ORR Press Release - Late-Notice Cancellations

[Bald Rick]

The ORR have set out their interpretation unambiguously that it's not compliant. That's obviously the interpretation taken by other operators such as CrossCountry, who have gone on record quite freely to say they wouldn't do this.

Cross Country do it, as do almost all operators. TPE have been doing it more than most recently, but it is very common practice, and has been for 25 years plus.

So the question is I guess why is your interpretation, and the presumed interpretation of TPE and as was FCC, so different?

It’s not an interpretation. The code is very clear that you can change the plan of the day, for any reason, up to 2200 the previous day. There is currently no method in industry performance systems of identifying why those changes have been made - perhaps there should be, but there isn’t right now. Neither is there any systematic way of notifying cancelled trains in systems before the day of operation.

Furthermore, they want changes to the reporting process to prevent TPE and others from pulling the wool over the eyes of the public by reporting on all cancellations for resource unavailability together with any cancellations made on the day.

But why just traincrew resource availability? What about cancellations caused by there being no signaller avaialble - there’s been lots of examples recently. Or what about shortage of resource on engineering works, that triggers an overrun, which causes P coded cancellations? That happens all the time, big example at Christmas.

Are we really saying “we want passengers to know what the real cancellation rate is, but only for cancellations caused by traincrew” ? Or put another way “cancellations made the day before don’t matter unless they are caused by traincrew”. Seems inconsistent to me.

As I've said before, it's incredible that you currently have better passenger rights flying with Ryanair than travelling on the railway.

Tell that to anyone on a Ryanair flight arriving 2h30 late.

 

[Starmill]

It might be within the letter of the Code, but clearly the ORR is of the view that it is not within its spirit. Whether that still allows them to take enforcement action if TOCs don't comply with their requests is unclear; their failure to warn of a penalty for non-compliance suggests that it probably doesn't.

This is true - but they're very clearly 'rolling the pitch' here to give themselves the power to do that in future.

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

Cross Country do it, as do almost all operators. TPE have been doing it more than most recently, but it is very common practice, and has been for 25 years plus.

Common practice doesn't equal contractually compliant though does it? Surely you appreciate that. CrossCountry specifically said to the Guardian they don't do it so it's a big accusation you're making of them.

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

It’s not an interpretation.

Either it's a difference of interpretation between you and ORR, or one of you is being deliberately misleading. Are you suggesting that ORR are seeking to deliberately mislead the public via this press release?

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

But why just traincrew resource availability? What about cancellations caused by there being no signaller avaialble - there’s been lots of examples recently. Or what about shortage of resource on engineering works, that triggers an overrun, which causes P coded cancellations? That happens all the time, big example at Christmas.

Are we really saying “we want passengers to know what the real cancellation rate is, but only for cancellations caused by traincrew” ? Or put another way “cancellations made the day before don’t matter unless they are caused by traincrew”. Seems inconsistent to me.

Why shouldn't a train be cancelled on the day due to a shortage of signalling staff or overrunning engineering works? Once again sounds like an attempt to pull the wool over the eyes of the public to me. These events, by definition, are unplanned disruption.

 

[Bald Rick]

Common practice doesn't equal contractually compliant though does it? Surely you appreciate that. CrossCountry specifically said to the Guardian they don't do it so it's a big accusation you're making of them.

AIUI It is contractually required of some operators to do this by their specifier.

Perhaps I haven’t explained very well. Cross Country - like many other operators - made multiple changes to their plan when their drivers stopped working rest days, and they did much of this at short notice, often on a weekly basis. These changes went through as P codes. I’d agree that Cross Country are not doing this now - they don’t need to - but that’s not to say it doesn’t happen.


I still don’t understand what the objectibe is. is it

1) to enable passengers to have better information about what services are running (or cancelled)

2) to improve the quality of the statistics.

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

Are you suggesting that ORR are seeking to deliberately mislead the public via this press release?

