Data Methodology

ElectioMap publishes election results sourced exclusively from official electoral authorities. This page explains exactly where our data comes from, how often it updates, and how we handle errors and corrections.

Written and maintained by Bartłomiej Paruzel, Election Data Analyst · Last reviewed June 2026.

Where our data comes from

Primary sources — official government authorities

For every election we cover, our first and preferred source is the official national electoral commission or equivalent government body. These are the authoritative institutions responsible for conducting and certifying elections. Examples include:

• Colombia — Registraduría Nacional del Estado Civil
• Poland — Państwowa Komisja Wyborcza (PKW)
• Armenia — Central Electoral Commission of Armenia
• Kosovo — Komisioni Qendror i Zgjedhjeve (KQZ)
• Hungary — Nemzeti Választási Iroda (NVI)

We link to the official source on each election page so you can verify the figures independently at any time.

Secondary sources — unofficial data, always cross-verified

In rare cases, an official source may not publish machine-readable data or may be temporarily unavailable. When we cannot use an official source, we may draw on reputable journalistic aggregators, major international wire services, or international election-observation bodies. In every such case, we:

1. Cross-verify the figures against at least two independent non-official sources
2. Clearly label the data as coming from a non-official source on the relevant page
3. Replace the data with the official figures as soon as they become available

We never publish figures from a single unofficial source without independent corroboration.

Live election coverage

Update frequency

During an active election count, our automated system fetches the latest results from the official source and updates the map approximately once per minute. This is the fastest interval that avoids placing excessive load on official government servers.

Some official commissions restrict how often their data may be accessed, or publish batched updates less frequently. In those cases, we fetch data as often as the source permits — typically every 2–5 minutes — and display the time of the last successful update on the results page.

What "live" data means

When the official authority certifies the final result, we update the election page to mark it as final. Preliminary figures are never used in historical archives — only certified final results are preserved.

Source unavailability

If the official source becomes temporarily unavailable, we display the last successfully retrieved figures alongside the exact timestamp of that retrieval. We do not interpolate, estimate or fill gaps with projected numbers. The page clearly shows when data was last refreshed.

Historical election data

Historical results are sourced from the official certified final results published by the relevant electoral authority. Where results are expressed as percentages, we record them to exactly two decimal places as published by the source — we do not round, aggregate or adjust figures. Vote totals and seat counts are recorded exactly as certified.

For older elections where the original official source is no longer available online, we use the best available archival record and note the source clearly on the page.

How live data flows to your screen

• Our server polls the official source API or results page on a ~1 minute schedule
• Raw figures are validated against our data schema and stored with an immutable timestamp
• Your browser polls a lightweight manifest file every minute to detect when new data arrives
• Only when the manifest signals new data does your browser fetch the updated results snapshot
• The map and results table update instantly — no page reload needed

How we classify and label results

Every figure on ElectioMap is labelled so you know exactly what you are looking at. We distinguish clearly between:

• Preliminary / provisional — running totals published by the authority during or shortly after the count, before certification. These can change as more ballots are tallied.
• Official final / certified — the legally binding result confirmed by the electoral authority once counting is complete and any legal challenges are resolved. Only these are kept in the historical archive.

Where an authority reports the share of ballots or precincts already counted, we display that figure too, so you can judge how complete a preliminary result is. Candidate and party names are shown exactly as registered with the electoral authority, with an English translation or faithful transliteration added where the original is not in the Latin alphabet. We do not apply nicknames, colloquial labels, or descriptions that could be read as politically loaded — a principle set out in full in our Editorial Policy.

Multi-round, multi-chamber and regional results

Many elections are not a single number. We model them as the official authority reports them:

• Two-round elections (such as most presidential run-offs) are stored as separate rounds, each with its own figures and, where available, its own map. A later round never overwrites the stored record of an earlier one.
• Multi-chamber legislatures (for example a separate Senate and lower house) are shown chamber by chamber, with seat allocations for each.
• Regional breakdowns are published only when the electoral authority releases official results by region, province or constituency. We map those figures using public administrative boundary data and colour each region by its leading candidate or party. When no official regional breakdown exists, we say so explicitly rather than estimating one.

Seat totals, majority thresholds and turnout are taken directly from the authority's published figures, not recalculated by us, except where a simple, clearly documented arithmetic aggregation is needed (for example summing constituency-level official results into a regional total).

Data precision and formatting

We treat the source figure as authoritative and avoid introducing rounding error. Percentages are recorded to two decimal places exactly as published; we do not pre-round source data and we do not inflate precision the source did not provide. Vote counts and seat numbers are recorded as integers exactly as certified. Numbers are then formatted for display according to each reader's language (for example, thousands separators and decimal marks follow local conventions), but the underlying stored value is always the official one.

Verification process

Accuracy is checked at several points before and after a result is published:

• Live figures are fetched directly from the official source by software, never typed in by hand — this removes transcription error from live data
• Every incoming record is validated against a strict results schema; data that does not conform is rejected rather than published
• Historical figures are checked against the authority's certified result and, where helpful, against reputable independent records before being added to the archive
• Each published page links to its official source so any reader can verify the numbers for themselves

Limitations and transparency

We are open about what our method can and cannot do. Preliminary results reflect only what the authority has released at a given moment and may shift; we therefore timestamp every live update and never present a provisional figure as final. For some elections — particularly in countries where the authority publishes only national totals — a regional map is simply not possible, and we state that on the page instead of filling the gap with guesses. For older elections whose original official source has gone offline, we rely on the best available archival record and identify the source used. We do not assess whether an election was free or fair; where credible observers have raised concerns, we report that context factually and let readers draw their own conclusions.

Authorship and accountability

ElectioMap's data methodology is defined and maintained by its editor, Bartłomiej Paruzel, who is named and contactable. Responsibility for sourcing, accuracy and corrections rests with a real, identifiable person — not an anonymous brand. If you believe a figure is wrong, you can reach the editor directly and expect a substantive response.

Corrections policy

We correct errors as soon as they are identified. Because our live data is fetched directly from official sources, corrections to live figures happen automatically when the source corrects its own data. For historical pages or metadata errors (candidate names, party labels, seat thresholds), corrections are applied manually and, where material, noted on the affected page.

If you spot an error, please contact us — we take data accuracy seriously and will investigate all reports promptly.

Questions about our data?

If you have questions about sources for a specific election, methodology details, or would like to report a discrepancy, please visit our Contact page. You may also find our Editorial Policy useful for understanding how we make coverage decisions.