How OSINT analysts verify claims, merge duplicate reports, and count independent sources to maintain accuracy on live conflict maps.
When an airstrike hits a building in a city, the first video arrives on Telegram within minutes. By the afternoon, the same footage has been reposted dozens of times across channels, social media accounts, news outlets, and forums. To someone monitoring a live conflict map in real time, this creates an immediate challenge: is each post a separate incident, or copies of the same event? Answering that question correctly is essential. Conflating duplicates distorts the scale of violence; missing genuine separate incidents does the opposite.
The duplication problem intensifies because incentives reward speed over accuracy. Telegram channels seeking subscribers post fast. News outlets republish trending footage. Activists amplify reports to draw attention. No single post carries metadata that proves when and where it originated. A video filename offers no reliable timestamp. A geolocation claim in a caption might be wrong. The analyst's job is to reconstruct what actually happened from this noisy, redundant stream.
The first step is asking: is this the same event or a different one? The answer depends on geography, timing, and the visible damage pattern. If five channels post video of a crater and destroyed building, and the crater, rubble, and surrounding buildings all match when compared frame by frame, they are almost certainly the same incident. If the location differs by several kilometres, or the building types and damage look different, they are likely separate strikes.
Experienced analysts use reverse-image search, geolocation tools, and reference imagery to anchor events in space and time. A strike on Monday morning in one neighbourhood is a distinct event from a strike in a different neighbourhood on Tuesday evening, even if the damage looks superficially similar. When timing is uncertain—a video posted Thursday might show damage from Tuesday—analysts cross-reference with local accounts, infrastructure reports, or witness testimony to establish a plausible window. The goal is not perfect certainty, which is rarely achievable, but reasonable confidence that two reports describe the same explosion or two different ones.
Once duplicates are merged, the next task is counting how many truly independent sources confirm the incident. This matters because source diversity strengthens credibility. An airstrike confirmed by a local resident's video, a journalist's on-the-ground account, and an official statement from a regional authority represents three independent streams of information. The same incident reported only through a single Telegram channel, then reposted twenty times, is still just one source—the original post.
Independence matters more than volume. A major news outlet that cites an eyewitness video is building on that video; it is not an independent source in itself. A military statement that references social media reports is repeating claims, not independently verifying them. True independence means each source gathered the information through its own means: a witness was there, a journalist conducted interviews, a satellite image was captured and analysed, an official statement came from a decision-making body. Counting sources correctly prevents the common error of treating repetition as validation.
Live conflict mapping adds time pressure. An incident might be reported by a local channel at 3 p.m., by a national news agency at 4 p.m., by international media at 5 p.m., and by government statements at 6 p.m. An analyst monitoring in real time must decide quickly whether to flag each report separately or wait for the full picture to emerge. Waiting ensures accuracy but sacrifices timeliness; publishing immediately risks amplifying unverified claims or redundant reports.
Most serious maps adopt a hybrid approach. They publish incidents when at least one credible source confirms them, but they mark reports as 'unconfirmed' or 'single-source' until additional independent verification arrives. As more sources report the same incident, the confidence rating updates upward. A report initially posted with one source might rise to 'confirmed' once a second source appears. Analysts also distinguish between claims about what happened—a building was hit—and claims about responsibility, casualties, or cause, which require more stringent verification. A building strike might be established through video; responsibility for the strike is far harder to confirm and often remains uncertain.
Duplication is not the only challenge. Sometimes independent sources describe the same incident differently. One source claims 'three buildings destroyed'; another says 'two'. One says 'civilian area'; another identifies 'military installation'. These conflicts often reflect genuine uncertainty rather than lying. Eyewitnesses have limited perspective. Damage assessment from a single photo is imprecise. Different sources may interpret the same evidence differently.
Responsible analysts note these discrepancies rather than averaging them or picking the most dramatic claim. A map might record: 'Local reports describe three buildings damaged; satellite imagery confirms at least two structures with visible damage.' This honesty about uncertainty is harder to publish than a single confident number, but it is more useful to readers. It signals what is known, what is disputed, and what remains unclear. Over time, if additional sources arrive, the picture often clarifies. If it does not, the persistent disagreement itself becomes informative.
Not all sources are equally reliable. A Telegram post from an unknown account claiming to be an eyewitness is a source, but a weaker one than a video geolocated to the exact address or a journalist who interviewed multiple witnesses. A single blurry photo is a source, but less useful than multiple angles of clear damage. Serious analysts develop informal hierarchies of credibility based on track records, transparency, and verifiability. A channel that has been consistently accurate and discloses its methods earns more weight than an anonymous account with a history of errors or sensationalism.
This hierarchy is never rigid. A source can be reliable on some topics and unreliable on others. A Ukrainian local account is valuable for reporting events in its own city but less useful for claims about distant events. A Russian military source has incentives to minimize losses but might accurately report operational moves. Analysts must assess credibility in context rather than applying blanket trust or distrust to any organization.
When you read a live conflict map, ask what you are seeing: a single report, multiple independent confirmations, or a merged cluster of duplicates? Look for source attribution. Does the map name its sources, and are those sources accessible for your own review? A map that says 'three sources confirm this incident' is more trustworthy than one that simply asserts facts without evidence. Be wary of round numbers and over-confident claims; conflict reporting is genuinely uncertain, and honest mapmakers say so. If an incident is marked 'unconfirmed' or 'single-source', it deserves skepticism until more evidence arrives—but it is still worth noting, because first reports are often later confirmed.
Duplication and source-counting are unglamorous work. They do not produce headlines. But they determine whether a map shows what actually happened or an amplified, distorted version of events. The best conflict maps are transparent about this process, showing readers both the raw reports and the analysis that shaped them.