asking you to trust a party on any of this. I’m asking you to hold the office accountable for solving real problems. That’s the job, and I intend to do it.
— Brad Lee Barnhill
Independent American Party of Nevada Candidate for Nevada Secretary of State
#YourIndependentVoice — Results, Not Noise