Not at all. But from what I can see, it seems that they have fired off a letter without first understanding what the underlying issues are, and how it might be solved.

Why shouldn't a train be cancelled on the day due to a shortage of signalling staff or overrunning engineering works?

Because if you know about it a day or two before, better to cancel it then so that intending passengers know about it asap.

Are you suggesting that when a TOC knows a train won’t be running, that it shouldn’t cancel it from the timetable Until the day it is due to run?

 

[Starmill]

Because if you know about it a day or two before, better to cancel it then so that intending passengers know about it asap.

Are you suggesting that when a TOC knows a train won’t be running, that it shouldn’t cancel it from the timetable Until the day it is due to run?

That can be done without p-coding though. It's the p-coding that ORR are objecting to, not giving notice.

ORR make this explicit. If the service is p-coded it will disappear entirely from the customer's app or the screen and announcements at the station. If it's cancelled this doesn't happen - that's the objective here.

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

I still don’t understand what the objectibe is. is it

It's objectionable because the position of ORR is that trains should simply be cancelled and not p-coded. That doesn't impede giving good notice. I'm not sure how this is controversial?

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

Not at all. But from what I can see, it seems that they have fired off a letter without first understanding what the underlying issues are, and how it might be solved.

Forgive me but I think that if you're going to take an interpretation of the rules which contradicts ORR's you perhaps could explain why that is a little bit more. You must understand that a neutral reader of the thread as it stands would opine that you and ORR cannot possibly both be right?

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

The ORR appears to be suggesting they are happy for resource related cancellations, and the concomitant gaslighting of passengers, to continue so long as they are done at least 72 hours in advance. I fear the outcome could therefore simply be to slightly increase the notice given for such "removals" from the timetable.

In theory, that 72 hours could be used to give publicity about the emergency timetable, if there were money put into doing that, and if it were kept as a genuine last resort yo introduce any such emergency timetable. However, I firmly agree that in reality this wouldn't be any improvement over the situation today.

ORR and many others have correctly identified that late the night before gives absolutely no scope for reasonably informing the customer of the change, they may already have retired for the night by 2200. It's so utterly unreasonable.

== Doublepost prevention - post automatically merged: 20 Jan 2023 ==

Tell that to anyone on a Ryanair flight arriving 2h30 late.

It was clear that the comment you're referencing was for notice of amendments, not delays on the day.

 

[Bald Rick]

That can be done without p-coding though. It's the p-coding that ORR are objecting to, not giving notice.

Unfortunately, that’s not the case, hence the problem. Any train cancelled from the timetable before 2200 the night before, for any reason (and regardless of what code is entered) is automatically removed from the timetable system for the next day and does not appear.

Also lots of services are cancelled at short notice for other reasons, including resource reasons as I mention above. What should happen with those?

 

[Parjon]

Unfortunately, that’s not the case, hence the problem. Any train cancelled from the timetable before 2200 the night before, for any reason (and regardless of what code is entered) is automatically removed from the timetable system for the next day and does not appear.

Also lots of services are cancelled at short notice for other reasons, including resource reasons as I mention above. What should happen with those?

Since they're cancelling trains just before 22:00, if the system is so restricted as you claim it to be, presumably they could cancel them at 22:01? What's the reason for not waiting those few minutes?...

Not only would it give a clear picture of which trains are and are not running, but it would ensure that performance statistics are accurate.

Were trains cancelled as a result of signalling issues, the affected train company would be claiming against network rail. The responsibility for these particular cancellations rests entirely with the train operating company.

The question remains as to whether any organisation or individual has profited from an official record which is totally misleading as to true cancellations.

Perhaps this is something else for Helen Pidd to look at, since the system appears uninterested in making the picture clear.

 

[Watershed]

Tell that to anyone on a Ryanair flight arriving 2h30 late.

Despite having been in that situation several times of late, I'd still go for Ryanair (and their ilk) every day of the week and twice on Sundays. They provide a far more reliable service than the railway, even if not necessarily quite as punctual.

