Rewind three Israeli elections. Compare two forecast approaches, follow each bloc’s chance of a majority, and see where the final predictions met reality.
THE MAJORITY PICTURE · 2 Mar
Netanyahu-aligned6.6%
Zionist non-Netanyahu grouping<0.1%
Chance of 61+ seats in each stated scenario
FINAL PREDICTION
Move across a trend graph or drag with a finger to explore each day. You can also focus a graph and use the arrow keys.
Netanyahu-aligned6.6%
Zionist non-Netanyahu grouping<0.1%
BLOC MAJORITY PROBABILITIES · Vote Share Only
What are the chances of reaching 61?
10,000simulated seat outcomes per available snapshot2 Mar · Baseline
The chance that the listed parties win at least 61 seats together. This counts their seats under the model’s assumptions; it does not predict whether they would agree to form a government. We show selected groupings, not every possible combination. Vote and Seat Fusion gives different weights to the same 10,000 simulated election results.
THE DAILY MAJORITY RACE
How the odds moved
Scale 0–40% · move the cursor or drag to choose a date
* About 80% of simulated seat outcomes fall in this range, from the 10th to the 90th percentile.
Available dates · dotted line = 50% · dot = selected date
How parties moving together changes the result: 11.9%–23.6%
These results test different assumptions about how parties’ support rises or falls together. They are separate from the main forecast, not a range of likely outcomes, and the main forecast may fall outside them.
Assumptions move this probability by at least 10 percentage points.
Available dates · dotted line = 50% · dot = selected date
How parties moving together changes the result: <0.1%–2.0%
These results test different assumptions about how parties’ support rises or falls together. They are separate from the main forecast, not a range of likely outcomes, and the main forecast may fall outside them.
Parties in this scenario (3)
Blue and White · Labor-Gesher-Meretz · Yisrael Beiteinu
Explore all six descriptive groups
These groups describe parts of the political landscape; they are not extra majority scenarios. Their ranges count seats from the same simulated elections together, rather than adding each party’s separate range.
Anti-Bibi Right Zionist
6.9 average seats
80% interval: 5–9
Yisrael Beiteinu
Arab / non-Zionist lists
13.1 average seats
80% interval: 11–15
Joint List
Center
35.0 average seats
80% interval: 32–38
Blue and White
Haredi
15.5 average seats
80% interval: 13–18
Shas · United Torah Judaism
Left Zionist
8.8 average seats
80% interval: 7–11
Labor-Gesher-Meretz
Pro-Bibi Right Zionist
40.6 average seats
80% interval: 37–44
Likud · Otzma Yehudit · Yamina
FOLLOW THE CAMPAIGN · Vote Share Only
The road to election day
Select parties to compare. Move across the chart or drag with a finger to choose a day.
2 MarBlue and White35Likud33Joint List13
Line: headline seats · ◆ Actual result. Gaps mean the party or forecast is unavailable. Uncertainty and threshold probabilities appear below.
PARTY FORECASTS · Vote Share Only
The final call. The actual result.
PARTIES FORECAST TO WIN SEATS8
9 parties in this date’s lineup
MEAN ABSOLUTE SEAT ERROR1.33 seats
9 comparable parties
K23 Vote Share Only forecast on 2020-03-02. Difference is forecast minus actual. Ranges cover the 10th–90th percentiles.
Party / uncertainty
SeatsForecast / actual
DifferenceSeats
Blue and White
80% range: 32-38Above threshold: 100.0%
35 / 33
+2
Likud
80% range: 30-36Above threshold: 100.0%
33 / 36
-3
Joint List
80% range: 11-15Above threshold: 100.0%
13 / 15
-2
Labor-Gesher-Meretz
80% range: 7-11Above threshold: >99.9%
9 / 7
+2
Shas
80% range: 6-10Above threshold: >99.9%
8 / 9
-1
United Torah Judaism
80% range: 6-9Above threshold: >99.9%
8 / 7
+1
Yamina
80% range: 5-9Above threshold: 99.8%
7 / 6
+1
Yisrael Beiteinu
80% range: 5-9Above threshold: 99.6%
7 / 7
0
Otzma Yehudit
80% range: 0, 4-9 · if representedAbove threshold: 10.2%
0 / 0
0
Bar = headline forecast · marker = actual. Each available headline totals 120 seats. “0, x–y” means zero seats remains possible; x–y is the 80% range conditional on representation. It is not a continuous interval.
POLLS AND METHODS
The evidence behind this forecast
VOTE-SHARE POLLS17
Polls informing the vote-share forecast
SEAT-POLL RESULTS—
Used only by Vote and Seat Fusion
How are polls counted?
Vote-share polls count distinct surveys used for Vote Share Only. Seat-poll results are the seat-based figures supplied to Vote and Seat Fusion; several results can come from one poll. The groups may overlap, so adding them would not give a reliable number of unique polls. The export does not identify pollsters for each day.
DOWNLOADABLE DATA, VISIBLE LIMITS
Inspect the evidence.
V7 attachment · 183 calendar dates · 181 available snapshots per approach · 13 frozen-horizon parity checks. Vote Share Only is the baseline; Vote and Seat Fusion is an experimental comparison. Exact dependence sensitivity is supplied only for the final Vote Share Only snapshots.