Lootura

Lootura on your home screen

Theme

Loots

A way to see what is making the news.

More coverage, more Loots

Loots are points that reflect recent media attention. The more distinct publications cover a story, the more it can earn. Repeating the same headline contributes less.

Stories lose momentum over time

Without new articles, Loots fall. This lets today’s events gradually make room for the next stories.

From a story to its subjects

Each link contributes Loots to its story. When it relates to an entity — a person, place or organization — it also helps that entity stand out. Entities at the top of the front page need a story covered by at least three distinct publications.

How to read the numbers

“24 LOOTS” is a story or entity’s score on a scale from 0 to 100. Beside each article, you can see its contribution. The ↑ and ↓ arrows show whether the score has risen or fallen since roughly six hours earlier. They only appear when comparable data exists and the score has changed.

Loots measure coverage, not a story’s accuracy or quality. Reader clicks do not increase the score.

The algorithm, step by step

The score is calculated from links and their publication dates. These are Lootura’s actual rules, without manual importance ratings.

See the formula, weights and an example

1. Each link’s contribution

Contribution = base × publication diversity × repetition × headline × breadth × age

Base
16 points per link before applying the factors.
Distinct publications
Ordered by their earliest recorded article: the first has weight 1, the second 1/√2, the third 1/√3… Each new publication adds less than the previous one.
Links from the same publication
The first has weight 1. Further links have weights 0.2, 0.1, 0.05… Many links from one publication do not count as many publications.
Headline
The first recorded article has weight 1.25. Subsequent distinct headlines have weight 1.15 and repeated ones 1. Case, accents and punctuation are ignored. This does not establish who broke the story.
Breadth
Multiply by 1 + min(0.2, 0.04 × log₂(number of publications)). This provides a bonus of up to 20% for diverse coverage.
Age
Multiply by 2^(−hours/8). A link retains half its weight after 8 hours and a quarter after 16.

Only links with valid titles and dates from the last 24 hours count. Each URL counts once. Sections of one publication are not separate publications. Links are first grouped by event: sharing a city or person is not enough.

2. The story’s Loots

Story Loots = min(100, round(sum of contributions))

A story needs at least 2 publications to score in Headlines and 3 to enter Front page. Displayed contributions are scaled to the total and allocated in tenths: first rounded down, then remaining tenths go to the largest remainders. They therefore add up to the displayed total.

3. The entity’s Loots

Up to 8 related events are ranked by Loots. The first contributes its full score, the second its score divided by √2, the third by √3… A bonus of up to 8 points is added, fading with the age of the newest link.

Entity Loots = min(100, round(weighted event scores + 8 × freshness of newest link))

The total is allocated again in tenths across links and the bonus. A link’s contribution to an entity can differ from its contribution to a story. Only entities with an event covered by at least three distinct publications can appear at the top of the front page.

An example using the real calculation

Fictional example: three publications, one link each, three distinct headlines about the same event, all freshly published. These are not real news data.

Publication 121.1 LOOTS

Publication 213.7 LOOTS

Publication 311.2 LOOTS

The story scores 46 Loots. If it is an entity’s only event, that entity scores 54: 46 from coverage and a 8 bonus. Without new links, after 8 hours the story scores 23 and the entity 27.

The scale is fixed, not rescaled to the day’s leader. When current coverage is insufficient, a labelled archive can use up to 72 hours, retaining age decay. Arrows compare real samples from six to eight hours earlier using the same algorithm version; without comparable data or a change, no arrow appears.

Calculation version: links-v2