Alaska 2026: How to Read Ranked-Choice Polls, and What the Repeal Measure Would Change
Dated editorial analysis; figures in this article are not recalculated automatically. Visit the linked race pages for current data.
Alaska is the one state in our tracker where a poll can report two different answers to the same question. Under the state's system, a voter ranks candidates, and the count can proceed through several rounds. A pollster can therefore publish first choices, a simulated final round, or both. Reading Alaska's numbers correctly starts with knowing which one you are looking at.
How the system works
Alaska voters approved a 2020 ballot measure that replaced party primaries with a single open primary. The four candidates with the most votes advance to the November general election, where voters may rank them in order of preference. If no candidate has more than half of the first-choice votes, the last-place candidate is eliminated and those ballots move to each voter's next-ranked choice. The process repeats until one candidate has a majority of the ballots still in play.
That has two practical effects. First, a candidate can lead on first choices and still finish second after the transfers. Second, the final tally can take longer than in most states; our election-night timing guide explains why some tabulations come after election night.
The repeal measure on the ballot
Alaskans are also voting on a 2026 ballot measure that would restore party primaries and end ranked-choice general elections. A 2024 repeal attempt was narrowly defeated, with about 49.9% voting yes and 50.1% voting no. The Alaska Supreme Court reviewed the 2026 ballot language this summer and ordered one addition: the summary must say that governor and lieutenant governor candidates would run separately if the measure passes. We take no position on the measure; the point for poll readers is that the rules for counting votes are themselves a question on the same ballot.
What the recent polls show
Our data date is October 28. Only one verified Alaska poll ended within 14 days of that date: Data for Progress, fielded October 8 to 16 among 694 likely voters, using an SMS and web panel. It covered both statewide races.
- In the Alaska Senate race, former U.S. Rep. Mary Peltola (D) and Sen. Dan Sullivan (R) were at 47% and 44% on first choices. In the simulated final round, Peltola was at 51% and Sullivan at 49%.
- In the Alaska governor's race, an open seat because Gov. Mike Dunleavy (R) is term-limited, the poll had Jonathan Kreiss-Tomkins (D) at 46% and Bernadette Wilson (R) at 28% on first choices. Because several Republicans are on the ballot, the simulated final round was Kreiss-Tomkins 53%, Wilson 47%.
With a single poll in the window, there is no average to compute, and one survey of fewer than 700 people carries a margin of error of roughly four percentage points on each candidate, so the final-round gaps are within that range. The Senate gap of two points is smaller than the sampling error of the poll.
Earlier polls, different formats
Our tracker holds three other verified Senate polls from September and early October, and they are not directly comparable. InsiderAdvantage (October 9 to 10, 1,200 likely voters) reported a final-round result of Peltola 50%, Sullivan 48%. Trafalgar Group (September 22 to 24, 1,079 likely voters) showed Peltola 45.5% and Sullivan 46.8% on a standard ballot. co/efficient (September 14 to 17, 799 likely voters) asked a two-candidate question and had Peltola at 46% and Sullivan at 49%. Mixing these formats would blur the picture, so we keep final-round simulations and first-choice figures apart rather than blending them into one number.
Why the ballot format matters for the governor's race
The governor's contest is a good illustration. Several Republicans and one Democrat are on the ballot, so the party-split first-choice numbers look very different from the final round. A first-choice figure of 28% for the leading Republican does not mean that candidate has a weak position in a head-to-head matchup; it reflects votes divided among more than one candidate from the same party. Conversely, a first-choice lead can narrow when eliminated candidates' ballots move. Neither number is wrong; they answer different questions.
What to watch
- Whether new polls report first choices, final rounds or both, and whether they say how they simulated the transfers.
- How large the share of respondents who rank only one candidate is, since ballots that stop ranking drop out of later rounds.
- The timing of the official count, which depends on the Division of Elections' schedule for absentee and overseas ballots.
Ratings from Cook Political Report and Sabato's Crystal Ball differ on the governor's race (Likely R and Lean R respectively), and both rate the Senate race a Toss-up. Ratings are forecasters' judgments, not results, and this site does not predict outcomes.
Poll data as of October 28, 2026. Only verified polls ending within 14 days of that date are counted; see each race page for full poll lists and sources.
Relevant race pages: Alaska Senate · Alaska Governor