Most importantly, 14 days before travel you know for certain what the schedule for the purposes of delay compensation and food/drink entitlement is. And whilst I've had my fair share of schedule changes, very few have been anywhere close to 14 days from departure - most at least a month or two ahead.

 

[Bald Rick]

Since they're cancelling trains just before 22:00

Are they?

 

[hick]

Unfortunately, that’s not the case, hence the problem. Any train cancelled from the timetable before 2200 the night before, for any reason (and regardless of what code is entered) is automatically removed from the timetable system for the next day and does not appear.

Also lots of services are cancelled at short notice for other reasons, including resource reasons as I mention above. What should happen with those?

Are you sure this is correct?

Darwin takes the CIF file from capacity planning. A P coded cancellation isn't in the CIF as it's being done under control arrangements and Darwin doesn't take cancellations from TRUST.

My understanding was the trains are manually cancelled in Darwin (or other systems that feed Darwin). Some TOCs are also then manually deleting the trains, but others don't.

 

[Mugby]

When short notice changes to the following days timetable are made, are the resultant timetables, complete with alterations, ever published in full?

 

[Starmill]

Unfortunately, that’s not the case, hence the problem. Any train cancelled from the timetable before 2200 the night before, for any reason (and regardless of what code is entered) is automatically removed from the timetable system for the next day and does not appear.

Well that's demonstrably wrong I'm afraid, because trains frequently appear cancelled on journey check the previous day and then show cancelled on the screen and live departures on the day of travel. This is common on Northern for example.

== Doublepost prevention - post automatically merged: 21 Jan 2023 ==

Also lots of services are cancelled at short notice for other reasons, including resource reasons as I mention above. What should happen with those?

They're cancelled at short notice and thus appear cancelled in the ordinary sense of the word. They're not withdrawn and p-coded. I don't understand how there's a problem with this?

== Doublepost prevention - post automatically merged: 21 Jan 2023 ==

Most importantly, 14 days before travel you know for certain what the schedule for the purposes of delay compensation and food/drink entitlement is. And whilst I've had my fair share of schedule changes, very few have been anywhere close to 14 days from departure - most at least a month or two ahead.

Precisely. And, since the Supreme Court's decision, even Ryanair will generally pay claims for food or drink or compensation for delay itself, regarding qualifying delays not announced until after the 14 day notice. Thus as long as you're at your origin on time, you can be certain where you will stand a full fourteen days out, and plan around that. To me, that seems highly preferable to the railway's offering. You can also use this to wait until 14 days before departure and then make bookings, paying whatever prices are necessary for that certainly, to go with your other plans such as hotels or connecting transport, and be safe in the knowledge that the airline may not change it substantially at their whim. The railway expects you to wait until just two hours before the day of travel if you wish to do this.

== Doublepost prevention - post automatically merged: 21 Jan 2023 ==

When short notice changes to the following days timetable are made, are the resultant timetables, complete with alterations, ever published in full?

I think it depends on what you mean by published. They're shown in the open data but there's no guarantee they'll be published in a timetable on the website for example. They're often not.

== Doublepost prevention - post automatically merged: 21 Jan 2023 ==

The question remains as to whether any organisation or individual has profited from an official record which is totally misleading as to true cancellations.

Perhaps this is something else for Helen Pidd to look at, since the system appears uninterested in making the picture clear.

It's certainly a possibility that TransPennine Express have received slightly more in management fees than they would have if they weren't p-coding such a large proportion of services. However, we cannot say that for sure without sight of the commercially confidential parts of their agreement with the Department, and in any case such fees are a very tiny proportion of the total costs of running the franchise, usually under 2%.

 

[Parjon]

The question remains as to whether any organisation or individual has profited from an official record which is totally misleading as to true cancellations.

Perhaps this is something else for Helen Pidd to look at, since the system appears uninterested in making the picture clear.

