publish-the-run-that-did-not-work
Report the experiment that failed, preserve the superseded one unchanged, and never re-roll for significance. Selection is the bias, not the number.
Conferida por quien revisó y leyó la tarjeta, no reclamada por el propio archivo. Una tarjeta que el agente destila durante una ejecución que consumió contenido no confiable nace contaminada y queda retenida para revisión antes de ser recuperada.
Cuándo viene a la mente
- the A/B came out negative
- rerunning the benchmark
- the result contradicts what we claimed
- writing up a measurement
Evita y Comprueba es donde suele estar el valor. Haz es la sección que todo el mundo escribe.
Disparador
You ran a comparison to decide whether a change helps, and the result is inconvenient: negative, not significant, or contradicting something already announced.
It does not apply to a run that was simply broken — a crashed harness measured nothing, and publishing it as a result would be its own kind of noise. Fix and rerun, and say that you did.
Haz
- Write down, before running, what outcome would kill the claim. If you cannot name it, the experiment cannot fail, and an experiment that cannot fail is not one.
- Report the effect with an interval, not a point estimate.
+9.8pp [−3.5, +16.7]says something+9.8ppdoes not. - When a later run supersedes an earlier one, keep the earlier one published and unedited, and link the correction. Amending it removes the only evidence that the correction happened.
- If a claim does not survive replication, retract it in the same place it was made, in the same size type.
Evita
Rerunning until it looks better and reporting the best run. That is not a bad measurement, it is a selected one, and selection cannot be detected in the number afterwards by anyone — including you.
Avoid quietly dropping the negative result while keeping the positive one from the same series. Avoid pooling a fresh sample with an old one after seeing that pooling helps; if pooling was not pre-registered as primary, it is secondary and must be labelled that way.
And avoid describing a non-significant delta with words that imply significance: "improved", "lifted", "up from". "Not significant on its own" is the sentence.
Comprueba
Ask what a sceptical reader could conclude from the artefacts you published. If the answer requires trusting that you did not discard anything, the artefacts are incomplete — publish the journal, the per-item outcomes, or the raw pairs.
Then check the claim in your own README against the results section. If one says a thing the other retracts, the retraction is the true one and the claim has to move.
Riesgo
This is slower, and it publishes work that looks like failure to anyone reading a single page out of context. Some of those readers will be evaluating whether to use the thing.
The subtler cost: a project that publishes negatives can be quoted against itself. That is a real price, and it is smaller than the alternative — one discovered overstatement makes every other number you have published worthless.
Cómo usarla
La tarjeta es un dato. Clona el repositorio e impórtala por ruta: lo que llega por la red se trata como contaminado y queda retenido a la espera de aprobación, que es el comportamiento deseable y la razón por la que aquí no hay un instalador de una línea.
git clone https://github.com/brcampidelli/chimera-agent.gitchimera skills-import chimera-agent/skills/publish-the-run-that-did-not-work/SKILL.mdIntegridad
SHA-256 del archivo tal como se publicó. Quien lo importe puede comprobar que lo que recibió es lo que mostró esta página.
a0a59845f9bef586b26fd372445c1e2688fa69882121a0f5e2ef8e4848b1218a