Fares oddities

[Southern]

Folkestone Stations - Brighton £19.10
Folkestone Stations - Chichester £19.10
Folkestone Stations - Havant £43.80
Folkestone Stations - Fareham £18.20
Folkestone Stations - Portsmouth Stations £20.00
Folkestone Stations - Salisbury £20.00

All these tickets are Off-Peak Day Returns.

 

[rjhglover]

Two Examples:

Stamford - Lincoln: £28.90

Stamford - Peterborough: £7.00
Peterborough - Lincoln: £13.90
Total: £20.90

Peterborough - Birmingham New Street: £31.40

Peterborough - Stamford: £7.00
Stamford - Birmingham New Street: £18.50
Total: £25.50

Both Off-Peak

 

[yorkie]

Bear in mind ATOC may be reading this!!

So no, I am not going to put any tickets that are clear anomalies here that could be useful to me

The Havant anomaly is because it is routed Not London meaning it is highly likely to be valid through the outskirts of London, and is therefore set by SET and is expensive.

The combination of tickets being cheaper than a through ticket in EMT land are not really anomalies (an anomaly, in my opinion, is where it is cheaper to go further) and there are numerous examples where it is cheaper to do this (ie, split the journey into multiple tickets).

But when it is cheaper to go further, that's where you can really say it's an anomaly, I think.

 

[wibble]

Folkestone Stations - Brighton £19.10 Priced by Southern
Folkestone Stations - Chichester £19.10Priced by Southern

Folkestone Stations - Havant £43.80Priced by Southeastern

Folkestone Stations - Fareham £18.20Priced by Southern

Folkestone Stations - Portsmouth Stations £20.00Priced by Southern

Folkestone Stations - Salisbury £20.00Priced by Southern

Southern have cut a lot of prices on their South Coast route:
Southampton Ctl - Worthing CDR £15.00 SVR £22.40
Southampton Ctl - Brighton CDR £15.00 SVR £22.40
Southampton Ctl - Eastbourne CDR £15.00 SVR £22.40
Southampton Ctl - Hastings CDR £15.00 SVR £22.40
Southampton Ctl - Ashford Intl CDR £15.00 SVR £22.40

Portmouth - Brighton CDR £14.90 SVR £22.40......

 

[wintonian]

There are several examples if you play around with the routing guide and the NFM.

But unfortunately I can see this thread serving to increase the cost of rail travel when the right people read it - there is a PM system.

 

[laird]

It does seem to be a common factor in these cases that where the fare setting operator differs so does the pricing structure.
Now it isn't confusing for the average customer because you turn up ask for a ticket to Havant and either buy the ticket or walk away. The problem from the railway viewpoint must be the loss of revenue.
If the fares were balanced so Folkestone-Havant takes the Portsmouth fare passenger numbers to Havant might rise, would this be sufficient to offset reduced price paid by each ticket holder.
It is a shame that we feel ATOC might take the view that the best solution is to increase all the fares to the higher level, perhaps if there was a way to challenge these anomolies we would see fewer of them next year?

 

[wellhouse]

I can't help wondering how many anomalies arise from the 'Basket of Fares' concept applied to the RPI (Retail Price Index) formula for permitted fares increases. It may well be in the interest of a TOC to hold or even reduce fares on certain flows if that allows higher than average increases on busier routes elsewhere.

 

[scotsman]

Please don't post these - there was much discussion on the ScotRail forum on one particular route's fares and ATOC went to great lengths to make sure that nobody could use the tricks.

 

[laird]

It does seem a strange idea that ATOC would gather their information on fares anomolies on public forums, surely they could develop a system for identifying split ticketing opportunities.
It should be even easier for them to pick out these anomolies from within the data they already hold where one station combination on a route is set by one operator where all the others are connected to a second?

This example probably came about when Southern reviewed their fares along the South Coast route, in bringing down their prices to increase volume of sales where operators like South Eastern who were probably increasing by RPI would end up with a greater differential in pricing, although it should now be stabilising as both generally increase to the future.

Thinking about it if we had a price per mile anomolies would dissapear but we know that would lead to some routes being overcrowded so I wonder whether now like the yield management concept used in the airline business if there could not be a flat very low price per mile charged but each individual train would be subject to an overcrowding charge with the aim of pushing the load throughout the day. It would have to be based on every station pair on each individual train I guess but at least these oddities would be a thing of the past.

 

[dzug2]

Thinking about it if we had a price per mile anomolies would dissapear but we know that would lead to some routes being overcrowded so I wonder whether now like the yield management concept used in the airline business if there could not be a flat very low price per mile charged but each individual train would be subject to an overcrowding charge with the aim of pushing the load throughout the day. It would have to be based on every station pair on each individual train I guess but at least these oddities would be a thing of the past.

Not really - they'd be replaced by other oddities and anomalies like differing London Birmingham prices according to which of the at least 6 possible routes you used (Actually there's more than 6 - 9 or 10 more like)

 

[Deerfold]

It does seem a strange idea that ATOC would gather their information on fares anomolies on public forums, surely they could develop a system for identifying split ticketing opportunities.