It's certainly a possibility that TransPennine Express have received slightly more in management fees than they would have if they weren't p-coding such a large proportion of services. However, we cannot say that for sure without sight of the commercially confidential parts of their agreement with the Department, and in any case such fees are a very tiny proportion of the total costs of running the franchise, usually under 2%.

The issue isn't whether any sums would be headlines grabbing, but that it would be public money being claimed/handed out.

We've all been on the receiving end of TOCs making operational decisions that result in a poor passenger experience for non-operational, profit maximising reasons.

There may not be anything stopping them from stiffing passengers through playing games with the rules, but being paid money for a picture of service that is manifestly false would be an entirely different matter.

The ORR may buy that all this P-coding was done with the passenger's best interests at heart, but I cannot see the logic in what they are saying.

Whether any money has changed hands on the basis of these figures may be central as to who they may need to convince.

 

[Starmill]

The issue isn't whether any sums would be headlines grabbing, but that it would be public money being claimed/handed out.

I agree. But it's not an easy question to answer without some serious digging.

== Doublepost prevention - post automatically merged: 21 Jan 2023 ==

The ORR may buy that all this P-coding was done with the passenger's best interests at heart, but I cannot see the logic in what they are saying.

To be fair to ORR, they do appear to have suddenly found a spine. Not that it excuses past inaction.

 

[D6700]

Are they?

To me, that reads like a flippant dismissal. I do hope it wasn't intended that way, given the correct answer to your question is "Yes, they are".

 

[PG]

There may not be anything stopping them from stiffing passengers through playing games with the rules, but being paid money for a picture of service that is manifestly false would be an entirely different matter.

A precedent was set with the withdrawal from Govia of Southeastern described by the transport secretary as an “appalling breach of trust”, which is basically fraud.

 

[Bald Rick]

Are you sure this is correct?

Well that's demonstrably wrong I'm afraid, because trains frequently appear cancelled on journey check the previous day and then show cancelled on the screen and live departures on the day of travel. This is common on Northern for example.

I should explain. In so far as performance systems are concerned - anything cancelled from the timetable for any reason prior to 2200 the night before does not count as a cancellation in the performance stats. ThAt is because the network code requires it. Anything cancelled before 2200 is automatically P coded - Cancellations can be retrospectively recoded, but that must be done manually.

What appears downstream is a similar process. Darwin can be amended to show cancellations for services that have been cancelled the night before, but that is a manual process. And it doesn’t change what appears in the performance stats.

They're cancelled at short notice and thus appear cancelled in the ordinary sense of the word. They're not withdrawn and p-coded. I don't understand how there's a problem with this?

But they are withdrawn and p coded, that’s the point. For all sorts of incidents.


Look, I’m not arguing against the need to change the process. What I‘m saying is that automatically measuring this, so that regular reporting is produced specifically on train crew and rolling stock cancellations done before the day of the rave (which is what the ORR want) is not something that can be easily done, not at all. It will mean lots of manual effort (which means cost), and changes to the network code and in some cases contracts between TOCs / funder / Network Rail. None of that will be quick.

I suspect the end result will be lots more cancellations on the day, and passengers being given a worse service as a result.

 

[IanXC]

I should explain. In so far as performance systems are concerned - anything cancelled from the timetable for any reason prior to 2200 the night before does not count as a cancellation in the performance stats. ThAt is because the network code requires it. Anything cancelled before 2200 is automatically P coded - Cancellations can be retrospectively recoded, but that must be done manually.

What appears downstream is a similar process. Darwin can be amended to show cancellations for services that have been cancelled the night before, but that is a manual process. And it doesn’t change what appears in the performance stats.




But they are withdrawn and p coded, that’s the point. For all sorts of incidents.


Look, I’m not arguing against the need to change the process. What I‘m saying is that automatically measuring this, so that regular reporting is produced specifically on train crew and rolling stock cancellations done before the day of the rave (which is what the ORR want) is not something that can be easily done, not at all. It will mean lots of manual effort (which means cost), and changes to the network code and in some cases contracts between TOCs / funder / Network Rail. None of that will be quick.

