Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
I apologise if this has been done before, but I can’t find any other threads that attempt to do what I’ve been looking into. My aim has been to analyse all possible ticket combinations between UK stations, and see the effects that split ticketing has on the overall statistics. It’s been an interesting challenge, and I think the results should probably be published somewhere properly, if I can get around to it. A byproduct of it is a database of all optimal splits for walk-up fares from any station to another. I can make this publicly available (if I’m allowed) if people would like to use it.
I’ve done the analysis using the available static data from the National Rail Data Feeds, using Off-Peak and Anytime fares only and valid as of the 5th July. For this post I’ll just look at the cheapest standard-class fares between stations.
This eliminates one of the main reasons for splitting, which involves departing at a time when either a (Super) Off Peak isn't valid, or the Railcard held isn't valid. This is an addition to another major reason: the unavailability of through Advance fares.
How would this work for Two Together Railcard holders travelling before 09:30, or what about three people travelling together for journeys involving a mixture of TOCs, such as Southeastern + LNER, etc? What about a Network Railcard holder travelling from Brighton to Peterborough, and how would you ensure that the resultant "through ticket" ensures travel on a train that calls at Huntingdon?
These are tricky. Technically it’s Camarthen-Aberystwyth, which is £91.20 single, but it can be split by using a bus to be £9. I’m not sure if this counts. If not, then it’s the same as the absolute ones.
If it does count, how would buses be brought into this, given there are many buses that could be used to supplement rail journeys all over Great Britain?
Either the component prices need to rise, or there is going to be a big financial shortfall (currently paid by people who are prepared to pay the higher price), so it could only happen if rail subsidy increased.
However, it seems unlikely if subsidies would increase, therefore this would only happen if shorter distance fares are increased in price, which would be political suicide for any party that implemented such a measure.
Without increased subsidy, it would be literally taking money from people who are doing shorter journeys, people generally from poorer backgrounds, people less likely to be on expenses, and handing that money to people with a greater willingness to pay for premium journeys, businesses who pay for their employees, etc. A very unpopular move, I would say!
This could be taken a step further by reducing any tickets using the route of said advance train by the same amount, to ensure completely fair system. One can hope.
I don't know what you mean by this, but I am not sure how any system could be "completely fair", though any attempts to make any system "fairer" by increasing the cost of currently popular good value fares would be deeply unpopular.
== Doublepost prevention - post automatically merged: ==
Although I'm one of them, I don't feel some people should get cheaper train tickets just because they know how the system works, it should just be easy and fair for all. I wouldn't propose that anything increases in price at all.
In that case, your plan requires an increase in rail subsidy, which is not likely to be forthcoming any time soon, so this is very much a case of a very long term strategy, and unlikely to ever happen.
== Doublepost prevention - post automatically merged: ==
Passengers can, by using a site such as the forum's ticketing site, but train companies such as LNER, Avanti etc are putting taxpayers funding into advertising and/or schemes to encourage people away from the sites that do that.
I asked TOC financial director (while on a train journey a couple of years or so ago) that split ticketing grew revenue and passenger numbers by market segmentation, and the response was yes he thought that it did indeed do that.
To undo that would be very difficult and the money would have to come from somewhere where it currently simply doesn't exist.
It's a shame that split ticketing is quite so necessary in this country. I've certainly done it abroad as well, but I find in GB that pretty much any medium to long distance journey is cheaper, often a lot cheaper, by splitting.
It's also absolutely necessary in France, Italy and many other places too; if the proposal is for a similar system here, then it would involve the abolishment of many longer distance fares (or even shorter ones that straddle journeys that are considered to be different 'sectors' / 'categories')
== Doublepost prevention - post automatically merged: ==
It doesn't. It gives, quite simply, the cheapest combination of (flexible) splits - that is what it does. And, as you admit, it doesn't mean the splits are even usable in a practical sense, because a lot of them will tie you to 500 miles on stopping trains so you call at Bere Alston, Ashchurch for Tewkesbury, Stone, Blackrod etc etc all the way up to Aberdeen.
This is quite limited for start, as most people splitting are doing so because they are price sensitive and therefore would consider Advances, which are quota controlled and train-specific.
While I agree that in an ideal world people shouldn't be advantaged simply because they are power users of a public good, we have the system we have, and I do not think this tool would bring big-picture fairness in the way you think it does.
What do you think the practical upshot of the industry knowing the cheapest combination of splits for every single station pair would be? That all the long distance fares would simply go down to combine all the splits into a through ticket, and the railway lose revenue? Or, as has been the case since the 90s, would many popular fare flows used for splitting simply rise to prevent creative use, making other people's short, unsplittable journeys more expensive?
For longer distance journeys, most journeys are now on Advance tickets.
This eliminates one of the main reasons for splitting, which involves departing at a time when either a (Super) Off Peak isn't valid, or the Railcard held isn't valid. This is an addition to another major reason: the unavailability of through Advance fares.
Thanks for all this, it's rather enlightened me on the complexity of railcard restrictions! I really didn't mean to imply anything beyond just the basic statistical analysis. Perhaps by saying 'undiscounted' to begin with I'd have avoided most of your criticisms... I do have a method for calculating fair fares for when there is a change of 'discount' during the journey (as in, a railcard/off peak tickets become valid), or using advance fares, but it's quite involved, and goes beyond the scope of the analysis I've posted here. Obviously you're quite correct that all of this increases the utility of splitting even further, but it doesn't invalidate what I've quoted for undiscounted, off-peak fares. I'm quite happy to have a debate about how to expand a 'fair' system further, but only from a mathematical point of view -- that's where my interest lies rather than discussing whether it's politically feasible or not.
What about a Network Railcard holder travelling from Brighton to Peterborough, and how would you ensure that the resultant "through ticket" ensures travel on a train that calls at Huntingdon?
I may as well explain my solution for this example, as it's a nice one, though (as I say) it's beyond the scope of my original post. Also, I had no idea a Network Railcard exists?! How on Earth is that a thing but there's no equivalent in any other urban areas?! (This is an aside, and I don't want to get carried away).
Cheapest Off-Peak (split) fare BTN-PBO: £41.70 (split at GTW, HIT if you're interested)
Cheapest Off-Peak (split) fare BTN-HUN: £37.60 (GTW, HIT, BIW)
Equivalent with Network railcard: £24.85 (I've made some rounding assumptions here but it shouldn't be far off)
Fair price = £41.70 - £37.60 + £24.85 = £28.95
There would be no need to call at Huntingdon -- that kind of restriction is just an uneccessary artifact of needing to split, and serves no benefit to either the passenger or the railway.
So you're essentially just paying the extra undiscounted £4.10 to get to Peterborough, which is cheaper than the £6.50 single between them you'd have to pay currently. Hopefully that makes some sense - please feel free to pick holes in the mathematical process (but don't just say "we could never get it to work like that" or equivalent).
Similarly it can be done with your other examples, but you have to make a call at which location your train stops being 'peak'. Based on a normal weekday timetable, and taking the fastest trains you can (which means the most expensive tickets, as you travel further under peak restrictions), the 'fair' prices are (I may have got something wrong here, naturally, or missed a discount I didn't know about. The Southeastern Kids thing was news to me, but I've included that):
Brighton to York at 07:00 (2 adults with two together railcard) - £68.05
Nottingham to Stevenage at 13:00 (3 adults travelling together) - £64.80 (Did I miss something here -- I don't think there are any valid discounts?)
Ashford to Long Buckby at 08:00 (1 adult with Senior Railcard) - £42.05
Ashford to Rugby at 08:00 (1 adult with Senior Railcard) - £42.40
Milton Keynes to Dover at 10:00 (1 adult and 1 child) - £52.35
If it does count, how would buses be brought into this, given there are many buses that could be used to supplement rail journeys all over Great Britain?
As far as I can tell, it's only very few bus services in West Wales that can be bought with a normal Anytime Single (eg. Camarthen-Bow Street). I've just included them because it would be a pain to filter them out. Part of the above journey is by train, which prevents it being an 'Anytime Bus Single', which I've not included.
Similarly to the above example (the BTN-YRK one), the opposite could be applied to Advance fares, allowing for 'through' advance tickets to practically anywhere. Hypothetically, say you wish to travel from Marlow-Hexham (Advances do exist for this combination, I am aware, but ignore that for now). Off-peak this is £102.40. The 'Advance' section is KGX-NCL, which is £89.00 off peak (both with no splits). I'd then propose you must be able to get an advance single from Marlow-Hexham for the price of any advance KGX-NCL, plus the difference in these fares (£13.40). As it stands, sometimes such fares do exist, sometimes they don't.
