Election polls, voting trends, win probabilities, pollster ratings, and real-time election results. All in one place.
Joined April 2022
- Tweets254
- Following8
- Followers13.1K
- Likes347
PollsterGraph retweeted
I find it troubling that after being the most precise pollster 4 election cycles in a row, AtlasIntel still gets quite low coverage in US media, in particular on TV. To address this issue, we’ll be developing our own alternative channels, with a focus on @PollsterGraph.
Our forecast gave Democrats an 88% chance of winning the Senate before election day, and AP’s calls now give them 54 seats. Our favorite won 63 of the 67 Senate and governor races called so far.
pollstergraph.com/results/us…
Election Night 2026 is underway!
PollsterGraph is providing live coverage of the US midterms, with real-time results, Senate and House races, gubernatorial contests, and our probabilistic election forecasts.
pollstergraph.com
Israel, partial count (31% of the electorate, 3,906 ballot boxes, read 4:51 am Israel time): @PollsterGraph projection gives Likud 27 seats (80% range 25–28) and Netanyahu's bloc 56 (54–58); 61 or more in <1% of simulations. pollstergraph.com/results/il…
🇮🇱 ISRAEL ELECTION UPDATE
With 6.5% of the electorate counted, PollsterGraph projects:
• Likud: 29 seats
• Yashar: 21 seats
Likud is projected to finish first in over 99% of simulations, but Netanyahu's bloc has just a 14% chance of reaching the 61-seat majority.
Who wins the most Knesset seats? Here, PollsterGraph and Polymarket answered essentially the same question, but reached opposite conclusions.
PollsterGraph: Likud 70.5%, Yashar 28.5%.
Polymarket: Likud 32.5%, Yashar 67.5%.
Full forecast: pollstergraph.com
Who becomes prime minister?
Our model, conditional on this Knesset forming a government: Netanyahu 41%, Eisenkot 33%.
Polymarket, predicting the first PM sworn in by the end of 2027, including after a possible new election: Eisenkot 58%, Netanyahu 25%.
How did the model perform historically?
Tested on seven Israeli elections from 2013 to 2022, the actual seat count for Netanyahu's camp fell within the model's 80% prediction range in 6 of 7 elections.
Each election was forecast using a model fitted on the other six.
Forming a government could be even harder.
Accounting for parties' publicly stated coalition red lines, a government formed in just 27% of simulations.
When those red lines were assumed to be three times more likely to break, that figure rose to 73%.
Forming a government could be even harder.
Accounting for parties' publicly stated coalition red lines, a government formed in just 27% of simulations.
When those red lines were assumed to be three times more likely to break, that figure rose to 73%.
Winning the most seats is not the same as securing a majority.
Netanyahu's bloc averaged 53.5 seats, reaching the 61-seat threshold in just 8% of simulations.
The opposition without Arab parties averaged 51.7 seats, reaching 61 in 5%, or 24% in a coalition with Ra'am.
Likud was projected to win 24.4 seats on average, followed by Yashar with 21.9 and Together with 12.4.
Four smaller parties faced uncertainty around Israel's 3.25% electoral threshold, where a fraction of a percentage point can determine whether a party enters parliament.
🇮🇱 Israel has voted. What did our model predict before the election?
Using 84 polls from 9 firms and 40,000 simulations, PollsterGraph gave Netanyahu's Likud a 70% chance of winning the most seats, against 29% for Eisenkot's Yashar.
A diferença final foi de 3,6 pontos, contra 3,4 projetados inicialmente.
Como funciona a projeção? O modelo combina pesquisas eleitorais, resultados históricos e votos apurados em cada estado para estimar os votos que ainda faltam ser contabilizados.
Às 19h14, com apenas 1,7% dos votos apurados, a projeção já indicava Lula 3,4 pontos à frente.
Às 20h09, com 31,2% dos votos contabilizados, a probabilidade de vitória de Lula já superava 95%.
A liderança na apuração oficial só veio às 20h24.
Na véspera da eleição, a média das pesquisas apontava Flávio Bolsonaro 1,7 ponto à frente de Lula, com 65% de chance de vitória.
Mas, com o início da apuração, os primeiros resultados começaram a indicar um cenário diferente.
O PollsterGraph antecipou a vitória de Lula 1h10 antes de ele assumir a liderança na apuração oficial.
Com apenas 1,7% dos votos apurados, já era possível projetar o vencedor da eleição presidencial.
Veja como foi.
PollsterGraph estima 95% de chance de vitória para Lula, contra 5% para Flávio Bolsonaro.
Com 33,2% dos votos esperados apurados, o modelo projeta Lula com 51,4% dos votos válidos.
Veja como mudou a margem de votos entre 2022 e 2026 nas mesmas seções eleitorais:
pollstergraph.com/analises/b…
