Manchester-Derby Any Permitted Route Query

[M60lad]

Does anybody know what routes a Manchester-Derby return is valid on? Is it just:

Manchester-Derby via Stoke
Manchester-Derby via Sheffield

Is it actually valid via Nottingham on EMT although I doubt it with the train not serving Derby first

 

[1B85]

It is valid via Nottingham (and it seems Loughborough also if someone can confirm?) via mapped route PR...

 

[ForTheLoveOf]

It is valid via Nottingham (and it seems Loughborough also if someone can confirm?) via mapped route PR...

Would seem to be permissible via Loughborough, as long as there is a train between Loughborough and Nottingham that doesn't call at or pass through Derby.

 

[mallard]

as long as there is a train between Loughborough and Nottingham that doesn't call at or pass through Derby.

There are no timetabled Loughborough - Nottingham services that go via Derby in either direction, so there are 36 (and 35 in the opposite direction) weekday services that meet that criteria!

There are two trains per day that do Loughborough - Nottingham - Derby; they are the only services that call at all three of those stations.

 

[1B85]

There are no timetabled Loughborough - Nottingham services that go via Derby in either direction, so there are 36 (and 35 in the opposite direction) weekday services that meet that criteria!

There are two trains per day that do Loughborough - Nottingham - Derby; they are the only services that call at all three of those stations.

Yes, but you can't do Manchester-Sheffield-Chesterfield-Nottingham-Loughborough and then Loughborough-Nottingham-Derby as that's a double-back to Nottingham, only if the train goes to Derby from Loughborough without calling at Nottingham...

 

[robbeech]

Derby Sheffield Manchester
Derby Nottingham Sheffield Manchester
Derby Stoke Manchester
Derby Crewe Manchester
Derby Tamworth Stoke Manchester (or change at Stafford)

 

[bb21]

It is valid via Nottingham (and it seems Loughborough also if someone can confirm?) via mapped route PR...

Would seem to be permissible via Loughborough, as long as there is a train between Loughborough and Nottingham that doesn't call at or pass through Derby.

There are no timetabled Loughborough - Nottingham services that go via Derby in either direction, so there are 36 (and 35 in the opposite direction) weekday services that meet that criteria!

There are two trains per day that do Loughborough - Nottingham - Derby; they are the only services that call at all three of those stations.

You cannot go via Loughborough. Map permission finishes at East Midlands Parkway.

The two routes between Loughborough and Nottingham/Derby only diverge at EMD, and the latter is not a member of Loughborough Group, as no such group exists, so this would be a barred doubleback at EMD.

 

[JB_B]

On one reading, I'm with you, bb21, but I have a fair bit of sympathy with the view that via Loughborough should be valid.

The public-facing routeing guide asks the passenger to trace a route on the map(s) between origin and destination routeing points which doesn't pass through the same station twice.

But the maps don't show the actual stations you'll pass through; instead, they show routeing points groups (a collection of stations), routeing points (individual stations) and interchanges (again, individual stations) - which between account for only around 500 of the 2,500+ stations on the network.

I'd say it's entirely possible that the (semi-mythical) author/s of the routeing guide actually intended "station" to be as a convenient short-hand for "routeing point group, routeing point or interchange" - ie the entities shown on the maps rather than meaning every possible station on the physical rail network. (Given you're performing this check while tracing a route on the map, it's reasonable to read in " on the map/s " after the word "station". )

In other words, the other 2000 odd stations aren't on the map so, while you might physically pass through them when on your journey, you don't pass though them while tracing your route on the map and that is where it matters.

( The mismatch between the map nodes and links and the underlying physical network when taking the strict reading could potentially advantage passengers in some unusual circumstances but it gives bizarre results - see e.g. https://www.railforums.co.uk/thread...-bhm-group-season-ticket.158061/#post-3246800 - in any case I'm not convinced that there is always an unambiguous way of mapping from the routeing-guide world down to the underlying geography of the rail network.)

 

[SSp]

I think the last post means via Aberdeen is fine.

 

[JB_B]

I think the last post means via Aberdeen is fine.

... nope.

 

[yorkie]

On one reading, I'm with you, bb21, but I have a fair bit of sympathy with the view that via Loughborough should be valid.

The public-facing routeing guide asks the passenger to trace a route on the map(s) between origin and destination routeing points which doesn't pass through the same station twice.

But the maps don't show the actual stations you'll pass through; instead, they show routeing points groups (a collection of stations), routeing points (individual stations) and interchanges (again, individual stations) - which between account for only around 500 of the 2,500+ stations on the network.

I'd say it's entirely possible that the (semi-mythical) author/s of the routeing guide actually intended "station" to be as a convenient short-hand for "routeing point group, routeing point or interchange" - ie the entities shown on the maps rather than meaning every possible station on the physical rail network. (Given you're performing this check while tracing a route on the map, it's reasonable to read in " on the map/s " after the word "station". )

In other words, the other 2000 odd stations aren't on the map so, while you might physically pass through them when on your journey, you don't pass though them while tracing your route on the map and that is where it matters.

