P-Coding - TransPennine example

[whole diary]

[grahame]

From The BBC^» (British Broadcasting Corporation - home page)

I don't think we've had a TransPennine thread but anyway ...

The government-controlled firm said services between Leeds and Manchester would be reduced from four to three per hour during off-peak times.

It will mean the company runs 300 services a day instead of 320.

Managing director Chris Jackson said the temporary change would clear the backlog of driver training and put the firm on a more "stable" footing.

He told BBC Look North: "This plan to slightly step back the timetable in December will help to fix the business in the long term and operate a much more stable and reliable timetable for customers.

...  and that may be on top of the withdrawal of the locomotive and coach trains  ... it goes on ...

Figures showed the equivalent of one-in-seven services being cancelled in the first four weeks of April, more than any other operator.

According to the government operator which now runs the routes, in early 2023 it operated 276 services on a daily basis with around 50 to 60 train cancellations each day.

The company also came under fire in February for its high use of p-coding, which is used to cancel trains the night before in response to poor weather or damage to rail infrastructure.

Because the announcements are not made on the day, these cancellations are not included in official performance statistics.

Them's up north so we haven't mentioned it (much) before, but this sounds like the official name for having a cancellation that's not a cancellation.   Let's look at the Wiltshire version - for the fist week of this month

A timetable published in May had 59 trains from Westbury to Swindon.
If 3 were If 56 had run, that's  95% which is not brilliant but understood.

A timetable published in May had 59 trains from Westbury to Swindon.
3 were declared cancelled, but another 14 were withdrawn from the timetable due to industrial action. So 42 ran.  Description ...

3 declared cancelled out or 59 - so that's a 95% record
3 declared cancelled out of 45 - so that's a 93.5% record
17 didn't run out of 59 - so that's a 71.2% record.

I always wondered ...

As a passenger, I understand a timetable that has 59 trains in it.  I can only use 42 when the time comes. So I thing 71.2% is the honest description.

[JayMac]

I'm reliably informed that up to a dozen drivers from one TPE^▸ (Trans Pennine Express) depot alone are moving to DRS^» (Direct Rail Services Ltd, a national rail freight operator providing a range of services with its head office in Carlisle and depots at the South East, York, Crewe, Sellafield, and Scotland. - about) with the Class 68s, or to other operators. Drivers who have competency on other stock with TPE, not just the 68s, but fed up with the management culture at TPE. A culture which hasn't improved with the switch to Directly Operated Railways.

I don't think the backlog of training, leading to the reduction in services at TPE, is as temporary as the MD is making out.

[CyclingSid]

Does 0% trains equal 100% reliability?

[grahame]

Does 0% trains equal 100% reliability?

Ah ... the philosophy and concept of infinity when you divide by zero with a scrubbed timetable!

Zero trains with even one in the timetable is 0% reliability
Zero trains without any in the timetable is mathematically undefined

You will see the absurdity of stats with a tiny timetable if you look at reliability stats for the likes of Shippea Hill or Pilning.