October 26, 2026. Final forecast before the October 27 election. Evidence cutoff: 2026-10-26T00:41:48+02:00 (Jerusalem time).
After months of polls, the final forecast leaves the central question unresolved. Neither the Netanyahu-aligned bloc nor the Zionist Opposition reaches 61 seats in either approach's central estimate. The campaign is ending; the arithmetic still offers no straightforward majority.
The vote-led R forecast puts the Netanyahu-aligned bloc at 53.6 expected seats and the Zionist Opposition at 51.6. D, which also uses eligible additional seat-poll evidence, puts them at 54.0 and 51.4. Both approaches are published side by side. Their ranges show how much room remains around these averages.
| Grouping | R: expected seats (80% range) | R: 61+ | D: expected seats (80% range) | D: 61+ |
|---|---|---|---|---|
| Netanyahu-aligned | 53.6 (48–59) | 5.4% | 54.0 (49–59) | 4.4% |
| Zionist Opposition | 51.6 (46–57) | 1.7% | 51.4 (47–56) | 0.6% |
| Zionist Opposition + Arab lists | 64.4 (59–70) | 81.4% | 64.5 (60–69) | 85.2% |
The sharper question is how often a grouping actually reaches a majority. The Netanyahu-aligned bloc reaches at least 61 in 5.4% of R simulations and 4.4% under D's weights. For the Zionist Opposition alone, the corresponding figures are 1.7% and 0.6%. A lead over the other bloc and a majority of the Knesset remain different thresholds to cross.
Adding Ra'am and the Joint List to the Zionist Opposition yields 64.4 expected seats in R and 64.5 in D, with 61-seat probabilities of 81.4% and 85.2%. That larger total makes the distinction between seats and political agreements consequential. This is an arithmetic grouping; the model does not estimate whether these parties would cooperate or form a government. The Zionist Opposition itself comprises Yashar, Beyachad, Yisrael Beiteinu and the Democrats.
This final release incorporates fuller evidence from polls already in the record. Eight official CEC disclosures replace eight earlier representations of the same surveys. Seven supply usable national vote-share tables. Their fieldwork belongs to October 21–23; the additional information was verified for this release. No survey receives a second budget merely because its PDF arrived later.
Compared with the October 25 issue, the R central estimates move from 54.5 to 53.6 for Netanyahu-aligned parties and from 51.5 to 51.6 for the Zionist Opposition. In D they move from 54.6 to 54.0 and 51.5 to 51.4, respectively. These changes reflect the newly verified vote evidence, its effect on calibration and the passage of a day. They should be read as a completion of the evidence record, rather than as a direct measurement of overnight opinion change.
The electoral threshold is one place where small differences can have large consequences. Votes for a list below 3.25% yield no seats, and representation changes how the 120 seats are distributed. For Amcha Israel, the final threshold probabilities are 28.3% in R and 23.3% in D; for the Reservists and Economic list, they are 35.5% and 28.9%. The full party picture is below.
| Party/list | R | D |
|---|---|---|
| Likud | 23 (20–26) | 23 (20–26) |
| Shas | 8 (0, 5–10) | 8 (0, 6–10) |
| United Torah Judaism | 8 (0, 5–10) | 8 (0, 6–10) |
| Otzma Yehudit | 8 (0, 6–10) | 8 (0, 6–10) |
| Religious Zionism–Zehut | 6 (0, 5–9) | 7 (0, 5–9) |
| Amcha Yisrael | 0 (0, 4–7) | 0 (0, 4–7) |
| Haredi Public | 0 (0) | 0 (0) |
| Beyachad | 13 (10–16) | 13 (10–15) |
| Yashar! | 23 (19–26) | 23 (19–25) |
| Yisrael Beiteinu | 8 (0, 5–10) | 7 (0, 5–9) |
| The Democrats | 9 (0, 7–12) | 9 (0, 7–11) |
| Raam | 5 (0, 4–8) | 5 (0, 4–7) |
| Joint List | 9 (0, 6–11) | 9 (0, 6–11) |
| Blue and White | 0 (0) | 0 (0) |
| Reservists–Economic | 0 (0, 4–7) | 0 (0, 4–7) |
The whole-seat headlines form a coherent 120-seat allocation. Bloc expected seats average the allocated outcomes across simulations, so adding the headline numbers can give a different bloc total. Party ranges display zero separately where applicable; the positive interval is the central 80% range conditional on representation. Bloc ranges are the unconditional central 80% intervals.
The denominator checks matter. Six newly admitted tables already express a 100% party-choice base and remain unchanged. Maagar Mochot explicitly reports 8% 'don't know' in its raw voting question; N1 removes that category from the denominator once, retaining generic other-party responses. The rounded source table totals 100.5%, within the existing rounding tolerance, and is not forced to close by changing reported cells. Separately reported hesitation, turnout intentions and recruitment refusals are preserved as distinct evidence.
The incomplete vote percentages in Next Data's disclosure 4254 remain held: the reported party values total 96.9%, with no explicit category explaining the remainder. Its seat evidence remains eligible. Disclosure 4256 concerns border communities and is outside the national population; DRI disclosure 4247 is excluded by the owner's instruction. The nine supplied PDFs have been reviewed, without claiming a fresh comprehensive sweep of the CEC register.
Both approaches use the same 100,000 simulated outcomes. R gives them equal weights; D changes their weights using its existing seat restraint. The public model's formulas are unchanged. The forecast retains its assumptions about untracked lists and the absence of surplus-vote agreements. Earlier issued snapshots, their original articles and their actual simulation counts remain available as the public record; today's information has not been written backwards into them.
Readers can inspect the source observations, model inputs, current input roles, normalization audit, and CSV/JSON documentation. The role table distinguishes current inputs from held or superseded evidence.
For months, the campaign has appeared here as probabilities and ranges. On election day, individual choices will turn that uncertainty into a result. This is the final forecast we can make from the evidence in hand. The next decisive numbers belong to the voters.