Routing adding validity to a ticket

[Benjwri]

I'm just wanting to check my understanding on what the process is when a ticket's routing adds to the validity given by mapped routes? To give an example an Anytime Return (SOR) from Reading to Worcester, which carries the routing 'EVESHAM/STROUD'.

The journey from Reading to Worcester is mapped via map LH, which works fine for the route via Evesham, but does not contain a route via Stroud, and Stroud is not the shortest route. In this case does this mean you can pass through Stroud using the shortest route that does past through Stroud, do you use the permitted routes between Reading and Stroud, then Stroud and Worcester, or is this route entirely invalid?

 

[alistairlees]

I'm just wanting to check my understanding on what the process is when a ticket's routing adds to the validity given by mapped routes? To give an example an Anytime Return (SOR) from Reading to Worcester, which carries the routing 'EVESHAM/STROUD'.

The journey from Reading to Worcester is mapped via map LH, which works fine for the route via Evesham, but does not contain a route via Stroud, and Stroud is not the shortest route. In this case does this mean you can pass through Stroud using the shortest route that does past through Stroud, do you use the permitted routes between Reading and Stroud, then Stroud and Worcester, or is this route entirely invalid?

You consider the routes:
- Reading to Evesham, and Evesham to Worcester (joining these up at Evesham); or
- Reading to Stroud, and Stroud to Worcester (joining these up at Stroud).

You do not consider Reading to Worcester.

 

[Benjwri]

You consider the routes:
- Reading to Evesham, and Evesham to Worcester (joining these up at Evesham); or
- Reading to Stroud, and Stroud to Worcester (joining these up at Stroud).

You do not consider Reading to Worcester.

Ahh okay thanks, in that case is it just shortest routes from each? Or do you consider the routeing points and subsequent mapped routes? Doesn't make a difference here but could imagine it might in other circumstances

 

[alistairlees]

Ahh okay thanks, in that case is it just shortest routes from each? Or do you consider the routeing points and subsequent mapped routes? Doesn't make a difference here but could imagine it might in other circumstances

As you would consider any journey for "Any permitted" - routeing points, nfm64, maps, distance etc.

 

[yorkie]

Ahh okay thanks, in that case is it just shortest routes from each? Or do you consider the routeing points and subsequent mapped routes? Doesn't make a difference here but could imagine it might in other circumstances

The permitted routes are calculated independently. Booking sites should handle this.

This is the sort of concept we can cover in fares workshops, but usually only at advanced fares workshops. The rules in this area are not very well known and indeed not all pricing managers realise this, so does lead to anomalies in some fares; if posted here, the anomalies would disappear, so I will refrain from divulging them!

 

[island]

You consider the routes:
- Reading to Evesham, and Evesham to Worcester (joining these up at Evesham); or
- Reading to Stroud, and Stroud to Worcester (joining these up at Stroud).

You do not consider Reading to Worcester.

I don't believe this is correct. My understanding is that as Reading to Worcester has a mapped route via Evesham, you use that. It is only because Reading to Worcester does not have a mapped route via Stroud that you would separately consider Reading to Stroud and Stroud to Worcester.

 

[yorkie]

I don't believe this is correct. My understanding is that as Reading to Worcester has a mapped route via Evesham, you use that. It is only because Reading to Worcester does not have a mapped route via Stroud that you would separately consider Reading to Stroud and Stroud to Worcester.

I can confirm that what @alistairlees says is correct in terms of combining the permitted routes; the issue is that this should be better documented.

The permitted routes are any route which is a permitted route for the throughout journey which go via the via point (if any) plus the routes formed by the split route check (i.e. permitted routes from origin to via point and from via point to destination)

 

[Watershed]

I don't believe this is correct. My understanding is that as Reading to Worcester has a mapped route via Evesham, you use that. It is only because Reading to Worcester does not have a mapped route via Stroud that you would separately consider Reading to Stroud and Stroud to Worcester.

I believe @yorkie has it right; both approaches are used to determine permitted routes. The Routeing Guide refers:

Where the fare specifies a particular route, there will be restrictions in the fare routes data that will affect thee [sic] routes listed in the Guide which pass through the station shown in the route description. The National Routeing Guide may be used to find out how to reach the station shown in the route description.

The word 'may' makes clear that the split route check is one option, but you are not compelled to use it. You can thus use both the permitted routes:

• from the origin to the destination, that pass through the via point(s), or
• from the origin to the via point(s), then between the via points if there's more than one, then from the last via point to the destination (i.e. the split route check),

whichever suits the passenger best.

There probably won't be a huge number of cases where the first approach is materially more generous than the second, but to give one example, walkup fares from Sheffield to Chester routed 'via Stockport' can only be used via Warrington through the first approach. If you purely determined permitted routes using the 'split route check', travel via Warrington would not be permitted.