In that case, your plan requires an increase in rail subsidy, which is not likely to be forthcoming any time soon, so this is very much a case of a very long term strategy, and unlikely to ever happen.
Sure, I get your point. But you could say the same about offering any kind of discounts. The Southeastern 'kids for a quid' promotion used above surely requires such an 'increase in rail subsidy' by the same logic, and I don't recall that being fiercely debated in parliament. I'd argue that the benefits of a simpler, fairer system would draw more people to the railways and the changes could pay for themselves. But such things are above my pay grade... I just do numbers.
One of the main problems I have with the current system (and of course I use splits for every long-distance journey, pretty much) is in the case of disruption (about every other long-distance journey, pretty much...). I recently had a journey with a leg Warrington-Carlisle, split at Oxenholme on advances. I missed my connection to get this train: the next one to Carlisle was in 20 minutes but the next one to Carlisle stopping at Oxenholme was in 2.5 hours. Luckily the guard was understanding and I just got on the first one with no issue. But I think you'd struggle to argue that it's advantageous to anyone that 'by the book' I'd have been 3 hours late to my destination as opposed to half an hour. I should have just been able to buy an advance for the whole journey for the same price. Do you disagree?
Also, I had no idea a Network Railcard exists?! How on Earth is that a thing but there's no equivalent in any other urban areas?! (This is an aside, and I don't want to get carried away).
You can get 1/3 to 50% off your train fares with a Regional Railcard, which gives you discounts in a specific area.
www.nationalrail.co.uk
It looks like you are doing a basic number crunching exercise and saying "look, it's simple to literally guarantee the best available fare" without any knowledge of how tickets work in practice or the complexity of the system. The sheer complexity of the system is why not even any third party retailer has guaranteed "the best available fare". The forum site will split your tickets but it will not ever guarantee you can't find a better combination. Have you incorporated day rangers etc into the splitting algorithm? What about multimodal tickets? How about overdistance ones?
There would be no need to call at Huntingdon -- that kind of restriction is just an uneccessary artifact of needing to split, and serves no benefit to either the passenger or the railway.
You can't just sneak in a bit of fares reform to make things work and assert that NRCOT 14.2 serves no benefit to the railway. Why do you think the railway - who set the passenger contract - have this in the conditions?
Sure, I get your point. But you could say the same about offering any kind of discounts. The Southeastern 'kids for a quid' promotion used above surely requires such an 'increase in rail subsidy' by the same logic, and I don't recall that being fiercely debated in parliament.
Why do you assume those discretionary promotions are some sort of charitable gimme, and lose money? They're revenue positive - they get one or two adults on a train who otherwise would have driven their kids on a day out. That's the point of those promotions.
That's naive and not borne out by any sort of evidence. It's just an assertion that the enormous subsidy gap would be magically filled by a huge load of passengers (which the railway in many cases couldn't physically carry!) and it would all "magically pay for itself, so please take my proposal seriously". There are two main reasons why the railway hasn't already adopted your method of "fairer fares" - one, they would lose a ton of money the taxpayers would be asked to further subsidise, and two, they already have plenty of experience in what commercial levers already work.
One of the main problems I have with the current system (and of course I use splits for every long-distance journey, pretty much) is in the case of disruption (about every other long-distance journey, pretty much...). I recently had a journey with a leg Warrington-Carlisle, split at Oxenholme on advances. I missed my connection to get this train: the next one to Carlisle was in 20 minutes but the next one to Carlisle stopping at Oxenholme was in 2.5 hours. Luckily the guard was understanding and I just got on the first one with no issue. But I think you'd struggle to argue that it's advantageous to anyone that 'by the book' I'd have been 3 hours late to my destination as opposed to half an hour. I should have just been able to buy an advance for the whole journey for the same price. Do you disagree?
That's not how it works. To quote @Watershed who gave the clearest advice about this:
When using split tickets your right to take the next train in the event of disruption is the same as if you held a through ticket - as long as you allow enough time for changing trains, of course. You can also claim Delay Repay based on all of the tickets you hold.
I really didn't think I'd encounter this level of hostility over something I just did for a bit of fun, to have a look at some potentially interesting statistics and make some nice graphs. I'm not sure why anything I've said has been so offensive that it's deserved this kind of response (apart from a couple of people last week who said it was interesting). I am not trying to change the world here, just trying to have some fun with some interesting data. Very sorry for ruining your day.
Do you have the source/thread for "When using split tickets your right to take the next train in the event of disruption is the same as if you held a through ticket", as I've been consistently been told for years (on this site, and on retailers themselves) that this isn't the case, and unless told otherwise your trains still need to stop at the split points (as if you had a a through ticket with the restriction "train must stop at Oxenholme" etc.). I'd have saved a lot of money/worry otherwise...
As I said above, and throughout this post, this is all purely hypothetical, and just a mathematical exercise to try to find 'ideal' fares in a complicated system.
== Doublepost prevention - post automatically merged: ==
It looks like you are doing a basic number crunching exercise and saying "look, it's simple to literally guarantee the best available fare" without any knowledge of how tickets work in practice or the complexity of the system. The sheer complexity of the system is why not even any third party retailer has guaranteed "the best available fare".
No. I've done a reasonably complicated number crunching exercise that nobody else appears to have done publicly, so I thought it would be interest. And I clearly stated it was for Anytime/Off Peak tickets and never mentioned any railcard discounts in the comparisons I made originally.
Indeed, and this line is particularly problematic due to the price difference between the slow/fast trains. I'm currently thinking about ways to incorporate such differentials properly into a unified system for my own curiosity, not to change the world. But I haven't done that, I've calculated it using the cheapest options available.
I'm sorry to be vaguely sarcastic in my reply, but you've really been quite harsh, I feel.
I really didn't think I'd encounter this level of hostility over something I just did for a bit of fun, to have a look at some potentially interesting statistics and make some nice graphs.
I think you implying that the railways "should" implement something like this is what got on the nerves of the long time regular users of the forum. I personally do agree that a fares reform would be quite unlikely to work, but I also think that they could have told you that in a kinder way
Do you have the source/thread for "When using split tickets your right to take the next train in the event of disruption is the same as if you held a through ticket", as I've been consistently been told for years (on this site, and on retailers themselves) that this isn't the case, and unless told otherwise your trains still need to stop at the split points (as if you had a a through ticket with the restriction "train must stop at Oxenholme" etc.). I'd have saved a lot of money/worry otherwise...
In your example you could have relied on your rights to rerouting under the PRO given the expected length of delay was over an hour - apologies, my advice was example-specific.
As I said above, and throughout this post, this is all purely hypothetical, and just a mathematical exercise to try to find 'ideal' fares in a complicated system.
== Doublepost prevention - post automatically merged: ==
But I think herein lies the nub of the issue - perhaps the mathematical model might be complex and you may believe it's "reasonably comprehensive", but I think you are labouring under the impression this is all much simpler than it is. If you go into this not knowing what Groupsave or a Network Railcard is, or not accounting for flexible tickets cheaper than the Off Peak becoming valid, or overdistance or zonal tickets (both *very* common methods of splitting!), or easements/negative easements, or the Routeing Guide - then I don't think you can claim this is in any way comprehensive or claim that people who have pooh-poohed this sort of exercise before weren't thinking hard enough.
If all you are doing is a mathematical analysis, then I am sure that has some value as a way of displaying your own competence in aggregating and displaying data - but you have posted this in the Fares section of the UK's biggest railway forum, so it is reasonable to expect some level of pushback. This is especially so if you say things like
shows that calculating such a thing takes a few minutes' computation rather than the standard line of 'this is so complex it would take aeons to unravel' line which is usually stated.
- well, it is incredibly complex and difficult, which is why nobody has *really* tried this before because doing so in a way to produce useful results is a truly enormous amount of work. There would be a fair bit of money to be had for a third party retailer who could do this - but none of them can.
and if you make your own proposals and repeatedly mention fairness, like
I’m a firm believer that if a cheaper walk-up split ticket for a journey exists, it is essential that this can be bought as a through ticket. The "fair" fare is easy enough to calculate (I’ve done it!) and so there’s no excuse for not doing so officially
I don't feel some people should get cheaper train tickets just because they know how the system works, it should just be easy and fair for all. I wouldn't propose that anything increases in price at all.
no need to call at Huntingdon -- that kind of restriction is just an uneccessary artifact of needing to split, and serves no benefit to either the passenger or the railway.
(proposing a reform to fares for the express purpose of making your solution work - 14.2 is there for good reason and it's because it often serves to protect the revenue of express operators, a clear benefit to the railway)
- then I think people are entitled to take the points up.