I suspect the end result will be lots more cancellations on the day, and passengers being given a worse service as a result.

This might be too simplistic but...

Currently we're talking about "PG Planned cancellation where that cancellation is identified and agreed prior to 22.00 the day before the schedule runs", if a new code were created (for the sake of argument) "PH" for TOC responsibility, with "PG" remaining in use for Network Rail responsibility then I think the data that is sought would be available?

It's also worth noting that the TOC need only have submitted the requests by 2200, Network Rail do not also have to have input them by that time.

Issues of passenger information via Darwin are in some ways a different problem - both using a much more modern and amendable system, but also related to processes undertaken by TOCs.

 

[Bald Rick]

Currently we're talking about "PG Planned cancellation where that cancellation is identified and agreed prior to 22.00 the day before the schedule runs", if a new code were created (for the sake of argument) "PH" for TOC responsibility, with "PG" remaining in use for Network Rail responsibility then I think the data that is sought would be available?

That’s an answer to part of it, but it doesn’t solve the contractual / network code issues. Nor does that help with passenger info.

 

[The Ham]

I should explain. In so far as performance systems are concerned - anything cancelled from the timetable for any reason prior to 2200 the night before does not count as a cancellation in the performance stats. ThAt is because the network code requires it. Anything cancelled before 2200 is automatically P coded - Cancellations can be retrospectively recoded, but that must be done manually.

What appears downstream is a similar process. Darwin can be amended to show cancellations for services that have been cancelled the night before, but that is a manual process. And it doesn’t change what appears in the performance stats.




But they are withdrawn and p coded, that’s the point. For all sorts of incidents.


Look, I’m not arguing against the need to change the process. What I‘m saying is that automatically measuring this, so that regular reporting is produced specifically on train crew and rolling stock cancellations done before the day of the rave (which is what the ORR want) is not something that can be easily done, not at all. It will mean lots of manual effort (which means cost), and changes to the network code and in some cases contracts between TOCs / funder / Network Rail. None of that will be quick.

I suspect the end result will be lots more cancellations on the day, and passengers being given a worse service as a result.

I'm curious and thinking "out loud".

If you could output the timetable (i.e. the one agreed from the previous May/December) as (say) an Excel file and then output the on the day timetable also in the same format you can then use Excel to compare the differences, that would flag up where there were changes wrote quickly.

Now Excel is unlikely to cope with 22,500 rows/columns of data, however it's likely that using a database could do a similar function.

Yes you'd still have to go through and see which were which (although I suspect you could create a filter to count each one which changed any was from each TOC).

You could then add the daily numbers to a running weekly/monthly/quarterly total.

Yes it would take time to setup and test, however once working you could be fairly sure that you've got the data working.

Once you've got that working you could then do some fairly clever stuff, like applying a less than 1 score for cancellations made before the day whilst a score of 1 for on the day. This would mean that there's still an advantage to the TOC to cancel the night before over on the day, however it would still count towards their poor service ratings.

For example anything cancelled pre service day -3 counts as zero anything service day -3 to -1 counts as 0.75 and anything service day counts as 1.

Whilst that would mean creating the file pre service day -4 and at the end of pre service day -1 and compare between the two, it's not being the realms of possibility.

Now having identified the "missing" service, the system could also be updated to show the real-time information (including station displays) that the train is no longer running.

Probably better to have any cancelled service as 2nd or 3rd and for a short time as 1st, but after a while display it as the "override information" in the third line (i.e. shows up like a 4th service) so people can see that it's not run for a bit longer but still allowing people to see what the next train actually is.

 

[Confused52]

Are you sure this is correct?

Darwin takes the CIF file from capacity planning. A P coded cancellation isn't in the CIF as it's being done under control arrangements and Darwin doesn't take cancellations from TRUST.

My understanding was the trains are manually cancelled in Darwin (or other systems that feed Darwin). Some TOCs are also then manually deleting the trains, but others don't.