This is a non-trivial task - you'd be suprised how hard it is to do with a computer program. This is why no-one's come up with a computer program designed to save you money on this method.

People on here are good at suggesting splits by having knowledge of where good places to split are.
--- old post above --- --- new post below ---

Thinking about it if we had a price per mile anomolies would dissapear but we know that would lead to some routes being overcrowded so I wonder whether now like the yield management concept used in the airline business if there could not be a flat very low price per mile charged but each individual train would be subject to an overcrowding charge with the aim of pushing the load throughout the day. It would have to be based on every station pair on each individual train I guess but at least these oddities would be a thing of the past.

So automatically, the busier your train the more you'd pay? I can see that being popular - and also a way of saying how very rural routes aren't paying their way as people would pay less per mile for them.

 

[laird]

Deerfold, I agree the split ticketing finder seems to be illusive, there are some well known locations that recur round the country of course. Indeed it probably is in the interest of ATOC and the rail companies to perpetuate this.

The idea would be that the cheaper trains would be those with seating capacity, some would be rural routes but as the price would be very low for these they will hopefully tempt the car driver into the train because the cost of rural motoring we are being told is so high. It does of course rely upon perfect economic theory but since it seems to work to keep the loads high in the airline market perhaps it could be true for the railways too.
Unless we switch to some form of electronic ticketing it wouldn't work easily so it is something of a dream.

Perhaps a simple step that ATOC could take is to look at which company is setting the fare for each route, that would be a quick way of getting rid of the irregularity of the Folkestone to Havant having ended up with South Eastern rather than perhaps the more logical Southern. To find those shouldn't be quite so bad and actually could be in the interest of ATOC.

 

[penaltyfines]

The reason why there's no automated system for calculating fares is that every possible permutation matches the Travelling Salesman Problem (TSP) [http://en.wikipedia.org/wiki/Travelling_salesman_problem] which is NP-hard.

To give you an idea, 10 stations, there are 3 628 800 possible ticket combinations that pass through all 10 points. Given the number of stations on the network, and the number of stations on or around a given route, calculating all possible purmutations would require masses of computation time.

 

[b0b]

This is a non-trivial task - you'd be suprised how hard it is to do with a computer program. This is why no-one's come up with a computer program designed to save you money on this method.

Part of the problem is the information required is not in a readily usable format for this kind of thing, unless you're willing to pay 600 quid for a data extract from ATOC.

For example, to find fare splits you need to know:

(a) which train(s) you are considering;
(b) all of the stations that the train calls at
(c) fares information for all of the potential split points

and that doesn't even cover some other tricks as using rovers and/or season tickets where even more information is required, for example, the intermediate stations a "fast" train passes through.

I and some other people have been investigating potential sources that can be manipulated into the form required to figure this stuff out, but it's been awful.

 

[Southern]

Sorry if I have caused any offense, didn't realise that ATOC read transport forums.

 

[Oracle]

They do in the same way the TfL managers read District Dave's. And also the erstwhile Routemaster Forum that was closed down some years ago.

 

[mallard]

Part of the problem is the information required is not in a readily usable format for this kind of thing, unless you're willing to pay 600 quid for a data extract from ATOC.

For example, to find fare splits you need to know:

(a) which train(s) you are considering;
(b) all of the stations that the train calls at
(c) fares information for all of the potential split points

and that doesn't even cover some other tricks as using rovers and/or season tickets where even more information is required, for example, the intermediate stations a "fast" train passes through.

I and some other people have been investigating potential sources that can be manipulated into the form required to figure this stuff out, but it's been awful.

For a while I've been pondering the idea of writing a program that would go through the fares database and sort the various walk-up tickets by £-per-mile (crow-flies distance would be easiest by far). All that would be needed is a way to read Avantix Traveller's database files and a list of geographical positions of stations...

The tickets that show up with low prices on this list would then be (in theory) good candidates for use as part of a splitting scheme.

 

[Paul Kelly]

All that would be needed is a way to read Avantix Traveller's database files and a list of geographical positions of stations...

http://librailfare.sourceforge.net/ and http://www.ordnancesurvey.co.uk/meridian2/

 

[mallard]

Look's like I've found something to do this weekend then!

 

[Mojo]

I disagree that the reason for nobody being able to develop a solution to cheaper rail fares is wholly down to the "travelling salesman problem," although that's part of it.

In order to do it, you'd need to:

• Buy a link to the fares and timetable database from Atoc (costs a fortune)
• Get the user to specify the date and time for journeys they are undertaking where Apex tickets are available or where different trains call at different stations (most routes)
• Take into account the fact that you can often get a cheaper ticket by buying a ticket for a longer distance than the journey you are actually undertaking
• Look at including rovers, rangers and PTE products
• Bear in mind that it can sometimes be cheaper to be "doubly valid"

 

[yorkie]

Sorry if I have caused any offense, didn't realise that ATOC read transport forums.

Don't worry, you certainly didn't cause offence.