Perhaps this has reached the wrong audience in that there are a few people here who aren't "getting it" - data for the sake of data, but if it really is just that then I don't think you can supplement the findings with other commentary around changing the rules or claim the results reflect "fairness".
Literally the best thing about the UK rail scene is the enthusiastic work being done by very capable people with rail open data, so I do hope you are not put off by tough lines of questioning and it isn't my intention to be harsh. The fact is I can't do this sort of coding/programming work and can't criticise your efforts from a technical standpoint, but I do see how these pet projects can feed into better railway services for people to use and be positively disruptive. So I would agree with @walkingsophie's encouragement that you should do similar things, and if I have come across as harsh then I apologise for that. You may be able to make something of use with the project you have started!
In your example you could have relied on your rights to rerouting under the PRO given the expected length of delay was over an hour - apologies, my advice was example-specific.
This is good to know, though presumably it would involve finding station staff who are willing to follow the rules. Thanks.
My mistake, it seems, was naming the thread as such, as you are undoubtedly correct in that I haven't taken into account many of the splitting opportunities. I use trains a fair bit, and very much know my stuff about the journeys I regularly make. But I never travel at peak times, and very rarely in the South or with groups, so naively assumed my usage was representative. I have renamed the thread as such...
The algorithms behind Fastjp are about as good as it's possible to get, I'd say. I'd love to figure out how they manage such things so quickly. I've got my own code for calculating splits for my own use, which a few years ago would consistently beat Trainsplit etc., but does so rarely these days (and it's much slower). It's very a different task to looking at all the station pairs - which is inherently not going to much use for actually planning journeys, as I said.
Whilst eliminating splits for unrestricted, undiscounted tickets could be done within the current framework (and I maintain that it should be), eliminating splits due to railcards or peak restrictions would be mathematically impossible with how things currently work, hence my thought experiments in how to 'fix' that when asked. This is completely separate to the initial large analysis (I just thought people might be interested in my opinion).
Perhaps this has reached the wrong audience in that there are a few people here who aren't "getting it" - data for the sake of data, but if it really is just that then I don't think you can supplement the findings with other commentary around changing the rules or claim the results reflect "fairness".
Yes, that's fair enough. Out of interest, in a perfect world would you want to change the way any of the fares system works? You seem quite certain that everything is the way it is for a reason, and that's how it should stay. If I'm correct, then fine, but if not then why not engage with hypothetical solutions? If there's anything I've learnt while wading through all this data, it's quite how outdated, random and inconsistent it all is. I suppose the best I could hope for in the near future would be a GBR (o.e.) retailer selling well-calculated splits by default, and packaging them all into one ticket.
I heartily appreciate a tough line of questioning (if you've got a problem with the maths or any of the logic, I want to know, so I can try to fix it!) - I spend much of my working life engaging in such things, and give as much as I receive. But if I reacted to opinions I didn't like from people I don't know with statements like "That's naive and not borne out by any sort of evidence" (importantly without saying something nice first, which basically gives you a free pass) then I'd soon find myself out of a job, I fear.
I'll try to integrate usage into the raw statistics, and possibly make the data available somehow, but that's probably all I'll do on this for now.
== Doublepost prevention - post automatically merged: ==
I need to work more on my program that finds these, there is way too much missed potential there
I think you implying that the railways "should" implement something like this is what got on the nerves of the long time regular users of the forum. I personally do agree that a fares reform would be quite unlikely to work, but I also think that they could have told you that in a kinder way
Please do keep doing similar things, the railway world need more fun statics and graphs in my opinion
Thank you, I'm glad someone appreciates what I was aiming for.
I also think it's very unlikely to work, but this has basically confirmed my opinion (which is perhaps naive and definitely not borne out by any evidence) that the main reason for this is that people aren't willing to regard it as even possible...
I'm pretty sure I could figure out all the overdistance fares over the shortest route/(within 3 miles) using some of the code I've already written - I'll give it a shot. But it might be too slow, and more complicated routes would be beyond me.
This is good to know, though presumably it would involve finding station staff who are willing to follow the rules. Thanks.
My mistake, it seems, was naming the thread as such, as you are undoubtedly correct in that I haven't taken into account many of the splitting opportunities. I use trains a fair bit, and very much know my stuff about the journeys I regularly make. But I never travel at peak times, and very rarely in the South or with groups, so naively assumed my usage was representative.
To put rail usage into context, one sixth of all National Rail journeys are on the Elizabeth Line alone, and the overwhelming majority of daily journeys in the UK are made using just one ticket and are over distances of 25 miles or fewer. And therein lies the rub - that isn't data for the sake of data, but it tells us something meaningful about what the railway does, where a lot of its funding ought to go, and which sorts of journeys ought to be protected from negative change.
Whilst eliminating splits for unrestricted, undiscounted tickets could be done within the current framework (and I maintain that it should be), eliminating splits due to railcards or peak restrictions would be mathematically impossible with how things currently work, hence my thought experiments in how to 'fix' that when asked.
Yes, it's a massive undertaking. And it's not just peak restrictions - you also have super off peak fares to consider, plus a variety of other weird products like Evening Out Returns.
This is completely separate to the initial large analysis (I just thought people might be interested in my opinion).
Yes, that's fair enough. Out of interest, in a perfect world would you want to change the way any of the fares system works? You seem quite certain that everything is the way it is for a reason, and that's how it should stay.
Everything is the way it is for a reason. The problem is those reasons are out of date and were baked into the fares system in the mid 1990s and the system is not at all fit for purpose. What is fairly clear to me is that railway is not in a political position to demand large amounts of subsidy to just make travel cheaper, which means at best, any changes will need to be revenue neutral.
My general thesis however is that the underlying complexity of fares is not an inherent issue; it's how they are packaged and sold. This is a good blog on the matter: https://www.freewheeling.info/blog/complexity-isnt-the-issue-with-train-fares (though the argument about arbitrariness falls short of making sense).
People frequently cite air fares as being vastly more simple than rail fares, which is actually not true at all. Air fares are extremely complex, but the way they are sold is very simple; passengers do not need to apprise themselves of every single rule. Here is what the rules to a normal BA economy ticket between London and New York look like:
British Airways (BA) OLN8E1B4 LON to NYC
General notes
BASIC SEASON ECONOMY RT UNBUNDLED FARES O
APPLIES FOR ROUND TRIP FARES
FOR ADULT
TICKET DESIGNATOR - J615
Category 3: Seasonal restrictions
FROM AREA 2 -
PERMITTED 10SEP THROUGH 09OCT OR 28OCT THROUGH 11DEC OR
30DEC THROUGH 19MAR OR 11APR THROUGH 15MAY OR 01SEP
THROUGH 09SEP FOR EACH TRANSATLANTIC SECTOR.
TO AREA 2 -
PERMITTED 10SEP THROUGH 23OCT OR 03NOV THROUGH 25DEC OR
13JAN THROUGH 31MAR OR 19APR THROUGH 22MAY OR 08SEP
THROUGH 09SEP FOR EACH TRANSATLANTIC SECTOR.
Category 4: Flight restrictions
THE FARE COMPONENT MUST NOT BE ON
ONE OR MORE OF THE FOLLOWING
BA FLIGHTS 8200 THROUGH 8299.
AND
THE FARE COMPONENT MUST BE ON
ONE OR MORE OF THE FOLLOWING
ANY AA FLIGHT OPERATED BY AA
ANY AY FLIGHT OPERATED BY AY
ANY BA FLIGHT OPERATED BY BA
ANY IB FLIGHT OPERATED BY IB
ANY AA FLIGHT OPERATED BY AY
ANY AA FLIGHT OPERATED BY BA
ANY AA FLIGHT OPERATED BY IB
ANY AY FLIGHT OPERATED BY AA
ANY AY FLIGHT OPERATED BY BA
ANY AY FLIGHT OPERATED BY IB
ANY BA FLIGHT OPERATED BY AA
ANY BA FLIGHT OPERATED BY AY
ANY BA FLIGHT OPERATED BY IB
ANY IB FLIGHT OPERATED BY AA
ANY IB FLIGHT OPERATED BY AY
ANY IB FLIGHT OPERATED BY BA.
RESERVATIONS FOR ALL SECTORS ARE REQUIRED AT LEAST 28 DAYS
BEFORE DEPARTURE.
TICKETING MUST BE COMPLETED WITHIN 3 DAYS AFTER
RESERVATIONS ARE MADE OR AT LEAST 28 DAYS BEFORE DEPARTURE
WHICHEVER IS EARLIER.
