A Knesset seat forecast from a synthetic panel
2,000 synthetic Israelis, and five ways of knowing how they vote. The forecast was frozen on 24 Oct 2026, before election day, and not changed since.
No bloc reaches 61 on its own
In all five forecasts, the Netanyahu bloc gets at most 58 seats and the Change bloc at most 57. Only the Change bloc together with the Arab parties crosses the line: 62 to 68 seats.
Five forecasts at a glance
All five start from the same 2,000 synthetic Israelis and the same seat calculator. They differ only in where each persona's vote comes from.
| Forecast | Netanyahu bloc | Change bloc | Arab parties | Chance of 61 |
|---|---|---|---|---|
| The panel aloneNo survey from the campaign at all | 53 | 57 | 10 | under 1% |
| Panel + opinion survey~1,500 respondents, late August and early October | 54 | 56 | 10 | 1% |
| Panel + latest wave751 respondents, 5–8 October | 58 | 53 | 9 | 6% |
| Media poll average52 polls, mid-September to 22 October | 53 | 53 | 14 | under 1% |
| Poll average without Channel 1440 polls | 52 | 54 | 14 | under 1% |
In detail
Who moved to which party since 2022, from real respondents in a recent opinion survey (two waves), applied to the personas.
Dashed line = 61 seats
| List | Seats | Range | |
|---|---|---|---|
| Yashar! | 23 | 19–26 | |
| Likud | 22 | 19–25 | |
| Together | 15 | 11–17 | |
| The Democrats | 12 | 9–14 | |
| Joint List | 10 | 9–12 | |
| Otzma Yehudit | 7 | 5–10 | |
| Religious Zionism–Zehut | 7 | 5–9 | |
| United Torah Judaism | 7 | 6–9 | |
| Shas | 7 | 5–8 | |
| Yisrael Beiteinu | 6 | 4–8 | |
| Amcha Yisraelon the edge · 71% | 4 | 0–6 | |
| Ra'amon the edge · 14% | 0 | 0–4 | |
| Reservists & Economicon the edge · 9% | 0 | 0–4 |
Range: 90% of simulations. "On the edge": a list whose chance of passing the 3.25% threshold is between 5% and 95%. Lists with at least 1% of the vote are shown.
How it works
- The personas: 2,000 synthetic Israelis built like the adult population (sector, religiosity, age, region), each knowing how they voted in 2022, matching the real results.
- Why an opinion survey: the personas know how they voted in 2022, but not what changed since — mergers, new parties, two years of war. A recent opinion survey shows who moved where, and the panel applies it to an accurate population mix instead of one sample's accidental mix.
- From votes to seats: different turnout in each sector, the 3.25% threshold, surplus-vote agreements and the official allocation method.
What the forecast doesn't know
- The panel alone rests on vote intentions from eighteen months ago: it doesn't know lists formed since, and it undercounts Arab voters.
- Lists near the threshold sometimes rest on a few dozen respondents. One seat either way is close to chance.
- The ranges reflect only what can be measured. If all the polls are wrong together in the same direction, the forecasts built on them will be too.
Synthetic data. An estimate, not an election poll. Sources: public surveys and datasets.