Rating Dutch pollsters: accuracy, the lead and openness
Abstract. Each Dutch pollster is rated out of 100 on accuracy, its error on the gap between the two largest parties, and openness, ranked against Dutch final polls since the rescale of 13 October 2026. Ipsos I&O and Verian, which publish reports with vote shares, samples and weighting notes, grade B; Peil.nl, which publishes seats only, grades C.
Summary
We rate each Dutch pollster out of 100 on the same scale as the British ratings (Rating the British pollsters: accuracy, the Labour–Conservative lead and openness): 50 points for the accuracy of its final Tweede Kamer polls since 2006, 20 for its error on the gap between the two leading parties, and 30 for how much of its working it publishes. The record is pulled towards a deliberately poor pollster, so that a pollster with one good election cannot score well by luck. The rating is shown for readers; it does not weight the polls.
The three pollsters that publish today are close on accuracy and differ on openness. Verian and Ipsos I&O publish reports with vote shares, margins, sample sizes and weighting notes, and are graded B. Peil.nl, whose final polls have been as accurate as anyone's, publishes seats only, with little about its method in its open posts, and is graded C. Dutch final polls miss by less than those of most countries on the site, so the "poor pollster" the ratings pull towards is not very poor, and Dutch grades read higher than Italian or Polish ones. They cannot be compared across countries.
What has changed since this paper was written
Updated 9 November 2026. The rest of the paper is as first published, except where a sentence below says it was corrected.
Pollster grades were rescaled on 13 October 2026. The first scale scored accuracy between fixed end points (a miss of 0.75 points or less earned full marks, 2.75 or more nothing) and pulled every short record towards a poor pollster, one that three in four final polls beat. Almost every pollster ended up with a D or an E, which told readers little. The rating now works like this:
- Accuracy (50 points) is ranked against the country's own checked final polls. If a share \(r\) of them missed by less than the pollster's average miss, it scores \(50\,(1-r)^{0.75}\): the typical (median) final poll earns about 30 and one in the best tenth about 46. Ranking within the country means a pollster is not marked down because its country is harder to poll.
- The lead (20 points) is scored the same way on the miss on the gap between the two leaders, \(20\,(1-r)^{0.75}\).
- A short record is pulled towards the typical (median) final poll, as if the pollster had two more elections there, not towards a poor one.
- Openness (30 points) is scored out of the items checked so far; an item not yet checked counts neither for nor against.
- Every pollster with a final poll on record gets a grade. One whose publications we have not reviewed is graded on its record alone (accuracy and lead scaled to 100) and marked "record only".
The grade boundaries are unchanged: A from 75, B 60–74, C 45–59, D 30–44, E below 30. Across the active pollsters of all ten countries the grades went from mostly D and E to A 8, B 68, C 76, D 31 and E 5. Where the sections below give the old end points, the poor-pollster prior or grades from before 13 October, they describe the first version of the scale; the live tables show today's.
In the Netherlands it gives B to Ipsos I&O (64) and Verian (62), and C to Peil.nl (48). These are the figures of 9 November 2026; the table on the pollsters page is current.
What the rating is for
The rating answers a narrow question: on the evidence we can check, how much should a reader trust this pollster's published figure, and how easily can they see how it was made? It does not say which pollster is right today, and it does not feed the forecast. The Dutch average gives every pollster's latest poll within 21 days of the newest the same weight (Reading Dutch polls: the correction and its test). Peil.nl, which publishes weekly, counts no more than Verian, which publishes monthly.
The evidence: final Tweede Kamer polls, 2006–2025
The accuracy parts of the rating use the 26 final polls in our record for the seven elections from 2006 to 2025, each taken the day before the vote (Measuring Dutch polling error, 2006–2025). Polls that published only seats are scored on seats ÷ 150 (Polls that publish only seats: deriving shares from seat headlines).
Mergers are handled as the pollsters handle them. Ipsos I&O, formed when Ipsos and I&O Research merged, has its own record from 2025; the two older firms keep theirs. Verian carries the record of NIPO, TNS NIPO and Kantar Public, which are the same Dutch team under earlier names.
For pollster \(j\) at election \(t\), the miss \(m_{j,t}\) is its mean absolute error over the parties measured that election, and the lead error \(\ell_{j,t}\) its error on the gap between the two parties that finished first and second.
Scoring, and a poor Dutch pollster
The scales are the British ones:
where the rated miss and lead error are pulled towards a poor pollster as if it had two more elections at the poor pollster's level (Efron and Morris, 1975):
The poor pollster is the 75th percentile of all Dutch final polls:
A poor Dutch final poll misses by less than a British one, so a pollster with no record starts with about 35 accuracy points and a few lead points, before openness. That is why Dutch grades read high next to Italian or Polish ones, and why the order within the Netherlands says more than the grades.