( The mismatch between the map nodes and links and the underlying physical network when taking the strict reading could potentially advantage passengers in some unusual circumstances but it gives bizarre results - see e.g. https://www.railforums.co.uk/thread...-bhm-group-season-ticket.158061/#post-3246800 - in any case I'm not convinced that there is always an unambiguous way of mapping from the routeing-guide world down to the underlying geography of the rail network.)

The electronic rules include that the following condition (among others) is satisfied: "the Map Journey does not pass through any station twice" (obviously that excludes situations where it is permitted for another reason, such as the Group Stations Rule, or by an easement)

I think the last post means via Aberdeen is fine.

No; Aberdeen does not appear on any of the relevant maps. But if did, plenty of intermediate stations would be shown on the relevant maps.

 

[JB_B]

Thanks, Yorkie - I'm happy to accept defeat on this one. It's a pity, because it makes the task of getting a visual overview of permitted routes significantly harder. It could be overcome in this case by, e.g, manually splitting links and inserting Loughborough as a 'virtual' interchange. Unfortunately, I suspect that this would become impractical for the densest parts of the network.

 

[bb21]

On one reading, I'm with you, bb21, but I have a fair bit of sympathy with the view that via Loughborough should be valid...

I can see where you are coming from, as the maps confused me initially too. The maps are utterly rubbish and there is simply no need for them to be that shabby.

I wouldn't want to be the one trying to argue against the doubleback though, mind. Others may fancy their chances.

 

[kieron]

The maps are just an attempt to communicate some complicated information in a simpler way.

To take an example, person A could use the routes on the YY map* could go between Pontefract Baghill and York on a train which stops at Church Fenton. Person B could use the same map to go between Leeds and Pontefract Baghill, this time changing trains at Church Fenton. Person C could use the map to go between Wakefield and York via Church Fenton, but would have to go via Castleford rather than Pontefract, and couldn't use any of the same train as persons A or B.

Perhaps there's a way to represent this which isn't "rubbish", but it's not obvious to me what this would be. I'm wary of the idea of just sticking a few intermediate points in here and there, as it's hard to check whether or not this changes anything.

* YY isn't a valid map combination for person A, but YA is, and includes the same route. Similarly, person C needs KY rather than YY. There aren't currently any routes which use the section of YY between Pontefract and York.​

 

[yorkie]

... it makes the task of getting a visual overview of permitted routes significantly harder. ...

It's exceptionally difficult to create a fares engine to calculate permitted routes accurately*. I don't think many people realise just how difficult it is!

(* Some accredited systems, e.g. Worldline/Atos WebTIS are not sufficiently 'accurate' in my opinion! I don't think the people who created/maintain that system are aware of just how many routes are calculated incorrectly!)

 

[ForTheLoveOf]

Would it be so difficult for ATOC/RDG to create a checker to spit out whether or not a certain intended routing is permitted, whether by route or by itinerary, to give a conclusive contractual answer?

 

[yorkie]

Would it be so difficult for ATOC/RDG to create a checker to spit out whether or not a certain intended routing is permitted, whether by route or by itinerary, to give a conclusive contractual answer?

I am not sure I understand the question. ATOC/RDG don't create stuff themselves; they contract third party companies to do it

Their NRE website uses Silverail IPTIS:
http://www.nationalrail.co.uk/100299.aspxT

The Online Journey Planner (OJP) is the engine used by National Rail Enquiries to plan routes, calculate fares and establish ticket availability. OJP derives routing information from SilverRail’s journey planning engine known as IPTIS (Integrated Passenger Transport Information System)

IPTIS is also used by the Govia owned TOCs (GTR and Southeastern) as well as Virgin Trains East Coast and a few others.

It's pretty accurate but it does give some incorrect results.

 

[ForTheLoveOf]

What I mean is - there are unofficial calculators out there that will tell you whether or not an itinerary or route is a permitted route for a specific ticket - why can't www.nationalrail.co.uk have a kind of 'Expert mode' for this?

 

[yorkie]

What I mean is - there are unofficial calculators out there that will tell you whether or not an itinerary or route is a permitted route for a specific ticket

NRE does do that.

 

[ForTheLoveOf]

NRE does do that.

Where? When I click on "other trains you can take with this ticket", it shows other trains of the same routeing as the search, not all the possible trains/routeing combinations. And if I click on the "other valid routes" button, nothing comes up. I've never seen the "other valid routes" button come up with anything, on any route, on any ticket, on any computer/browser!

 

[yorkie]

Where?

You earlier asked if it will say 'whether or not an itinerary or route is a permitted route'; it does do that (albeit with not quite 100% accuracy).

When I click on "other trains you can take with this ticket", it shows other trains of the same routeing as the search, not all the possible trains/routeing combinations. And if I click on the "other valid routes" button, nothing comes up. I've never seen the "other valid routes" button come up with anything, on any route, on any ticket, on any computer/browser!