NOTE -
DUE TO TTL ROBOT PROCESS DIFFERENCES MAY EXIST
BETWEEN THE CRS LAST TICKETING DATE AND TTL ROBOT
REMARK. THE INFORMATION BY SSR IN THE PNR
PREVAILS.
Category 6: Minimum stay requirements
TRAVEL FROM INBOUND TRANSATLANTIC SECTOR MUST COMMENCE NO
EARLIER THAN THE FIRST SUN AFTER ARRIVAL AT TURNAROUND.
AND - TRAVEL FROM TURNAROUND MUST COMMENCE NO EARLIER THAN
24 HOURS AFTER ARRIVAL AT TURNAROUND.
Category 7: Maximum stay requirements
TRAVEL FROM LAST STOPOVER MUST COMMENCE NO LATER THAN 12
MONTHS AFTER DEPARTURE FROM FARE ORIGIN.
Category 8: Stopover restrictions
STOPOVERS NOT PERMITTED ON THE FARE COMPONENT.
Category 9: Transfer restrictions
2 TRANSFERS PERMITTED IN EACH DIRECTION FREE
2 FREE ON 9B/AA/AY/BA/EI/GR/IB ONLY IN EUROPE IN EACH
DIRECTION.
AND - 2 TRANSFERS PERMITTED IN EACH DIRECTION FREE
1 FREE BETWEEN AA/BA/IB AND B6 IN BOS/JFK IN
EACH DIRECTION
1 FREE BETWEEN AA/BA/IB AND AS IN LAX/PDX/SEA IN
EACH DIRECTION
1 FREE BETWEEN AA/BA/IB AND WS IN CANADA IN EACH
DIRECTION
1 FREE BETWEEN AA/BA/IB AND PD IN CANADA IN EACH
DIRECTION
1 FREE BETWEEN AA/BA/IB AND HA IN HAWAII IN EACH
DIRECTION
2 FREE ON AA/BA/IB ONLY IN THE UNITED STATES/
CANADA IN EACH DIRECTION
FARE BREAK SURFACE SECTORS NOT PERMITTED AND
EMBEDDED SURFACE SECTORS PERMITTED ON THE FARE
COMPONENT.
Category 10: Combinability
APPLICABLE ADD-ON CONSTRUCTION IS ADDRESSED IN
MISCELLANEOUS PROVISIONS - CATEGORY 23.
END-ON-END
END-ON-END COMBINATIONS PERMITTED WITH DOMESTIC FARES.
VALIDATE ALL FARE COMPONENTS. SIDE TRIPS PERMITTED.
END-ON-END NOTE -
ADDONS PERMITTED
PROVIDED -
COMBINATIONS ARE WITH ANY ECONOMY OW UNBUNDLED/ECONOMY
RT UNBUNDLED-TYPE FARES FOR CARRIER BA WITHIN UNITED
KINGDOM IN RULE 7901 IN TARIFF
EURDBAR - WITHIN EUROPE-DOMESTIC
FBRA2P - WITHIN AREA 2.
END-ON-END
END-ON-END COMBINATIONS PERMITTED WITH FARES WITHIN
EUROPE. VALIDATE ALL FARE COMPONENTS. SIDE TRIPS
PERMITTED.
PROVIDED -
COMBINATIONS ARE FOR CARRIER BA IN RULE UPSL IN TARIFF
FBRA2P - WITHIN AREA 2.
OPEN JAWS
FARES MAY BE COMBINED ON A HALF ROUND TRIP BASIS
-TO FORM SINGLE OR DOUBLE OPEN JAWS.
A MAXIMUM OF TWO INTERNATIONAL FARE COMPONENTS
PERMITTED.
MILEAGE OF THE OPEN SEGMENT MUST BE EQUAL/LESS THAN
MILEAGE OF THE LONGEST FLOWN FARE COMPONENT.
PROVIDED -
THE OPEN SEGMENT MUST BE
WITHIN
-WITHIN AREA 1 OR WITHIN UNITED KINGDOM
COMBINATIONS ARE WITH ANY ECONOMY OW UNBUNDLED/ECONOMY
RT UNBUNDLED-TYPE FARES FOR CARRIER BA/AA/AY/IB BETWEEN
AREA 1/AREA 2 IN ANY RULE AND TARIFF.
ROUND TRIPS/CIRCLE TRIPS
FARES MAY BE COMBINED ON A HALF ROUND TRIP BASIS
-TO FORM ROUND TRIPS
-TO FORM CIRCLE TRIPS
A MAXIMUM OF TWO INTERNATIONAL FARE COMPONENTS
PERMITTED.
PROVIDED -
COMBINATIONS ARE WITH ANY ECONOMY OW UNBUNDLED/ECONOMY
RT UNBUNDLED-TYPE FARES FOR CARRIER BA/AA/AY/IB BETWEEN
AREA 1/AREA 2 IN ANY RULE AND TARIFF.
COMBINATIONS ARE WITH ANY FARE FOR ANY CARRIER IN RULE
UPSL IN TARIFF
FBRINPV - BETWEEN USA/CA-AREA 1/2/3
TAFPBAR - BETWEEN AREA 1-AREA 2/3 AND GUAM-AREA 2.
Category 12: Surcharges
IF INFANT UNDER 02 WITHOUT A SEAT.
OR - INE UNDER 02.
OR - INFANT WITHOUT A SEAT RESIDENT UNDER 02.
OR - INCLUSIVE TOUR INFANT WITHOUT A SEAT UNDER 02.
OR - TIMESAVER INFANT WITHOUT A SEAT UNDER 02.
OR - CONTRACT BULK INFANT WITHOUT A SEAT UNDER 02.
OR - MISSIONARY INFANT WITHOUT A SEAT UNDER 02.
OR - MILITARY INFANT WITHOUT A SEAT UNDER 02.
OR - PIF UNDER 02.
OR - FREQUENT FLYER INFANT WITHOUT A SEAT UNDER 02.
THERE IS NO CHARGE FOR TRAVEL.
THE PROVISIONS BELOW APPLY ONLY AS FOLLOWS -
SALE IS RESTRICTED TO SPECIFIC AGENTS.
OR - SALE IS RESTRICTED TO SPECIFIC AGENTS.
THERE IS NO CHARGE FOR TRAVEL.
THE PROVISIONS BELOW APPLY ONLY AS FOLLOWS -
TICKETS MUST BE ISSUED ON BA. SALE IS RESTRICTED TO
SPECIFIC AGENTS.
THERE IS NO CHARGE FOR TRAVEL.
THE PROVISIONS BELOW APPLY ONLY AS FOLLOWS -
SALE IS RESTRICTED TO SPECIFIC AGENTS.
OR - FARES MAY ONLY BE SOLD BY BA OR IB.
SALE IS RESTRICTED TO SPECIFIC AGENTS.
OR - FARES MAY ONLY BE SOLD BY BA OR IB.
SALE IS RESTRICTED TO SPECIFIC AGENTS.
THERE IS NO CHARGE FOR TRAVEL.
THE PROVISIONS BELOW APPLY ONLY AS FOLLOWS -
SALE IS RESTRICTED TO SPECIFIC AGENTS.
TICKETS MAY ONLY BE SOLD IN CHINA.
A SURCHARGE OF CNY 33 PER COUPON WILL BE ADDED TO THE
APPLICABLE FARE FOR TRAVEL.
ORIGINATING UNITED KINGDOM -
A SURCHARGE OF GBP 13.00 PER FARE COMPONENT WILL BE
ADDED TO THE APPLICABLE FARE FOR TRAVEL.
Category 15: Sales restrictions
Category 15: Sales restrictions
TICKETS MUST BE ISSUED ON BA OR AA.
TICKETS MAY NOT BE ISSUED BY PTA. TICKETS MUST BE ISSUED
BY ELECTRONIC TICKETING. EXTENSION OF TICKET VALIDITY IS
NOT PERMITTED.
OR - TICKETS MUST BE ISSUED ON BA OR AY.
TICKETS MAY NOT BE ISSUED BY PTA. TICKETS MUST BE
ISSUED BY ELECTRONIC TICKETING. EXTENSION OF TICKET
VALIDITY IS NOT PERMITTED.
OR - TICKETS MUST BE ISSUED ON BA OR IB.
TICKETS MAY NOT BE ISSUED BY PTA. TICKETS MUST BE
ISSUED BY ELECTRONIC TICKETING. EXTENSION OF TICKET
VALIDITY IS NOT PERMITTED.
OR - TICKETS MUST BE ISSUED ON BA OR EI.
