Claims under audit

Part of github.com/debedb/elephants · the method, in full →

Thirty claims, drawn at random. The events are real. The sentences assert more than the documents do.

Loading.

This is a finding about sentences. It is not a finding about the classifications the tracker assigns, which we do not adjudicate, nor about its method, which publishes its criteria and marks its judgements as opinion. Thirty rows out of 3,466 cannot tell you whether any of this holds generally. Everything below is the working.

Everything is arguable. The field is even, the ball is round. Everything must be in the open.

— a football saying. We do not know whose.

We tried three. Sepp Herberger, whose 1954 der Ball ist rund is repeated everywhere and sourced nowhere we could reach. Nikolai Ozerov, offered in review, whose documented catchphrases are other lines. Vadim Sinyavsky, offered next, with the person offering him saying plainly that he might be misremembering. The Russian form мяч круглый, поле ровное carries both halves and Russian sources call it simply an old saying, attached to no one. Three names, no primary source, and a sentence that is probably older than all of them.

This is the whole method, failing in public, on its own epigraph. The rule we work by says a half-remembered citation is a lead rather than a problem to be repaired: you do not go and verify who said it, you go and find out whether the artifact exists. We went. It does not. So the honest citation for the sentence this page is built on is that we cannot give you one — and saying that is worth more than a confident name would have been, because a confident name is exactly what we would have printed if we had not looked.

We took thirty claims at random from a tracker of 3,466, went looking for the primary document behind each one, and put the two side by side.

Our position

The epigraph above, unpacked.

Hidden methodology is discounted. Open methodology can be disputed. That is the whole rule, and it runs the opposite way to the usual instinct that concealing how you reached a conclusion protects it. Concealment does not protect a method; it removes its standing. A method nobody can inspect cannot be argued with, and a claim that cannot be argued with cannot be relied on either \u2014 it is not refuted, it is discounted. Publishing a method exposes it to attack, and being attackable is what makes it worth something.

The field is even because nobody here has standing. We claim no authority and hold no office. Nothing on this page is binding, nothing is remanded, and no one owes it a response. What we have is what we can show you \u2014 which is also all anyone else in this argument has, which is what makes the field level. Authority would tilt it.

The ball is round because a well-argued case can lose. Six of our own findings have been overturned by our own checks; one rubric change reversed a verdict we had already published. If the outcome were settled in advance there would be no reason to run the tests, and we would be doing something other than what we say we are doing.

Applied without exception, including to us: our rules are frozen in public before the work and every correction is logged, so they are disputable and we would rather they were disputed. The upstream tracker publishes the criteria behind its classifications, which is more than most, and does not publish its data anywhere version-controlled, so that half of it can only be taken on trust. A dashboard built on that data forecasts institutional collapse before 2028 with no stated method, so there is nothing there to dispute. And a directive decided by classified order has a method nobody outside government can see, which is why we record that we cannot check it rather than pretending its silence is evidence.

On the line itself, corrected. We first wrote that "the ball is round" comes from Sepp Herberger, coach of the West Germany side that won the 1954 World Cup \u2014 der Ball ist rund und das Spiel dauert 90 Minuten. That attribution is widely repeated in secondary sources and we did not reach a primary record of him saying it. It is also probably the wrong lineage for the sentence above, which reads "the field is even, the ball is round" \u2014 both halves of the Russian football commonplace \u043c\u044f\u0447 \u043a\u0440\u0443\u0433\u043b\u044b\u0439, \u043f\u043e\u043b\u0435 \u0440\u043e\u0432\u043d\u043e\u0435, which Russian sources describe as an old saying and attach to no one. A competing attribution to the Soviet commentator Nikolai Ozerov was raised and we could not support it: his documented catchphrases are others. So there are two traditions, we have a primary source for neither, and we have taken a football aphorism and hung a nationality on it that it may not have. Recorded rather than quietly fixed, because it happened in the paragraph announcing this page\u2019s thesis, and because the failure \u2014 repeating a confident attribution from secondary sources \u2014 is the exact one the page exists to catch.

The ledger

Each row opens with the claim and the document, quoted. Read those two and you have the finding; everything after them is us showing our work, and you can ignore it or attack it.

Drawn, not yet adjudicated

In the sample, not yet worked. Listed by index so the draw stays verifiable.

Claims quoted from

The Trump Action Tracker, by Christina Pagel — trumpactiontracker.info, export trump-actions-6-26-26.csv, accessed 2026-06-27, pinned at sha256 6fa721f4…2cc8f2. Quoted verbatim for examination, credited and linked; the dataset is not mirrored here. Her glossary defines every domain and states that the classifications are expressions of opinion, and her about page declares the scope. We read the export before we read those pages, and got things wrong as a result →