Where our data comes from

Transport for London

Live line status comes from Transport for London’s Unified API, which is published as open data. It covers the eleven Underground lines, the Elizabeth line, the DLR and the London Overground. Powered by TfL Open Data.

We use it as TfL publishes it. Where this site quotes a reason for a delay, those are TfL’s own words rather than our summary of them, because a paraphrased cause is a cause we cannot stand behind.

What we keep, and why

TfL’s API answers one question: what is happening now. It holds no history, and there is no endpoint for what a line was like last week. That absence is the reason this site exists — the question people actually have is not whether the Central line is delayed this morning but whether it is always like this, and nothing on the internet currently answers it.

So we ask every few minutes and write down the answer whenever it changes. A line running normally for a week is one record, not thousands. A spell of disruption is a record with a start and an end, which is what makes its length measured rather than estimated.

What the reliability figures count

• Only unplanned disruption. TfL marks planned engineering work separately, and it is excluded everywhere on this site. A line closed every Sunday for upgrades is not unreliable, and counting scheduled possessions as failures would make the table meaningless.
• Only statuses that were actually in force. A closure announced for next weekend carries a severity now but is not today’s disruption, so it is not counted until its window opens.
• The worst status per line. TfL returns a separate entry for each affected route, so one disrupted line can arrive four or five times. We take the most severe.
• Measured duration. A spell runs from when a line drops below Good Service to when it returns, rather than being rounded to hours or days.

What the figures cannot tell you

They describe what TfL reported, not what passengers experienced. A line reported as running a good service can still be crowded, hot or slow, and none of that appears in a status feed.

They also start when we did. A line that had a terrible year before we began recording looks clean here, and the record only becomes genuinely useful as it lengthens.

And a status feed is a report, not a measurement. If TfL is slow to update a line, we record the delay in the reporting rather than the delay on the railway.

Using these numbers

You are welcome to cite them, with attribution to London Travel News and a note that they are our record rather than an official TfL statistic. If you quote a specific figure, give the date you took it — the totals move as disruptions end.

London Travel News is not affiliated with Transport for London or any operator. Where our figures and an operator’s own published statistics disagree, the operator’s are the ones to cite for official purposes.

Other sources

Reporting on this site also draws on published journalism from the outlets we credit on each article, and on official announcements from the Department for Transport and London’s boroughs. Every story links to what it was written from.