TICKETS MAY NOT BE ISSUED BY PTA. TICKETS MUST BE
ISSUED BY ELECTRONIC TICKETING. EXTENSION OF TICKET
VALIDITY IS NOT PERMITTED.
OR - TICKETS MUST BE ISSUED ON BA OR PD.
TICKETS MAY NOT BE ISSUED BY PTA. TICKETS MUST BE
ISSUED BY ELECTRONIC TICKETING. EXTENSION OF TICKET
VALIDITY IS NOT PERMITTED.
Category 16: Penalties
CANCELLATIONS
ANY TIME
TICKET IS NON-REFUNDABLE.
NOTE -
FARE COMPONENT IS NON-REFUNDABLE.
----------------------------------------------
WAIVED FOR DEATH OF A PASSENGER AND PASSENGERS
TRAVELLING COMPANIONS.
----------------------------------------------
WHEN COMBINING NON-REFUNDABLE FARES WITH
REFUNDABLE FARES
1. THE AMOUNT PAID ON EACH REFUNDABLE FARE
COMPONENT IS REFUNDED.
2. THE AMOUNT PAID ON EACH NON-REFUNDABLE FARE
COMPONENT WILL NOT BE REFUNDED.
3. WHEN COMBINING FARES CHARGE THE SUM OF THE
CANCELLATION FEES OF ALL CANCELLED FARE
COMPONENTS.
----------------------------------------------
REFUND OF UNUSED TAXES FEES AND CHARGES PAID TO
THIRD PARTIES PERMITTED. ASSOCIATED CARRIER
IMPOSED CHARGES WILL NOT BE REFUNDED.
----------------------------------------------
REFUND PERMITTED WITHIN TICKET VALIDITY.
----------------------------------------------
ANY NON-REFUNDABLE AMOUNT FROM A PREVIOUS TICKET
REMAINS NON-REFUNDABLE FOLLOWING A CHANGE.
----------------------------------------------
TICKET IS NOT TRANSFERABLE TO ANOTHER PERSON.
----------------------------------------------
-------CANCELLATION REPRICING CONDITIONS------
FLOWN COUPONS MUST BE REPRICED USING HISTORICAL
FARES IN EFFECT ON THE PREVIOUS TICKETING DATE.
THE FARE FOR THE JOURNEY TRAVELLED MUST BE CAPPED
AT THE TOTAL FARE AMOUNT PLUS CARRIER IMPOSED
CHARGE PAID ON THE TICKET BEING PRESENTED FOR
REFUND.
FULLY FLOWN FARE COMPONENTS MAY BE REPRICED USING
ANY BOOKING CODE WITHIN THE SAME CABIN PROVIDED
THE NEW FARE AMOUNT IS EQUAL OR HIGHER THAN
ORIGINAL.
PARTIALLY FLOWN FARE COMPONENTS MUST BE REPRICED
USING THE SAME OR HIGHER BOOKING CODE.
CHANGES
ANY TIME
CHARGE GBP 150.00 FOR REISSUE/REVALIDATION.
NOTE -
WITH THE EXCEPTION OF TICKETS CHANGED IN THE US
ANY PENALTY FEE MUST BE COLLECTED VIA AN EMD.
REFER TO SPEEDBIRDCLUB.COM OR BATRAVELTRADE.COM
AND YOUR GDS FOR INSTRUCTIONS.
----------------------------------------------
CHARGE APPLIES PER TRANSACTION - PER PERSON FOR
ALL PASSENGER TYPES.
INFANTS WITHOUT A SEAT - NO CHARGE.
--------------------------------------------------
A CHANGE IS A DATE/FLIGHT/ROUTING/BOOKING CODE
CHANGE. NEW RESERVATION AND REISSUE/REVALIDATION
MUST BE MADE ON THE SAME DAY.
--------------------------------------------------
REISSUE MUST BE MADE THE SAME DAY AS CHANGE OF
RESERVATION BUT NO LATER THAN SCHEDULED DEPARTURE
TIME OF FLIGHT BEING CHANGED.
OTHERWISE THE TICKET WILL ONLY BE VALID FOR
REFUND IF APPLICABLE.
--------------------------------------------------
WHEN MORE THAN ONE FARE COMPONENT IS CHANGED THE
HIGHEST PENALTY OF ALL CHANGED FARE COMPONENTS
WITHIN THE JOURNEY APPLIES.
--------------------------------------------------
--- REPRICING CONDITIONS ---
A. BEFORE DEPARTURE OF JOURNEY WHEN THE FIRST
FARE COMPONENT IS CHANGED THE ITINERARY MUST BE
RE-PRICED USING CURRENT FARES IN EFFECT ON THE
DATE THE TICKET IS REISSUED.
B. BEFORE DEPARTURE OF JOURNEY WHEN CHANGES ARE
TO BOOKING CODE ONLY IN THE FIRST FARE COMPONENT
AND RESULT IN A HIGHER FARE THE ITINERARY MUST BE
RE-PRICED USING HISTORICAL FARES IN EFFECT ON THE
PREVIOUS TICKETING DATE OR USING CURRENT FARES IN
EFFECT ON THE DATE THE TICKET IS REISSUED -
WHICHEVER IS LOWER.
C. BEFORE DEPARTURE OF JOURNEY WHEN THERE ARE NO
CHANGES TO THE FIRST FARE COMPONENT BUT OTHER
FARE COMPONENTS ARE CHANGED THE ITINERARY MUST BE
RE-PRICED USING HISTORICAL FARES IN EFFECT ON THE
PREVIOUS TICKETING DATE OR USING CURRENT FARES IN
EFFECT ON THE DATE THE TICKET IS REISSUED-
WHICHEVER IS LOWER.
D. AFTER DEPARTURE OF JOURNEY THE ITINERARY MUST
BE RE-PRICED USING HISTORICAL FARES IN EFFECT ON
THE PREVIOUS TICKETING DATE.
--------------------------------------------------
1. IF SAME BOOKING CLASS IS USED NEW TICKET VALUE
MUST BE LOWER - EQUAL OR HIGHER THAN PREVIOUS AND
MUST COMPLY WITH ALL PROVISIONS OF THE NEW FARE
BEING APPLIED.
2. IF A DIFFERENT BOOKING CLASS IS USED NEW
TICKET VALUE MUST BE EQUAL OR HIGHER THAN
PREVIOUS AND MUST COMPLY WITH ALL PROVISIONS OF
THE NEW FARE BEING APPLIED.
--------------------------------------------------
WHEN THE ITINERARY RESULTS IN A HIGHER FARE THE
DIFFERENCE WILL BE COLLECTED. ANY APPLICABLE
CHANGE FEE STILL APPLIES.
--------------------------------------------------
ANY NON-REFUNDABLE AMOUNT FROM A PREVIOUS TICKET
REMAINS NON-REFUNDABLE FOLLOWING A CHANGE.
--------------------------------------------------
NO-SHOWS FOR A FLIGHT ARE CONSIDERED A
CANCELLATION AFTER DEPARTURE AND CHANGES ARE NOT
PERMITTED.
--------------------------------------------------
TICKET IS NOT TRANSFERABLE TO ANOTHER PERSON.
Category 17: HIP exceptions
THE HIGHER INTERMEDIATE POINT RULE DOES NOT APPLY FOR
CONNECTIONS.
AND - THE HIGHER INTERMEDIATE POINT RULE DOES NOT APPLY
FOR STOPOVERS.
Category 18: Endorsements
THE ORIGINAL AND THE REISSUED TICKET MUST BE ANNOTATED -
NONREFUNDABLE - AND - RESTRICTIONS APPLY - IN THE
ENDORSEMENT BOX.
Category 19: Discounts
ACCOMPANIED CHILD 2-11 - CHARGE 75 PERCENT OF THE FARE.
TICKET DESIGNATOR - CH.
MUST BE ACCOMPANIED ON ALL FLIGHTS BY ADULT
OR - ACCOMPANIED CHILD 12-15 - CHARGE 100 PERCENT OF THE
FARE.
TICKET DESIGNATOR - YP
OR - INFANT UNDER 2 WITH A SEAT - CHARGE 75 PERCENT OF THE
FARE.
TICKET DESIGNATOR - CH.
MUST BE ACCOMPANIED ON ALL FLIGHTS BY ADULT
OR - INFANT UNDER 2 WITHOUT A SEAT - CHARGE 10 PERCENT OF
THE FARE.
TICKET DESIGNATOR - IN.
MUST BE ACCOMPANIED ON ALL FLIGHTS BY ADULT.
