HS2 reaches Euston in the 2040s. Fares, for once, did not rise.

Measured at every stop on the journey, which is a harder test than terminating punctuality. Cancelled services are excluded from this figure entirely, so cancelling a train improves the punctuality number.

A weighted score in which a full cancellation counts as one and a part cancellation as a half. Roughly one train in every thirty-one.

The denominator behind both figures above. A 3.2% cancellation score across 1.9 million planned services is roughly 61,000 trains that did not run as planned, in three months.

The first freeze in thirty years. Without it regulated fares would have risen 5.8%. Across Great Britain the fares index was unchanged while RPI rose 4.1%, so fares fell in real terms. A promise made and a promise delivered, recorded here as plainly as the broken ones.

Up 2% on the year before, but still 89% of the pre-pandemic level, and 86% outside London. The recovery is real and incomplete at the same time.

Up 1% on the previous year. Buses carry far more passenger journeys than the railway, and attract a fraction of the attention.

Phase One was originally to open in 2026. The NAO puts the slippage at between around three and thirteen years, depending on which promised date you measure from.

Until then, high-speed passengers change at Old Oak Common in west London. A child starting school in 2026 will have left university before the line reaches central London.

£3.8bn already spent, with a further £4.1bn anticipated, for a station that will not open until the 2040s.

What HS2 Ltd expects to spend on reorganising the project. Not track, not trains, not stations. The NAO's verdict is that the department is making progress but has to get this reset right, having failed before.