Threshold Simulator

One party dropping below 3.25% can flip the whole map. How close to the line is it really?

We took a real seat poll and simulated thousands of "elections" in which every party drifts randomly within the poll's margin of error. The result shows how much a party hovering right around the electoral threshold β€” and the balance between the blocs β€” rides on sampling luck alone.

Channel 13 News Β· Jul 22, 2026 Β· sample 988, margin Β±3.5%

No bloc is locked in.

Zionist Home sits right on the line β€” a 50% chance of clearing the threshold. If it falls, its votes are redistributed and the balance between the blocs shifts.

Share the simulator

One party on the line changes the whole picture β€” let friends play with the scenarios.

How the Knesset splits on average

The expected split of all 120 seats across every simulation.

Average split across the simulations. 61 seats are needed for a majority.
  • Yashar!22
  • Likud22
  • Beyachad14
  • The Democrats11
  • Yisrael Beiteinu10
  • Shas8
  • United Torah Judaism8
  • Otzma Yehudit7
  • Hadash-Ta'al6
  • Ra'am5
  • Religious Zionism5
  • Zionist Home2

The blocs and the race to a majority

For each bloc: its expected seats, likely range, and the chance of reaching 61 on its own.

Anti-Netanyahu bloc60
expected seats
Likely range: 56–64Majority odds: 42%
Arab parties10
expected seats
Likely range: 6–13Majority odds: 0%
Pro-Netanyahu bloc50
expected seats
Likely range: 46–54Majority odds: 0%

On the bubble

Each party near the 3.3% threshold, and its odds of crossing. Force a party in or out to see how it changes everything.

Religious Zionism90%
Polled at 4%
to cross
Ra'am90%
Polled at 4%
to cross
Zionist Home50%
Polled at 3.2%
to cross
Balad0%
Out in this poll Β· 1.9%
to cross
Blue and White0%
Out in this poll Β· 1%
to cross

Based on 5,000 simulations per calculation.

How it works (and its limits)

  • Each party starts from its polled vote share, and in every run it drifts randomly by the standard error derived from the sample size β€” a small party near the threshold is more sensitive to noise than a large bloc.
  • A party that drops below 3.25% of the vote loses its seats, and they are redistributed among the parties that cleared the bar β€” using the D'Hondt (Bader-Ofer) method, as in a real election.
  • The probabilities are frequencies over thousands of runs, not a forecast. Sampling error is only one source of uncertainty; there are also pollster bias and last-minute shifts.
  • Bloc assignments reflect each party's declared stance as of the poll date, and ignore surplus-vote agreements.

This is a statistical illustration for understanding, not an election forecast. The numbers reconstruct a single public poll β€” don't treat them as a basis for how to vote.