Category 23: Miscellaneous Fare Tags
THIS FARE MUST NOT BE USED AS THE HIGH OR THE LOW FARE
WHEN CALCULATING A DIFFERENTIAL. THIS FARE MUST NOT BE
USED AS THE THROUGH FARE WHEN PRICING A FARE COMPONENT
WITH A DIFFERENTIAL.
Category 31: Voluntary changes
IN THE EVENT OF CHANGES TO TICKETED FLIGHTS
BEFORE DEPARTURE OF JOURNEY - APPLIES WITHIN TKT VALIDITY
CERTAIN DOMESTIC REISSUE PROVISIONS MAY BE OVERRIDDEN BY
THOSE OF BA INTERNATIONAL FARES
CHARGE GBP 150.00 OR HIGHEST FEE OF ALL CHANGED FARE
COMPONENTS - DISCOUNTS APPLY - NO FEE FOR INFANTS W/O
SEAT AND
REPRICE USING FARES IN EFFECT WHEN TKT WAS ISSUED
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. NO CHANGE TO 1ST FARE COMPONENT
2. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
3. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
4. ANY CARRIER O- FARE FAMILY IS USED
5. ALL RULE AND BOOKING CODE PROVISIONS ARE MET
6. ADV RES IS MEASURED FROM ORIGINAL TKT DATE TO
DEPARTURE OF PRICING UNIT
OR -
REPRICE USING FARES IN EFFECT WHEN TKT WAS ISSUED
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. NO CHANGE TO 1ST FARE COMPONENT
2. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
3. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
4. ANY CARRIER FARES ARE USED
5. NEW TKT HAS EQUAL OR HIGHER VALUE THAN PREVIOUS
TKT
6. ALL RULE AND BOOKING CODE PROVISIONS ARE MET
7. ADV RES IS MEASURED FROM ORIGINAL TKT DATE TO
DEPARTURE OF PRICING UNIT
OR -
REPRICE USING FARES IN EFFECT TODAY
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
2. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
3. ANY CARRIER O- FARE FAMILY IS USED
4. ADV RES IS MEASURED FROM REISSUE DATE TO DEPARTURE
OF PRICING UNIT
OR -
REPRICE USING FARES IN EFFECT TODAY
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
2. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
3. ANY CARRIER FARES ARE USED
4. NEW TKT HAS EQUAL OR HIGHER VALUE THAN PREVIOUS
TKT
5. ADV RES IS MEASURED FROM REISSUE DATE TO DEPARTURE
OF PRICING UNIT
WHEN CHANGE RESULTS IN LOWER FARE IGNORE RESIDUAL THEN
ADD-COLLECT
ENDORSEMENT BOX- HIGHER NON-REF AMT AND NEW ENDORSEMENTS.
OR -
BEFORE DEPARTURE OF JOURNEY - APPLIES WITHIN TKT VALIDITY
CHANGES NOT PERMITTED/REFUND TKT-ANY REMAINING AMT WILL
APPLY TO NEW TKT
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. CHANGE IS AFTER ORIGINAL SCHEDULED FLIGHT.
OR -
AFTER DEPARTURE OF JOURNEY - APPLIES WITHIN TKT VALIDITY
CERTAIN DOMESTIC REISSUE PROVISIONS MAY BE OVERRIDDEN BY
THOSE OF BA INTERNATIONAL FARES
CHARGE GBP 150.00 OR HIGHEST FEE OF ALL CHANGED FARE
COMPONENTS - DISCOUNTS APPLY - NO FEE FOR INFANTS W/O
SEAT AND
REPRICE USING FARES IN EFFECT WHEN TKT WAS ISSUED
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
2. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
3. ANY CARRIER O- FARE FAMILY IS USED
4. ALL RULE AND BOOKING CODE PROVISIONS ARE MET
5. ADV RES IS MEASURED FROM ORIGINAL TKT DATE TO
DEPARTURE OF PRICING UNIT
OR -
REPRICE USING FARES IN EFFECT WHEN TKT WAS ISSUED
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. WHEN NO INTL COUPONS REMAIN - ALL NEW TRAVEL MUST
BE DOMESTIC
2. CHANGE IS BEFORE ORIGINAL SCHEDULED FLIGHT
3. ANY CARRIER FARES ARE USED
4. NEW TKT HAS EQUAL OR HIGHER VALUE THAN PREVIOUS
TKT
5. ALL RULE AND BOOKING CODE PROVISIONS ARE MET
6. ADV RES IS MEASURED FROM ORIGINAL TKT DATE TO
DEPARTURE OF PRICING UNIT
WHEN CHANGE RESULTS IN LOWER FARE IGNORE RESIDUAL THEN
ADD-COLLECT
ENDORSEMENT BOX- HIGHER NON-REF AMT AND NEW ENDORSEMENTS.
OR -
AFTER DEPARTURE OF JOURNEY - APPLIES WITHIN TKT VALIDITY
CHANGES NOT PERMITTED/REFUND TKT-ANY REMAINING AMT WILL
APPLY TO NEW TKT
PROVIDED ALL OF THE FOLLOWING CONDITIONS ARE MET-
1. CHANGE IS AFTER ORIGINAL SCHEDULED FLIGHT.
OR -
ANYTIME
CHANGES EXCEPT RBD NOT PERMITTED TO THE JOURNEY -
CHARGE HIGHEST FEE OF ALL CHANGED FARE COMPONENTS.
Category 33: Voluntary cancellations
APPLIES FOR FARE COMPONENTS THAT ARE FULLY FLOWN.
NO CHARGE. IF ALL PENALTIES ARE PER FARE COMPONENT COLLECT
EACH. IF THERE IS A MIX OF PENALTY APPLICATIONS, THE MOST
RESTRICTIVE APPLIES TO THE ENTIRE JOURNEY.
REFUND REQUEST MUST BE LESS THAN ONE YEAR AFTER TRAVEL
COMMENCEMENT DATE IF PARTIALLY USED OR ONE YEAR AFTER
TICKET ISSUANCE IF TICKET UNUSED. FORM OF REFUND -
ORIGINAL FORM OF PAYMENT. ONLY VALIDATING CARRIER MAY
REFUND TICKET.
REPRICE FLOWN PORTION USING FARES IN EFFECT ON TICKET
ISSUANCE DATE. FOR FULLY FLOWN FARE COMPONENTS FARE BREAK
POINTS MAY NOT BE CHANGED. FOR PARTIALLY FLOWN FARE
COMPONENTS ONLY DESTINATION FARE BREAK POINTS MAY BE
CHANGED. REPRICE USING NORMAL/SPECIAL ONE WAY/ROUND TRIP
FARES/ANY RULE/FARE CLASS/EQUAL OR HIGHER RBD. PUBLIC
FARES ARE USED IF TICKETED FARE IS IN PUBLIC TARIFF.
QUALIFIED PRIVATE FARES OR PUBLIC FARES ARE USED IF
TICKETED FARE IS IN PRIVATE TARIFF. NEW FARE FOR FULLY
FLOWN FARE COMPONENTS MUST BE EQUAL TO OR HIGHER THAN
TICKETED FARE.
OR -
FARE IS NONREFUNDABLE. IF THERE IS A MIX OF PENALTY
APPLICATIONS, THE MOST RESTRICTIVE APPLIES TO THE ENTIRE
JOURNEY.
REFUND REQUEST MUST BE LESS THAN ONE YEAR AFTER TRAVEL
COMMENCEMENT DATE IF PARTIALLY USED OR ONE YEAR AFTER
TICKET ISSUANCE IF TICKET UNUSED.
REPRICE USING EQUAL OR HIGHER RBD.
OR -
REFUND MAY BE REQUESTED ANYTIME.
PRICING UNIT IS NONREFUNDABLE. IF THERE IS A MIX OF
PENALTY APPLICATIONS, THE MOST RESTRICTIVE APPLIES TO THE
ENTIRE JOURNEY.
Category 35: Negotiated Fares
IF THIS FARE IS VALID FOR A SPECIFIED ACCOUNT CODE.
CONTACT CARRIER FOR DETAILS.
SALE IS RESTRICTED TO SPECIFIC AGENTS
IF THIS FARE IS VALID FOR A SPECIFIED ACCOUNT CODE.
CONTACT CARRIER FOR DETAILS.
SALE IS RESTRICTED TO SPECIFIC AGENTS
SALE IS RESTRICTED TO SPECIFIC AGENTS
SALE IS RESTRICTED TO SPECIFIC AGENTS
Category 50: Application
ECONOMY HAND BAGGAGE ONLY FARES
/ZERO BAGGAGE ALLOWANCE MUST BE DISCLOSED TO CUSTOMER/
APPLICATION
AREA
THESE FARES APPLY
BETWEEN AREA 1/AREA 2.