RTT shows 1T66 LIV-MCO due dep 0614 cancelled code PG on 21/1/23.It also shows up on on OTT. Is there may be a way of creating that elusive PToD, perhaps based on time of changes? The old website with changes files might have given me the answer but it isn't there any more.

 

[Starmill]

But they are withdrawn and p coded, that’s the point. For all sorts of incidents.

I feel like you're not really listening to the thing people are saying here. That's fine if you want to stick with whatever you understand the situation to be but there's little purpose in discussing the matter if so. I've explained above that ORR really are explicit about this, they don't want customers to see their trains vanishing a few hours before they're due to travel. They should be marked as cancelled, as this will allow them to claim compensation for delays, it will push notifications through to apps and on email from retailers who have scripts which do that (while simply deleting the train often won't trigger that), and it will not result in the station staff simply shrugging at the customer and saying "there's no train at that time, you're just mistaken". How you can object to this I don't understand.

 

[Watershed]

I'm curious and thinking "out loud".

If you could output the timetable (i.e. the one agreed from the previous May/December) as (say) an Excel file and then output the on the day timetable also in the same format you can then use Excel to compare the differences, that would flag up where there were changes wrote quickly.

Now Excel is unlikely to cope with 22,500 rows/columns of data, however it's likely that using a database could do a similar function.

Yes you'd still have to go through and see which were which (although I suspect you could create a filter to count each one which changed any was from each TOC).

You could then add the daily numbers to a running weekly/monthly/quarterly total.

Yes it would take time to setup and test, however once working you could be fairly sure that you've got the data working.

Once you've got that working you could then do some fairly clever stuff, like applying a less than 1 score for cancellations made before the day whilst a score of 1 for on the day. This would mean that there's still an advantage to the TOC to cancel the night before over on the day, however it would still count towards their poor service ratings.

For example anything cancelled pre service day -3 counts as zero anything service day -3 to -1 counts as 0.75 and anything service day counts as 1.

Whilst that would mean creating the file pre service day -4 and at the end of pre service day -1 and compare between the two, it's not being the realms of possibility.

Now having identified the "missing" service, the system could also be updated to show the real-time information (including station displays) that the train is no longer running.

Probably better to have any cancelled service as 2nd or 3rd and for a short time as 1st, but after a while display it as the "override information" in the third line (i.e. shows up like a 4th service) so people can see that it's not run for a bit longer but still allowing people to see what the next train actually is.

Unfortunately it's not that simple, because the timetable that is in place at the May or December change is often not the timetable that the TOC had intended to run.

You could of course take a snapshot of the timetable 8 weeks out, when STP alterations are due to be published, but that would in some cases still not reflect strike timetables and the like; even though it's a breach of contract, the government are never going to agree to base compensation and cancellation statistics etc. on 'normal' timetables on strike days.

This is a much more complicated exercise than it may appear to be at first glance, which is why it's understandable that the ORR is asking TOCs to come up with the data, as they will generally be best placed to provide it. Of course, if TOCs commit to no longer removing trains from the timetable, but instead (correctly) marking them as cancelled, then they can simply tell the ORR this and don't need to collect any data...

 

[Snow1964]

I think the problem is more headline Grabbing for early morning trains cancelled late. ORR references this with go to bed expecting to get early train and it vanishes.

Maybe cancellation statistics need to be adjusted to 24 hour minimum (so an 08:00 train would need to be P-coded by 07:59 previous day, not 14 hours later as now)

 

[The Ham]

Unfortunately it's not that simple, because the timetable that is in place at the May or December change is often not the timetable that the TOC had intended to run.

You could of course take a snapshot of the timetable 8 weeks out, when STP alterations are due to be published, but that would in some cases still not reflect strike timetables and the like; even though it's a breach of contract, the government are never going to agree to base compensation and cancellation statistics etc. on 'normal' timetables on strike days.

