Cross-country overcrowding: what could be done?

[NCT]

A reliable takt-timetable at an attractive frequency and appropriate capacity is more important than a low-frequency direct train at any cost.

 

[35B]

A reliable takt-timetable at an attractive frequency and appropriate capacity is more important than a low-frequency direct train at any cost.

In general, yes. But that low frequency direct train may in some cases be a much more useful service - the railway has a growing habit of expecting passengers to bear the costs of changing rather than considering the value of through workings.

 

[Starmill]

In general, yes. But that low frequency direct train may in some cases be a much more useful service - the railway has a growing habit of expecting passengers to bear the costs of changing rather than considering the value of through workings.

There's lots of value in a "takt" running with maximum length trains hourly or two-hourly, which is then given an overlay of three or six train pairs a day running on especially high-demand corridors where you can't add more carriages, or on medium-demand ones where there's a market for a through service which doesn't fit into an hourly service. In the latter case it can also be used to give secondary stations a long-distance service which they otherwise wouldn't justify. The "overlay" trains can also be selective in the days they run, for example Fridays and Sundays only on leisure-focused destinations, or Tuesdays - Fridays only for joint business and leisure routes. They can also easily switch in and out at different weeks of the year, such as not running in January and February for destinations that mostly attract people when it's sunny.

 

[35B]

There's lots of value in a "takt" running with maximum length trains hourly or two-hourly, which is then given an overlay of three or six train pairs a day running on especially high-demand corridors where you can't add more carriages, or on medium-demand ones where there's a market for a through service which doesn't fit into an hourly service. In the latter case it can also be used to give secondary stations a long-distance service which they otherwise wouldn't justify.

Precisely.

 

[Topological]

This is by far the most important part of this. The 0742 from Birmingham New Street to Bristol Temple Meads really ought to revert to starting back at Manchester Piccadilly as well, as 1012 is rather late for the first arrival in Bristol. Unfortunately an Avanti service is in the way of an 0700 departure from Stafford to Wolverhampton.

Does the connection not work?

I know they send an Avanti in the slot to provide capacity. It is almost like that could have been the answer for the rest of the day too

 

[Starmill]

Does the connection not work?

I know they send an Avanti in the slot to provide capacity. It is almost like that could have been the answer for the rest of the day too

The Avanti service is not really there to offer capacity, otherwise it would be on the train at around 7am, the one at 6am is so early that the extra space is wasted. It's there to balance Avanti's unit workings.

Either way it makes no difference. The Avanti wouldn't arrive in Birmingham in time to make the connection even if it had half a minute shorter dwell at Wolverhampton and skipped its call at Sandwell & Dudley, and the 0610 from Shrewsbury would clash with it anyway in that case, so there's no hope.

 

[NCT]

There's lots of value in a "takt" running with maximum length trains hourly or two-hourly, which is then given an overlay of three or six train pairs a day running on especially high-demand corridors where you can't add more carriages, or on medium-demand ones where there's a market for a through service which doesn't fit into an hourly service. In the latter case it can also be used to give secondary stations a long-distance service which they otherwise wouldn't justify. The "overlay" trains can also be selective in the days they run, for example Fridays and Sundays only on leisure-focused destinations, or Tuesdays - Fridays only for joint business and leisure routes. They can also easily switch in and out at different weeks of the year, such as not running in January and February for destinations that mostly attract people when it's sunny.

This is so far from the British norm as to be near enough irrelevant.

The standard frequency on the Cross Country network is half-hourly (will be in the May 26 timetable anyway) with some section with 2 sets of 2tph combined to give 4tph (Derby - Birmingham, Stafford - Birmingham for example). The overly complex Cross Country timetable structure (and other services) mean some key cross-Birmingham journeys only have hourly frequencies, because connections in the opposite half hour don't work. This is a classic takt degradation. It's the same problem reflected in the Leeds - St Pancras thread - Sheffield isn't operating as a takt-node should - despite most segments operating 2tph they are not all 30 minutes apart and the paths don't line up properly.