CLASS OF SERVICE
THESE FARES APPLY FOR ECONOMY CLASS SERVICE.
TYPES OF TRANSPORTATION
FARES GOVERNED BY THIS RULE CAN BE USED TO CREATE
ROUND-TRIP/CIRCLE-TRIP/OPEN-JAW JOURNEYS.
THE TICKET IS INVALID FOR TRAVEL IF FLIGHT
COUPONS ARE NOT USED IN SEQUENCE
CAPACITY LIMITATIONS
THE CARRIER SHALL LIMIT THE NUMBER OF PASSENGERS CARRIED
ON ANY ONE FLIGHT AT FARES GOVERNED BY THIS RULE AND SUCH
FARES WILL NOT NECESSARILY BE AVAILABLE ON ALL FLIGHTS.
THE NUMBER OF SEATS WHICH THE CARRIER SHALL MAKE
AVAILABLE ON A GIVEN FLIGHT WILL BE DETERMINED BY THE
CARRIERS BEST JUDGMENT
RULES NOT APPLICABLE
35 PASSENGER EXPENSES EN ROUTE
So in this sense packaging splits into a single coupon as you suggest, makes a lot of sense (I would prefer this to happen tomorrow!).
The real issue is the cognitive effort and jeopardy around rail ticketing make the display of choice worrisome for many users ("I told the machine I was travelling now - but it has an Off Peak and an Anytime - why is it telling me about the Anytime? Who would buy that? Maybe I do need it - but then, why is the Off Peak being sold as well at the same time?") etc etc.
Fairness is very subjective but not likely to come from the moving of split ticketing from a niche practice to the default method of calculating fares. And Advance tickets, which account for a vast proportion of long distance tickets now sold, are not "fair" - you pay different prices depending on how much the assets are being sweated.
Distance is not a good barometer of price because a mile in London isn't worth the same as a mile in remote Scotland, and it's not how most ordinary people think about paying for their travel - but we digress from the original premise of the thread. I think that given the UK's unique rail geography (heavily centralised, radial, but widely distributed across a country which has diverse needs and diverse income levels) a level of "fairness" will not come from blunt tool "make it so" changes but political wrangling to reflect the differing values of different types of journeys. Some of this "make it so" stuff like removing the Off Peak ticket (which effectively used to cap Advance fares) has started at LNER with very mixed results; some of the changes are very anti-passenger but it is also not clear at all to me that the average punter really cares. The trains are still basically full, and as far as UK plc is concerned carrying around higher income, higher productivity individuals to and from the capital instead of selling half the seats for £20 twelve weeks out to students and dilettantes makes better use of very expensive and very heavily subsidised assets.
If I'm correct, then fine, but if not then why not engage with hypothetical solutions? If there's anything I've learnt while wading through all this data, it's quite how outdated, random and inconsistent it all is. I suppose the best I could hope for in the near future would be a GBR (o.e.) retailer selling well-calculated splits by default, and packaging them all into one ticket.
Well, they could do this, but the idea this will become the default *without raising intermediate tickets which are commonly used to split* (for argument's sake: your Brighton to Gatwick and Hitchin to Biggleswade tickets upthread) to make up for the immediate shortfall in long distance revenue is fanciful. This harms people travelling short distances who have their fares pushed up, which is mad because short distances are where rail is least competitive with other modes.
I heartily appreciate a tough line of questioning (if you've got a problem with the maths or any of the logic, I want to know, so I can try to fix it!) - I spend much of my working life engaging in such things, and give as much as I receive. But if I reacted to opinions I didn't like from people I don't know with statements like "That's naive and not borne out by any sort of evidence" (importantly without saying something nice first, which basically gives you a free pass) then I'd soon find myself out of a job, I fear.
Well, this isn't the office where we all have to work on the same team - it's a discussion forum, so it is what it is. I'm not in a position to take on the maths or logic side - I'll diplomatically say it isn't my skill set. But what I do have experience in is rail fares, including in a professional context (usually explaining to customers why the system is awful, on behalf of my employer!), and I understand how and why the system has come to be how it is and the political barriers in the way of reform. If you just say something along the lines of "well if you make all the fares cheaper there will just be more passengers and then the system will pay for itself" then, unfortunately, I can only say that is naive and not something which is borne out by evidence. The railway has significant issues with capacity in key areas and is not an unlimited teleportation machine, so you can't just keep carrying more and more people at lower and lower fares. A good thought experiment is what do you think would actually happen if all McDonald's meals were suddenly £1? There are loads of things which could happen, but unaltered profit levels and optimum use of the company's resources is probably not one of them.
They're not at all bad - and probably significantly better than much of the competition, though as with any system it has its limitations and trade-offs.
And it can't ever make that guarantee*, because in many cases it won't find an optimal combination (some cases deliberately so, others due to upstream limitations).
* Unless you want a watered down, pretty useless guarantee that "it won't be beaten on price provided the tickets being compared against are exactly the same ticket type, and the same split points are used"... Can't think of any third party retailers who would play that game, of course
Anyway, @johnsmith147, definitely keep playing with the data and building cool things. My own reply wasn't intended to pick holes in something being done for free, but merely to highlight there are far more complexities that have not been considered against the backdrop of fares reform proposals.
There are too many technically-ill-informed conversations around reform already happening within the industry, in my personal view. Step 1 for any body trying to make changes should be understanding the current state of affairs, properly. The data, the standards, the functionality supported by the schemas underpinning it all etc etc.
shows that calculating such a thing takes a few minutes' computation rather than the standard line of 'this is so complex it would take aeons to unravel' line which is usually stated.
So, I can see what you meant with this. That is, it's computationally a tractable problem to solve. Which, I mostly agree with, to be honest? We're not talking volumes of data that commodity hardware can't process. But, I object a little bit more to the implication that a (correct) implementation covering the cases discussed in this thread is also straightforward from an engineering perspective. I say this not because I think you don't understand, but because I think some readers may not understand that.
Something that @SickyNicky was arguing for at least 5 years ago now! I'm pretty sure a joint written TrainSplit-Raileasy proposal went to DfT at the time.
For walk-up single fares, it should be guaranteed the cheapest. (Edit: for one passenger only; it already can get pretty complicated with two are travelling together)
As the creator of many of them, I disagree! Many compromises were made: calculation of the cheapest price for a return journey is provably an NP-hard problem, made much more complicated by the lack of an Availability Distribution Service to share availability of advance fares.
Every stage of a fares inquiry is broken down into whether it's a serial or parallel task, and processed in a highly multi-threaded way with synchronisation points between each serial/parallel task. At least that's the basic principle.
To put rail usage into context, one sixth of all National Rail journeys are on the Elizabeth Line alone, and the overwhelming majority of daily journeys in the UK are made using just one ticket and are over distances of 25 miles or fewer. And therein lies the rub - that isn't data for the sake of data, but it tells us something meaningful about what the railway does, where a lot of its funding ought to go, and which sorts of journeys ought to be protected from negative change.
Indeed, my analysis needs to be combined with usage statistics to be meaningful, as has been pointed out. But of course it is almost always the case that people will manipulate the statistics to their own aims -- of course while I've shown at least 90% of station-station pairs can be made cheaper by splitting, weighted by usage this number would be a lot smaller. But I'd imagine still higher than the average person would assume, I'd wager.
Yes, it's a massive undertaking. And it's not just peak restrictions - you also have super off peak fares to consider, plus a variety of other weird products like Evening Out Returns.
Not just a massive undertaking, but I think not actually possible without introducing more types of tickets (eg. with restrictions like 'valid on peak trains as far as Peterborough'), which currently don't exist but would need to to avoid splitting entirely. There may be scope for such things being introduced, but that's an entirely different question and incredibly unlikely.
Everything is the way it is for a reason. The problem is those reasons are out of date and were baked into the fares system in the mid 1990s and the system is not at all fit for purpose. What is fairly clear to me is that railway is not in a political position to demand large amounts of subsidy to just make travel cheaper, which means at best, any changes will need to be revenue neutral.
My general thesis however is that the underlying complexity of fares is not an inherent issue; it's how they are packaged and sold. This is a good blog on the matter: https://www.freewheeling.info/blog/complexity-isnt-the-issue-with-train-fares (though the argument about arbitrariness falls short of making sense).
So in this sense packaging splits into a single coupon as you suggest, makes a lot of sense (I would prefer this to happen tomorrow!).