This is a much more complicated exercise than it may appear to be at first glance, which is why it's understandable that the ORR is asking TOCs to come up with the data, as they will generally be best placed to provide it. Of course, if TOCs commit to no longer removing trains from the timetable, but instead (correctly) marking them as cancelled, then they can simply tell the ORR this and don't need to collect any data...

Thanks, I did suspect that it would be complex (hence referencing that there might have to be a capture of the timetable at -3 days - obviously to reduce the number of baseline timetables this could be captured and then used for 6 weeks so the baseline timetable is never more than 8 weeks from the day of travel).

The general point that I was getting at was that it should be fairly easy (in that it's just a data set which computers could provide an output summary) to get an answer to, and one which encouraged TOC's not to cancel on the day if they could cancel sooner whilst still encouraging them not to cancel within the the three days before travel too.

 

[Baxenden Bank]

Unfortunately it's not that simple, because the timetable that is in place at the May or December change is often not the timetable that the TOC had intended to run.

You could of course take a snapshot of the timetable 8 weeks out, when STP alterations are due to be published, but that would in some cases still not reflect strike timetables and the like; even though it's a breach of contract, the government are never going to agree to base compensation and cancellation statistics etc. on 'normal' timetables on strike days.

This is a much more complicated exercise than it may appear to be at first glance, which is why it's understandable that the ORR is asking TOCs to come up with the data, as they will generally be best placed to provide it. Of course, if TOCs commit to no longer removing trains from the timetable, but instead (correctly) marking them as cancelled, then they can simply tell the ORR this and don't need to collect any data...

Problem solved?
Cancel and show as cancelled, rather than removing from view in a Stalinesque manner. For younger readers the recent picture of the space rocket in Cornwall, with or without Boris Johnson applies. Airbrushing trains from history is part of the problem, not their cancelling - for whatever reason and however far out.

There ought to be a commitment to sticking with the published timetable. That is currently May and December but other and more frequent dates could be set - but not daily and certainly not 2200 the previous day. Publish and be damned, or at least commit to deliver what you said you would deliver and bear the costs of that.

Does 'p-coding' remove the need to provide rail replacement buses or arrange ticket acceptance?

 

[Starmill]

Does 'p-coding' remove the need to provide rail replacement buses or arrange ticket acceptance?

No. But these are next to never done anyway for a single cancellation. They sometimes aren't done for multiple cancellations.

 

[The Ham]

No. But these are next to never done anyway for a single cancellation. They sometimes aren't done for multiple cancellations.

Sometimes they're not even done when a landslip closes a line.

 

[Starmill]

Sometimes they're not even done when a landslip closes a line.

Indeed. Outside of 25 and 26 December, it used to be vanishingly rare to be told there's no service and we're not going to do anything about it, go home. Sadly it has become more common.

 

[Adrian1980uk]

Surely this is simple, take away p coding etc. Apply a simple measure of did the train run, as a passenger it doesn't help me if the reason was planned engineering works, weather or just that the TOC can't be bothered to run it.. i.e. no staff or technical failures.

You then have the expectations of bad figures if the country has bad weather but in general you get a more realistic view

 

[Bikeman78]

Precisely. And, since the Supreme Court's decision, even Ryanair will generally pay claims for food or drink or compensation for delay itself, regarding qualifying delays not announced until after the 14 day notice. Thus as long as you're at your origin on time, you can be certain where you will stand a full fourteen days out, and plan around that. To me, that seems highly preferable to the railway's offering. You can also use this to wait until 14 days before departure and then make bookings, paying whatever prices are necessary for that certainly, to go with your other plans such as hotels or connecting transport, and be safe in the knowledge that the airline may not change it substantially at their whim. The railway expects you to wait until just two hours before the day of travel if you wish to do this.

20 years ago, I used to ride around the country planning trips with nothing but a National Rail Timetable for information. The majority of trips went as planned. Imagine trying to do that now. Fine on some TOCs but certainly not Avati, Northern or TPE.