ELSEIF
Your brief EB
183 stories from 71 feeds 32 clusters Refreshed 8 minutes ago next pull 13:20

METHOD

How ranking works

Every story on elseif carries a signal score between 0 and 1000. The score is a weighted sum of five named components, and each one is recorded on the story so you can see exactly why something ranked where it did.

The components

Recency

How recently it was published, decaying on a fixed half-life rather than falling off a cliff.

40%
Cross-feed agreement

How many independent feeds carried the same story. Agreement reached separately is the strongest signal we have.

22%
Source authority

Editorial trust in the reporting feed. Primary engineering write-ups outrank rewrite desks.

20%
Early velocity

Discussion the story drew at its origin, not on elseif. We never rank on our own click data.

12%
Topic clarity

How confidently the story classified. A story we cannot place is unlikely to be what you came for.

6%

What we deliberately do not use

  • Payment. No source can buy a position. There is no promoted placement anywhere on this site.
  • Time on page. Ranking never reads how long you stayed. Optimising for that produces content designed to detain you.
  • Clicks on elseif. Using our own click data would create a feedback loop where whatever ranked first keeps ranking first.
  • Headline sentiment. Anger travels further than accuracy. We do not reward it.

Recency decay

Freshness decays on a half-life of 14 hours: a story is worth half as much 14 hours after publication, a quarter after twice that. This is why a well-corroborated story from this morning can outrank a single-source story from ten minutes ago.

Cross-feed agreement

The component that most distinguishes elseif is corroboration. When 5 or more independent feeds carry the same story, it earns full credit for that component. One feed earns none. This is a deliberate bet that independent editorial judgement, arrived at separately, is the best available signal that something matters.

It also has a failure mode worth stating: a story that only one outlet has the sources to break will rank lower than it deserves until others follow. We accept that trade because the alternative rewards volume.

Where the summaries come from

The summary and the three key points on each story are written by elseif from the cluster, not copied from any source. Generated text is checked against the source material before publication, and any assertion that cannot be traced back is rejected and replaced with a plain factual note. Headlines always link to the outlet that wrote them.