Statistical Glitch in Israeli Poll Highlights Limits of Election Forecasts and Reveals New Party Voter Profile

A recent statistical error in an Israeli election poll involving respondent number 148711 caused a significant shift in seat projections between two major parties, illustrating the inherent volatility and noise in polling data. The partial completion of this respondent's questionnaire initially shifted one mandate from Likud to Religious Zionism, but once fully completed, the effect reversed, underscoring the sensitivity of seat allocation algorithms to individual responses. This incident serves as a reminder that weekly fluctuations in poll results often reflect statistical noise rather than substantive changes in voter sentiment.

The poll also examined the emerging political party led by Gilad Erdan and Yuli Edelstein, which has yet to cross the electoral threshold but shows potential support from about 10% of respondents considering their vote. This party's potential electorate is distinct from both smaller centrist parties and the current right-wing government base, with over half identifying as moderate or center-right, and only a small fraction as far-right.

Key issues defining this group's profile include a preference for judicial reform only with broad consensus, contrasting with the more uncompromising stances of Likud supporters and the opposition party Yisrael Beiteinu. Additionally, 59% support mandatory enlistment for Haredim with sanctions for non-compliance, a position far more stringent than that of Likud voters. Notably, 40% of this group's potential voters favor a broad government coalition excluding both Haredim and Arab parties, diverging from the current right-wing coalition.

On the contentious issue of the West Bank, this electorate is almost evenly split between those favoring expanded control and settlement growth (42%) and those advocating full separation (41%), a unique polarization not seen in other party supporters. This suggests that the Erdan-Edelstein party's base resembles a coalition of diverse views rather than a unified ideological bloc.

Overall, the poll findings highlight the challenges of interpreting election surveys, the importance of large sample sizes to mitigate statistical noise, and the nuanced voter landscape emerging around new political formations ahead of the 2026 elections.