On this we agree entirely, it seems. But how would you argue that selling such a repackaged split ticket results in a different amount of 'lost revenue' than slightly altering the component prices to eliminate the need for the split in the first place? In such a system, the Brighton-Gatwick-Hitchin-Biggleswade-Peterborough would be sold as one ticket and the railway would lose out on the income in any case -- you argue this would necessitate the shorter journeys becoming more expensive to compensate. What's the difference?
This paper: https://arxiv.org/pdf/2502.08228 takes this problem into account by finding fare structures which satisfy the essential (to me) conditions, namely that splits are never advantageous and that overdistance tickets are never cheaper, but instead of reducing prices to the minimum as I have done, they seek to minimise the difference between such a constrained system and the existing fare structure. This is a far more complex problem and they say even 20 stations 'cannot be solved in reasonable time' (to find the optimal solution, at least, though a good approximation is undoubtedly easier). But still very applicable to zone-structured systems.
I wish I understood more about how these work, as they seem to do an incredible job of maximising revenue in mysterious ways. But they are fundamentally simpler than trains in that you are (always/usually?) constrained to a specific flight, and it generally doesn't stop in lots of places. Pricing a hypothetical open ticket on a route like LHR-JFK would be an interesting exercise... Overdistance fares being cheaper is quite common with connecting flights, though, if I understand correctly.
Distance is not a good barometer of price because a mile in London isn't worth the same as a mile in remote Scotland, and it's not how most ordinary people think about paying for their travel
Of course, they're not comparable. But actually I'd say this is exactly how the average punter thinks about it -- they're usually comparing with driving instead, for which (for any long journey at least) the cost is proportional to distance. Even if nobody is actually doing the maths, just saying 'you could drive there cheaper than that' is enough that they wouldn't get the train.
They really really do! I have had many pub chats with people who care little about trains other than getting them to places, and they find the lack of open tickets very irritating (I live on an LNER route). I am now far more regularly called upon to find tickets for people for trains which are 'sold out', usually by finding Off-peaks from odd combinations of stations including the 'sold out' section. There are often plenty of seats still...
Anyway, @johnsmith147, definitely keep playing with the data and building cool things. My own reply wasn't intended to pick holes in something being done for free, but merely to highlight there are far more complexities that have not been considered against the backdrop of fares reform proposals.
Thanks, I definitely will do. But will be a lot more careful in my wording in future when posting on here... I am aware of (though obviously do not know everything about) how complex it can be, and my post was intended as 'basically 'even at the very simple level, ignoring most of the complexities, it still doesn't make much sense'. I didn't make that clear enough.
But, I object a little bit more to the implication that a (correct) implementation covering the cases discussed in this thread is also straightforward from an engineering perspective.
I know nothing about how such things would actually be implemented, sure. I imagine it's fare more complex than anyone would want it to be. But I'm glad you agree that the number-crunching bit is probably theoretically doable.
== Doublepost prevention - post automatically merged: ==
I've never seen this before! It basically achieves what I was aiming for, it seems, and a lot more thought has been put into making it accessible than I would bother with. I'll think about generating some maps as an end goal. It seems they've gone through a lot of the same processes as I have (including various misunderstandings...). I didn't realise that if a specific station-station flow exists then it overrides a cluster-cluster/cluster-station flow over the same route, even if it's more expensive. This explains a couple of the inconsistencies I've found in my results. I shall have a good poke through the source code...
As the creator of many of them, I disagree! Many compromises were made: calculation of the cheapest price for a return journey is provably an NP-hard problem, made much more complicated by the lack of an Availability Distribution Service to share availability of advance fares.
I'd be very interested in such proofs, I've searched reasonably hard for any papers on the algorithms used commercially but with little success (probably not helped that they have different names for split ticketing abroad, it seems). For a given itinerary (or for open tickets), I'd have thought an upper bound on the number of operations (checking a price combination, o.e) would go like log(n)*n**3 for n stops (just back-of-the-envelope, I'm quite happy to be proven wrong). But of course the number of possible itineraries can quickly become immense.
Almost all of the times my splitting code still 'beats' trainsplit are using split stations which I imagine you'd have discounted as being unlikely to be worth checking, but these are becoming increasingly rare. The most recent resulted in the (completely legitimate regarding change times etc.) itinerary DHM-NCL-CAR-GLC-ANL-PPK-ANL-DMR-TUL, with tickets DHM-HDB-PPK-ANL-TUL. I wish I'd had the time to actually carry this out, but I just coughed up the extra 60p to go via Edinburgh.
The issue with the computational complexity of return journeys is about the non-additive weights when using a return. The cheapest price depends not only on attributes of the outward journey but also on attributes of the return journey, so you can't do a simple shortest path calculation. The simplest example is something like suppose I went from Reading to Didcot today and came back tomorrow. Each leg would cost £9.50 for the off-peak day single fare. Total price £19. But if I came back today then the Off-Peak Day Return is valid for just 10p more, so the return leg is effectively discounted from £9.50 to 10p.
I am not a computer scientist and can't really explain it better than that but this paper by (ex?) forum member tony_mac explains it: https://www.cs.ox.ac.uk/people/paul.goldberg/papers/CIAC13-gm.pdf - in computer science terms the proof is a reduction from "Minimum Set Cover".
The issue with the computational complexity of return journeys is about the non-additive weights when using a return. The cheapest price depends not only on attributes of the outward journey but also on attributes of the return journey, so you can't do a simple shortest path calculation. The simplest example is something like suppose I went from Reading to Didcot today and came back tomorrow. Each leg would cost £9.50 for the off-peak day single fare. Total price £19. But if I came back today then the Off-Peak Day Return is valid for just 10p more, so the return leg is effectively discounted from £9.50 to 10p.
I am not a computer scientist and can't really explain it better than that but this paper by (ex?) forum member tony_mac explains it: https://www.cs.ox.ac.uk/people/paul.goldberg/papers/CIAC13-gm.pdf - in computer science terms the proof is a reduction from "Minimum Set Cover".
The simplest example is something like suppose I went from Reading to Didcot today and came back tomorrow. Each leg would cost £9.50 for the off-peak day single fare. Total price £19. But if I came back today then the Off-Peak Day Return is valid for just 10p more
I am not a computer scientist and can't really explain it better than that but this paper by (ex?) forum member tony_mac explains it: https://www.cs.ox.ac.uk/people/paul.goldberg/papers/CIAC13-gm.pdf - in computer science terms the proof is a reduction from "Minimum Set Cover".
I has suspected this was the case - the proof is beyond me but good to know it exists I guess. All the code I've got using advances just does single journeys, partly for this reason.
However, in the special case (which is what I'm caring about for now) that every single ticket lying along the route (between each station) is the same price for the outward as the return, does this not simplify things to be the same complexity as a single journey? Whereby you take a return price over each leg as the minimum of either the return ticket, or two singles, and just solve for that? Obviously not allowing doubling back. Have I missed something here?
I've decided to expand the scope of my original project in light of the various comments, including route/TOC restrictions and overdistance fares (a brief search shows these to be far more common than I thought!). But this requires a fast implementation of the routeing guide, which will take me some time.
However, in the special case (which is what I'm caring about for now) that every single ticket lying along the route (between each station) is the same price for the outward as the return, does this not simplify things to be the same complexity as a single journey? Whereby you take a return price over each leg as the minimum of either the return ticket, or two singles, and just solve for that?
I've decided to expand the scope of my original project in light of the various comments, including route/TOC restrictions and overdistance fares (a brief search shows these to be far more common than I thought!). But this requires a fast implementation of the routeing guide, which will take me some time.
There is quite a bit of edge cases and undefined behaviour in the routeing guide, so if you want someone to help test/bug fix your implementation, I would be very happy to volunteer.
There is quite a bit of edge cases and undefined behaviour in the routeing guide, so if you want someone to help test/bug fix your implementation, I would be very happy to volunteer.
This could be very helpful, thank you. I've been reading up on it, though haven't started writing anything yet. As far as I can tell, despite all the marvellous pedantry there are still some things which can be open to interpretation...
Incidentally, looking at cheapest standard singles along shortest routes only (so the numbers would be higher with more flexibility), the number of station-station pairs which can be made cheaper by overdistancing is ~24%. More than I'd have thought, but will probably more interesting when combined with splitting (I've not done so yet). The most egregious I've found so far is ~£170.
I have also found there are only 85 station-station pairs for which you can buy an open return but not an open single. I presume these are probably all mistakes.
We are aware of an issue with emails from the Forum to Microsoft-based email accounts (hotmail/outlook/live.com email addresses). This is being looked into currently, thanks for your patience meanwhile.
