The match guide that
updates itself
Every football tournament has the same problem: dozens of matches, limited evenings, and no honest way for a neutral fan to know which games are worth watching. Official fixtures tell you when a match kicks off. They do not tell you whether to care.
MatchRated answers that question automatically. Every match gets a watchability rating built from what neutral fans actually care about: stakes, drama, form and star quality. Ratings refresh hourly through an automated pipeline, and every match card can be shared as a graphic.
It is our own product, built and operated end to end, and it ran unattended through an entire World Cup. Nobody edits it. The system is the editor.
Editorial output with
no editors
A tournament moves hourly. Scores change, group standings shift, knockout paths form, and a rating that was right this morning is wrong by tea time. Keeping up manually would need a newsroom. The entire point of the product was to need nobody.
The engineering challenge was a pipeline that fetches fresh data on a schedule, respects the rate limits of upstream football data APIs (including backing off politely when they push back), recalculates every rating, and publishes the result, with the site staying fast and stable under match-day traffic spikes.
Every feature earned its place
Watchability rating engine
A scoring model that turns fixtures, form, stakes and star quality into one honest number per match. Built for neutral fans deciding what to watch, not for pundits.
Hourly update bot
An automated pipeline that refreshes data every hour, with rate-limit-aware retries so a busy upstream API never becomes a broken site. Runs without supervision.
Shareable match graphics
Every match generates a share-ready PNG card with the rating and the web address on it, so every share is also a small advert.
TV channel guide
Ratings are only useful if you can act on them. Each match shows where to watch it, updated alongside the data.
The Wheel
An interactive picker for the gloriously undecided: spin, get a match, watch it. Product personality matters even in data products.
CI deployment pipeline
The bot commits fresh data and the deployment pipeline ships it automatically. Data changes become live pages with no human in the loop.
A platform that ran a
tournament by itself
MatchRated ran through an entire World Cup without manual intervention: data refreshed hourly, ratings recalculated, graphics generated, deploys shipped. Match-day traffic spikes were absorbed by the static architecture without drama.
This is what we mean by systems that run themselves. The same pattern of automated ingestion, scheduled processing and hands-free publishing applies directly to price lists, availability calendars, market reports, or any business content that changes faster than a human wants to update it.