 

[ThisIsSurbiton]

It is only because Reading to Worcester does not have a mapped route via Stroud that you would separately consider Reading to Stroud and Stroud to Worcester

Interesting! The formerly top secret guide sent to developers of routeing guide implementations, which I think is about to be included in the public guide to the data format RSPS5047 says:

If a routed fare is found to be invalid, then the journey can be split at the via locations contained in the route descriptor. Each part of the journey must follow a permitted route. If all part journeys are permitted, then the routed fare is valid, and the journey is deemed to be “split for routeing”.

I read this as saying that it's routeing guide for the overall journey OR separately valid journeys meeting at the split, as @island says

 

[Andrew1395]

All permitted routes have to follow the conditions of the fare. So a ticket which has an instruction via Evesham. Your ticket must include travel through Evesham.

Simplistically, journey planners calculate every possible combination of journeys in the timetable then apply the fare and routeing permissions to edit them down to the valid ones, before presenting itineraries to the customer.

Sometimes price means that routed fares can be more expensive than any permitted fare. So then that also needs to be taken into consideration.

the key principle is that the conditions of the fare is the starting point for its validity on any journey. So VIA, Not VIA, TOC specific, Mode specific, etc instructions are taken into consideration first.

if you look at the wording to Journey planners it says “ …the route descriptor associated with the fare may contain locations through which the the journey must pass..”

so if doing the normal Routeing Guide checks Reading to Worcester (mapped journey), they did not include map permissions for travel through Evesham (a requirement of the fare) - And therefore at that point invalid - You go on to do the journey splitting. You calculate the permitted routes from origin to Evesham then add on the permitted routes from Evesham to destination, to find a valid itinerary.

 

[JB_B]

There are some disagreements about how to approach this. I think that partly depends on whether you're talking about what the human-facing routeing guide specifies (or intended) or what a journey planner should allow - the latter is now the de facto standard but both are important.


If a ticket from S to T has a geographical restriction "via X" then you potentially need to consider two sets: A and B where

A is those permitted routes from S to T - considered as a whole - which pass through X; and

B is the product of the permitted routes from S to X combined with the permitted routes from X to T.

In each case the permitted routes must meet any other requirements of the ticket and would be the usual trio of: direct trains, shortest routes (+3m) and mapped routes (subject to any relevant easement.)


For any given potential journey, journey planner developers are effectively told to check if the journey is in A and, if it isn't, see if it's in B.

I think that the net effect of that instruction is that the overall set permitted routes is always the union A of B.

Sometimes A and B are not the same ; nor is one necessarily a subset of the other.

However, some interpreters of the human-facing routeing guide take the view that the permitted routes are just A and only if A is empty do you substitute the routes in B. I think that is @island 's position expressed in post #6. ( Apologies if I've misunderstood.)


It's possible that @alistairlees in post #2 is taking a different view again - you could take that post as saying "you only [ever] consider [set B]" and never consider A but given that set A is empty in this particular case (at least for X = STROUD ), it's potentially also compatible with either view above.

 

[Paul Kelly]

However, some interpreters of the human-facing routeing guide take the view that the permitted routes are just A and only if A is empty do you substitute the routes in B. I think that is @island 's position expressed in post #6.

It was also my position but I've just re-read the routeing guide and now I am more on the fence. The routeing guide used to say:

Where the fare specifies a particular route, you may choose only those routes listed in the Guide which pass through the station shown in the route description.

I would have thought this was fairly unambiguous, i.e. it disallows "split via routeing" unless there are no permitted routes at all via the the station in the restriction.

But since (according to farehistory.info) 26 August 2015, the above sentence has been amended to read:

Where the fare specifies a particular route, there will be restrictions in the fare routes data that will affect the routes listed in the Guide which pass through the station shown in the route description.

which is rather waffly and hard to interpret!

 

[Cantab26]

This is actually one of the simpler nuances of the routeing guide.

In summary, when travelling from A to B using a routed fare (“via X”), the passenger may:

1. Use any permitted route from A to B which passes through X
2. Use any permitted route (provided for in the routeing guide) from A to X and then X to B
3. Use any other permitted route from A to B which does not travel via X provided that the Any Permitted fare or a routed fare (“via Y”) permits travel via that route AND that fare is the same price or lower than the via X fare
4. In case 3., if the same or lower price fare is routed “via Y”, use any permitted route (provided for in the routeing guide) from A to Y and then Y to B

 

[Watershed]

This is actually one of the simpler nuances of the routeing guide.

In summary, when travelling from A to B using a routed fare (“via X”), the passenger may:

1. Use any permitted route from A to B which passes through X
2. Use any permitted route (provided for in the routeing guide) from A to X and then X to B
3. Use any other permitted route from A to B which does not travel via X provided that the Any Permitted fare or a routed fare (“via Y”) permits travel via that route AND that fare is the same price or lower than the via X fare
4. In case 3., if the same or lower price fare is routed “via Y”, use any permitted route (provided for in the routeing guide) from A to Y and then Y to B

To paraphrase Sir Humphrey, there is some difficulty in justifiably assigning the epithet "simple", inasmuch as this lays a heavier burden upon the logical and semantic resources of the English language than they can reasonably be expected to bear.