ATOC probably don't go round reading transport forums, in general, but they do read this forum (to what extent is unknown).

Your post basically demonstrated that one particular fare is set far higher than fares to similar destinations. In this case, it is clear that the fare to Havant is far too high. This is because it is assumed that travel will be via a longer route (ie, not via the coast) and there is no routeing (such as Rte Barnham) for this particular flow that are comparable with the other flows. The fare is set by SET. It's basically an 'odd one out'. Avantix Traveller software is very handy in situations like this as it reveals who sets the fare and what the routeing is at a quick glance.

In general though it's unwise to post lists of tickets where it is cheaper to go further for a particular journey as the TOC(s) in question may read it and may increase fares or introduce restrictions. In this case there is one ticket that is too high, I see no problem revealing that.

 

[wibble]

I can't help wondering how many anomalies arise from the 'Basket of Fares' concept applied to the RPI (Retail Price Index) formula for permitted fares increases. It may well be in the interest of a TOC to hold or even reduce fares on certain flows if that allows higher than average increases on busier routes elsewhere.

That's not strictly true - the RPI+X% formula is the average increase applied to regulated fares. Some regulated fares may increase by more than RPI+X% as long as the basket doesn't exceed the limit.
--- old post above --- --- new post below ---

In general though it's unwise to post lists of tickets where it is cheaper to go further for a particular journey as the TOC(s) in question may read it and may increase fares or introduce restrictions. In this case there is one ticket that is too high, I see no problem revealing that.

TOCs should be keeping an eye out for inconsitencies in prices, like those mentioned above, but as the journey priced by SET is not one of their core journeys, I don't think they'll be in a rush to correct it!

 

[yorkie]

TOCs should be keeping an eye out for inconsitencies in prices, like those mentioned above, but as the journey priced by SET is not one of their core journeys, I don't think they'll be in a rush to correct it!

I believe Southern could apply to ATOC to introduce a Route Barnham ticket for that flow. That would be a sensible solution, but as you say they may not be in a rush.

 

[laird]

Rather than waiting for an operator to introduce a fare, why can't we have a scheme to propose routed fares like this? Or does such a scheme exist?

 

[wibble]

I believe Southern could apply to ATOC to introduce a Route Barnham ticket for that flow. That would be a sensible solution, but as you say they may not be in a rush.

ATOC have no involvement in the creation of fares. Southern could create a route Barnham fare if they wished and should, as a courtesy, let SET know it was happening.
--- old post above --- --- new post below ---

For a while I've been pondering the idea of writing a program that would go through the fares database and sort the various walk-up tickets by £-per-mile (crow-flies distance would be easiest by far). All that would be needed is a way to read Avantix Traveller's database files and a list of geographical positions of stations...

The mileages between stations are already publically available in the National Rail Timetable... Mileage based on 'crow fly' distances doesn't actually reflect the cost the journey being made.

 

[mallard]

The mileages between stations are already publically available in the National Rail Timetable... Mileage based on 'crow fly' distances doesn't actually reflect the cost the journey being made.

In some cases, probably not. However, to calculate the straight-line distance between any two stations, all I need are their geographical co-ordinates, I don't need to know anything about routes or the rail system at all.

All my program would do is show how to get the furthest for the least amount of money. Tickets that show up high in that list may well be good candidates for use in splitting, cutting short, etc.

One example of such a ticket is the well-known East Midlands-North Wales "off peak" (actual restriction is after 0730 I believe) ticket that is valid over a large portion of Greater Manchester and Merseyside.

 

[MKB]

On Southern's website, they are showing London Bridge to Gatwick Airport as:

£7.90 - Standard Anytime Day Single (Southern Trains only)
£8.00 - Standard Off-Peak Day Single (Southern Trains only) after online 10% discount

Why would the Off-Peak cost more than the Anytime on a ticket in the same class and with the same routeing restrictions and with fare set by the same TOC?

 

[wintonian]

Avantix says

£7.90 SDS

The only £8 single I can find is a; SN SUPER GROUP which has restriction SV

Restriction : SV
Valid on Southern
(TOC CodeN) services only
unless stated on the ticket.
ADVANCE & ADVANCE FIRST CLASS

Ticket:
Ticket Class: STANDARD
Ticket Name: ADVANCE
Single/Return SINGLE
Validity:-
Date and time shown on
accompanying reservation.
Break of Journey: NO (See
notes below)

Booking and Reservations:
Ticket must be booked and
seats reserved before
departure.
Tickets sold subject to
availability.
Not available on the day of
travel.

Reservations Compulsory: YES
Child Discount: YES
Railcard Discount: (See notes
below)

 

[island]

The £8 ticket is for two adults. There's an equivalent one for five people for £15 and for ten people for £20. See http://www.southernrailway.com/offers/supergroup. They are advance tickets, not off-peak day singles.

 

[MKB]

Since the mysterious £8 off-peak day single ticket on the Southern website is after a 10% online discount, the non-discounted fare must be circa £8.89. This ticket doesn't show up on other booking engines.