1st place. Rafał Trzaskowski – 50.3% 

2nd place. Karol Nawrocki – 49.7%

Czy podoba ci się sondażowy wynik wyborów?

Nie
56.38% 168
Tak
42.28% 126
Jest mi obojętne
1.34% 4

What is an exit poll and how does it work?

An exit poll is a survey conducted on election day directly outside polling stations. Pollsters ask every nth voter selected at random—usually every 10th voter—whom they voted for, and respondents independently place the completed, anonymous ballot in a special ballot box. The aim is to quickly estimate the percentage support for candidates or committees and turnout—the results are known just minutes after polling stations close, long before the National Electoral Commission (PKW) announces its data.

Why is an exit poll so valuable?

Forecast + context—besides voting results, the survey provides extensive information about the structure of electorates (gender, age, education and place of residence), voter shifts between previous and current elections, and turnout in individual groups. PKW does not publish this information. (Ipsos)
Speed and accuracy—with a properly designed sample, the statistical error averages ±2 percentage points, which makes it possible to reproduce the result very precisely on election night.

Key differences between an exit poll and a pre-election poll

Timing of data collection—preference polls are conducted weeks or days before voting, when some respondents are still undecided. An exit poll is conducted after the vote has been cast, when the decision is final. Selection of respondents—an exit poll includes only people who actually voted; a traditional poll includes adult citizens regardless of whether they will go to the polls.

How are representativeness and reliability ensured?

Selecting polling stations—Ipsos selects approximately 800–900 precinct election commissions from across Poland using a stratified proportional method (voivodeship × size of locality), excluding closed and foreign precincts.

Pollsters’ work—at each selected polling station, a pair or trio of pollsters is on duty, inviting respondents according to an established pattern (e.g. every 10th voter).

Collecting and transmitting data—paper questionnaires are immediately coded in a mobile application, and the records flow in real time to a central system, where a team of several dozen analysts checks the quality and calculates preliminary results.

Exit poll, late poll and margin of error

A late poll is an update of the exit poll using official protocols from the same commissions—it is even more accurate because it eliminates the problem of refusals to participate and sampling. Statistical error is unavoidable because Ipsos relies on a sample rather than a complete count of voters. When differences are small (e.g. a committee with 5.1% support against a 5% threshold), the ±2 percentage-point margin must be taken into account and interpretations should be cautious. Refusals to participate, weather conditions or meticulous adherence to the procedure may affect accuracy. In a very close contest, an exit poll tends to show a trend rather than a certain outcome—the final result is always confirmed by PKW after all votes have been counted.