Guides
Longer pieces on how the data behind this site actually behaves — written because most of it is either undocumented, documented misleadingly, or buried in a specification nobody reads for pleasure.
Britain publishes a remarkable amount of open transport data. Very little of it comes with an explanation, and some of what does is wrong in ways that only show up once you have loaded the real files and compared them against the railway. These guides are what we learned doing that.
They are deliberately longer and more specific than the FAQ, which answers questions in a paragraph, and than how it works, which describes each mechanism once. Where a guide states a figure, it is one we measured against a real feed rather than one taken from a specification.
Rail
Why a train on the map is never exactly where the train is
Live train positions come from signalling berths, not GPS. What that means in practice: why the dot jumps, why half of all reported movements do not move it at all, and why some lines have no dots whatsoever.
How the railway names its own track
An ELR is the four-character code the industry uses to identify a stretch of railway. What they mean, why MLN1 is not the Midland Main Line, and why the West Coast Main Line is sixteen different codes.
How a delay is measured, and why every source disagrees
Three feeds report the same late train differently, and the published punctuality figure measures something else again. Where each number comes from, and which one answers your question.
Headcodes, UIDs, and why the same code runs twice
A headcode identifies a train to a signaller, not to you: the same four characters run in several places at once. What each identifier is for, and how to search for the train you actually mean.
One station, five different codes
The industry has at least five ways to identify the same station, because five different systems needed one. What each is for, and why searching a station name returns signals and junctions too.
What a train is made of, and why the data lies about it
Which carriages form a train is published data, but a cab flag does not mean a cab, a repeated allocation turns six coaches into twenty-four, and the accessibility letters have never been decoded. What the feed says, and where it misleads.
Weekend engineering work, and why your train is a bus
Most weekends part of the network is closed for repairs. Where the information comes from, why notices run for months rather than days, and why a rail replacement coach is so much harder to follow than a train.
Where the railway makes trains slow down
Network Rail publishes every stretch of track with a reduced speed limit. Only about a quarter of them state when they will lift, and the oldest still in force dates from 2004 — what the notice says and how to read it.
The three different things “cancelled” means
A cancelled train might never have left, been abandoned half way, or still be running and simply not stopping where you are. Why a part-cancelled run often carries no reason at all, and how the feeds tell them apart.
What the railway wrote down before the day began
551,000 schedules a day, most of them not about passengers. Working times booked to the half minute, padding recorded by name, and the associations that turn one train into another.
What a map knows that a timetable cannot
94% of British railway ways record how they are electrified and 76% their line speed — and a quarter of the network is not a running line at all. What that data supports, and the one statistic it could not.
When the dots stop, and what that actually means
A still dot is usually a signalling quirk, sometimes a line with no coverage, and occasionally an entire describer area that has stopped reporting — which looks identical to a stationary railway unless something says otherwise.
Bus
Why buses run early, and what punctuality really measures
About a quarter of buses arrive ahead of time, which is a timetabling artefact rather than a fault. Why that happens, and why our punctuality figure is lower than the one your operator publishes.
Related reading
Where the data comes from lists every feed behind the site, what each one covers and the licence it is used under. How it works is the shorter overview of the same mechanics. The daily archive records how the railway actually ran on a given date, which is the raw material a lot of these explanations are drawn from.