A worked example
Peil.nl has final polls at six elections, with an average miss of 1.06 points per party. Its rated miss is
which earns \(50 \times (2.75 - 1.14)/2 \approx 40\) accuracy points, the joint highest with Ipsos. With 12 lead points and 4 for openness, its score is 56: a C. Verian, with a larger average miss (1.22 at five elections) but 20 points for openness, scores 69: a B.
Openness
The openness items are the same in every country:
| Item | Points | Why it counts |
|---|---|---|
| Raw or unweighted figures | 8 | Shows what respondents said before the pollster's own adjustments |
| Full tables published | 8 | Sub-groups and weighting can be checked |
| Method note | 4 | How the sample was drawn and weighted |
| Sample size | 3 | The basics for judging a poll |
| Undecided share | 3 | Shows how firm the figures are |
| Fieldwork dates | 2 | When the poll was taken |
| Member of the industry body | 2 | Signs up to its standards |
Table: Openness items. A "partial" answer earns a third of the points; an item we could not check earns none.
What we found:
- Verian's monthly report for EenVandaag is a PDF with seats and vote shares (this month, last month and 2025), the sample, the fieldwork dates, the weighting variables, the share of floating voters and the expected turnout. Uncertain voters count for less in its figures, and it says so. Its reports allow reuse with the source named.
- Ipsos I&O's report is a PDF with a seat table, vote shares with 95% margins, the total sample and the number with a party preference, the fieldwork dates and the weighting (including the 2025 vote), and a table of how 2025 voters would vote now with a "don't know / won't vote" row.
- Peil.nl's open posts give a seat table with the change since 2025 and a trend line, the date, and sometimes the share of each party's 2025 voters who stay with it. They give no vote shares, no sample size and no weighting note. Most posts are for subscribers, and we did not read those.
- Ipsos (before the merger) and I&O Research published seats with vote shares, samples and dates; we checked one release of each.
No Dutch pollster publishes unweighted figures, so that item scores nothing for anyone. We did not check membership of the MOA, the Dutch market-research association, so that item also scores nothing.
Grades and the current table
| Pollster | Grade | Score | Accuracy | Lead | Openness | Elections |
|---|---|---|---|---|---|---|
| Ipsos I&O | B | 64 | 30 | 15 | 19 | 1 |
| Verian | B | 62 | 29 | 11 | 21 | 6 |
| Ipsos | C | 58 | 32 | 12 | 14 | 7 |
| I&O Research | C | 58 | 27 | 10 | 20 | 3 |
| Peil.nl | C | 48 | 32 | 12 | 5 | 6 |
| De Stemming | C | 47 | 22 | 11 | – | 1 |
| EenVandaag Opiniepanel | C | 47 | 22 | 11 | – | 1 |
| LISS panel | C | 47 | 23 | 10 | – | 1 |
The accuracy column is close for the pollsters with a long record, between 37 and 40 points. The difference in grades comes from openness: Verian and Ipsos I&O publish reports that can be checked; Peil.nl publishes a seat table. The pollsters of 2017 with a single final poll (De Stemming, the EenVandaag panel and the LISS panel) no longer publish and were not reviewed for openness, so they have no grade.
Each pollster's lean now
Beside the rating the site shows each pollster's latest poll within the 21-day window minus the average. With three pollsters, each lean moves the average by a third of itself. In the latest polls Peil.nl has had FvD several seats higher than Ipsos I&O and Verian, and the CDA lower. A lean is not an error: nobody knows yet who is right. Each pollster's page shows its lean now beside its errors at past elections.
What the rating leaves out
As elsewhere, the rating leaves out how often a pollster polls, who commissions it, and mode (all three use online panels). It also leaves out the exit poll, which Ipsos I&O runs for NOS and RTL, and provincial and municipal elections, which are polled rarely.
Limitations
Three pollsters. The ratings compare three active pollsters; a new entrant would start near the poor pollster.
Seven elections. A single late mover can decide a pollster's record at an election, and every pollster missed the same late movers.
Seats-derived shares. Peil.nl and most older final polls are scored on seats ÷ 150, which adds about a quarter of a point of rounding to their miss.
Openness from open pages. Peil.nl's subscriber posts may publish more than its open ones.
The first real test of whether any rating predicts accuracy in the Netherlands will be the next Tweede Kamer election.
References
Efron, B. and Morris, C. (1975). Data analysis using Stein's estimator and its generalizations. Journal of the American Statistical Association, 70(350), 311–319.
How to cite
Polling Forecast (2026). Rating Dutch pollsters: accuracy, the lead and openness. Retrieved 21 November 2026 from https://www.pollingforecast.com/research/netherlands-pollsters. Live figures change as new polls arrive; the date above is when the text was last revised.


