How to read this website
A short guide to the numbers and controls. For how the calculation works, see the methodology.
What is this project for?
To 120 combines polling evidence to show party forecasts, bloc majority chances and the uncertainty around them. Israel’s parliament, the Knesset, has 120 seats; 61 is a majority. A chance of reaching 61 seats is not a chance of forming a government.
How do I explore another date?
Move the mouse across a graph without clicking. On a phone, drag sideways; vertical movement scrolls the page. All forecast figures update together. Leaving the graph keeps the selected date. Use Return to latest to reset it, or open the optional exact-date selector. With a keyboard, focus the graph and use the arrows, Home and End.
How do I inspect a party?
Open Parties, choose a list, then choose seats, vote share or chance of entry. Changing the measure, party or approach keeps the selected date. Share this view copies a link to those choices; switching language preserves the same view.
What do the different numbers mean?
Each measure answers a different question:
- Vote share
- Estimated percentage of valid votes for a list; this is support, not its chance of entry.
- Headline seats
- The supplied allocation of 120 seats using the average vote shares across simulations, not necessarily the most likely individual result.
- Average seats
- Seats averaged across simulated elections. This can differ from headline seats.
- Chance of entry
- The proportion of simulated results in which the list clears the electoral threshold.
- Supplied seat range
- Positive party seat ranges are conditional on entering parliament; zero is shown separately where applicable. An 80% model range is not guaranteed coverage of the real election result.
Why don’t bloc probabilities add to 100%?
The composition chart divides the 120 seats into separate groups. Majority cards also ask overlapping seat-count questions. Their probabilities do not need to add to 100%, and adding parties’ seats does not assume they will agree to govern together.
How can I check the evidence?
Browse polls, retained dates and their roles in the calculation on the evidence page. Download the full CSV and JSON files for deeper analysis. Missing values are unknown, not zero; not every recorded poll contributes to every party. Historical reconstructions are identified separately from forecasts published at the time.