I agree that this feature rarely works. It was buggy at the best of times but it looks like it may have completely stopped working altogether now. The NRE website is full of bugs and should be completely replaced in my opinion.

There may not currently be a functional publicly available website that shows all the mapped routes available for any given fare at present.

Is there a particular fare/route you had in mind for us to check?

 

[ForTheLoveOf]

You earlier asked if it will say 'whether or not an itinerary or route is a permitted route'; it does do that (albeit with not quite 100% accuracy).

I agree that this feature rarely works. It was buggy at the best of times but it looks like it may have completely stopped working altogether now. The NRE website is full of bugs and should be completely replaced in my opinion.

There may not currently be a functional publicly available website that shows all the mapped routes available for any given fare at present.

Is there a particular fare/route you had in mind for us to check?

No, but I often check a lot of routes on mobile, where it's rather a hassle to use the Yellow Pages, the ATOC RP Calc and the Routeing Maps!

 

[JB_B]

The maps are just an attempt to communicate some complicated information in a simpler way.
...

Indeed - they're actually a rather elegant way of showing the information. It's just a pity they don't include all the information needed to check for double-backs.

 

[JB_B]

It's exceptionally difficult to create a fares engine to calculate permitted routes accurately*. I don't think many people realise just how difficult it is!...

I'd say you're correct on both counts.

 

[JB_B]

Would it be so difficult for ATOC/RDG to create a checker to spit out whether or not a certain intended routing is permitted, whether by route or by itinerary, to give a conclusive contractual answer?

The standard answer that RDG would offer ( and the answer that DfT are all too happy to accept ) is that any accredited TIS will offer to tell you whether or not any given journey is valid for a given ticket.

In one sense, that's fair enough - that's what most passengers want to know most of the time.

In a wider sense, this sort of black-box "computer says 'yes'" / "computer says 'no'" approach is wholly inadequate (and of course it's increasingly problematic in the wider world.)

TISs don't show their working - so you can't see what routes they've considered and which constraints they've taken into account.

It's extremely tedious to try to tease out how the result is arrived at because the "yes"/"no" result if a function of the various routeing and fares rules together with the timetable.

That means it's almost impossible to check whether they are accurate or not. It certainly cannot be the case that all TISs are wholly accurate to a single standard ( given you can get different results for the same queries across TISs ).

I understand the TIS developers have access to a tighter definition of permitted route than that provided to the public in the ( hopelessly underdetermined ) routeing guide but even so, I'd guess it's inevitable that differing interpretations, implicit assumptions (and maybe even outright errors?) will lead to different results.

A first step towards transparency would be to be able to visualize permitted routes before taking into account timetabling (and fares) constraints ... but even that is far from straightforward.

 

[kieron]

Indeed - they're actually a rather elegant way of showing the information. It's just a pity they don't include all the information needed to check for double-backs.

What I was attempting to show was that they can't always show this.

In this scenario, there is a route for which Church Fenton is between Pontefract Group and York, and one in which it is between Leeds Group and Pontefract Group.

Imagine Church Fenton was added to the map, linked to Leeds Group, Pontefract Group and York. The Pontefract Group-York link could then be removed, although the Leeds Group-Pontefract Group one could not (for the trains which go through Castleford if nothing else).

This would show the Leeds-Church Fenton-Pontefract Baghill-Church Fenton-York double back. It would also create a Leeds Group-Church Fenton-York route which doesn't exist on the current map. If you don't want to create this route, the only way to do it (without changing the map format) is to leave Church Fenton off the map entirely.

Adding East Midlands Parkway to PR wouldn't add any new routes to the map itself (I haven't checked multi-map sequences), so is less of a problem. If you know the maps don't show every double back, you can't rely on them to show you whether or not a route involves a double back. If you can't rely on them, you have to check elsewhere, and having to check Derby-Loughborough-Nottingham as well is no real hardship.

 

[JB_B]

What I was attempting to show was that they can't always show this.
...
This would show the Leeds-Church Fenton-Pontefract Baghill-Church Fenton-York double back. It would also create a Leeds Group-Church Fenton-York route which doesn't exist on the current map. If you don't want to create this route, the only way to do it (without changing the map format) is to leave Church Fenton off the map entirely....

Yes, I was thinking that (if it turned out to be practical) I'd have to introduce any such additional "pseudo-interchanges" as a separate class to the mapped interchanges and routeing points and use them for pruning out double-backs only (to avoid creating spurious new routes.)

Would it be reasonable to disregard the converse case, where repeated visits to different members of an (origin or dest) routeing point group don't result in a double -back (as in Birmingham to Stafford via Smethwick , Birmingham and Stafford) ?

 

[infobleep]

No, but I often check a lot of routes on mobile, where it's rather a hassle to use the Yellow Pages, the ATOC RP Calc and the Routeing Maps!

Alas the answer right now is to use a laptop or desktop computer. Perhaps RDG don't think mobiles are popular enough yet or going to be popular enough to create a mobile friendly version