The crime isn't so much the first Manchester - Bristol service arrives at 1012, but the earlier 'almost half hourly' paths from Manchester don't arrive at Birmingham at the right time because the Birmingham takt is compromised.

 

[Bletchleyite]

This is so far from the British norm as to be near enough irrelevant.

Except it isn't. "Takt" with peak extras on top is how the south WCML is now.

 

[35B]

Except it isn't. "Takt" with peak extras on top is how the south WCML is now.

No, it's not "takt" but regular interval. The integration implied by the concept of Takt is so far away from the operating model that the concept is a misnomer when actually looking at how the timetables are structrued.

 

[NCT]

Except it isn't. "Takt" with peak extras on top is how the south WCML is now.

The typical regular 'rhythm' on the British timetable is half hourly, not 1 or 2 hourly with overlays.

 

[Bletchleyite]

The typical regular 'rhythm' on the British timetable is half hourly, not 1 or 2 hourly with overlays.

Depends where you are. Manchester-London is even 3tph, while most of the north is based on hourly services from everywhere to everywhere, as the phrase goes. WMT is largely half hourly patterns though, give or take a few services.

 

[NCT]

Depends where you are. Manchester-London is even 3tph, while most of the north is based on hourly services from everywhere to everywhere, as the phrase goes. WMT is largely half hourly patterns though, give or take a few services.

I said typical, and within the context of where Cross Country operates ... Gordon Bennett.

 

[Snow1964]

This is so far from the British norm as to be near enough irrelevant.

The standard frequency on the Cross Country network is half-hourly (will be in the May 26 timetable anyway)

I would question the half hourly from May

eg at Cheltenham Spa
northbound : 11:12, 11:40, 11:59, 12:12 (28, 19, 13 minute gaps)
southbound : 11:15, 11:29, 11:52, 12:15 (14, 23, 23 minute gaps)

that is neither half hourly, or a regular 20 minute intervals, not a single service that anything like 30 minute intervals.

 

[NCT]

I would question the half hourly from May

eg at Cheltenham Spa
northbound : 11:12, 11:40, 11:59, 12:12 (28, 19, 13 minute gaps)
southbound : 11:15, 11:29, 11:52, 12:15 (14, 23, 23 minute gaps)

that is neither half hourly, or a regular 20 minute intervals, not a single service that anything like 30 minute intervals.

It's a half-hourly Birmingham - Bristol with a bonus hourly Cardiff on top, with typical British compromises - which is my entire point.

 

[Starmill]

The crime isn't so much the first Manchester - Bristol service arrives at 1012, but the earlier 'almost half hourly' paths from Manchester don't arrive at Birmingham at the right time because the Birmingham takt is compromised.

This isn't true though. It used to be a through service, from Manchester to Bristol, but it didn't have 12 minutes clear between the arrival and departure. If it had been split but left as two XC services in XC's old timings, they still wouldn't have connected anyway as there'd be only 9 min to change which isn't enough.

Unless you're saying we should go Swiss and significantly lower the minimum connection times? Which I would say were a great idea in theory but in practice we can be confident all that will happen is a huge increase in missed connections. They ain't going to be waiting like they would do in Switzerland.

The typical regular 'rhythm' on the British timetable is half hourly, not 1 or 2 hourly with overlays.

It's generally not. It's usually hourly or with multiple hourly services that dovetail nicely into a very rough half-hourly (or indeed which don't dovetail at all and leave 50 min gaps). Either way, whether it's a base of 4tph, 2tph, 1tph or 0.5tph the point I was making still stands.

 

[NCT]

This isn't true though. It used to be a through service, from Manchester to Bristol, but it didn't have 12 minutes clear between the arrival and departure. If it had been split but left as two XC services in XC's old timings, they still wouldn't have connected anyway as there'd be only 9 min to change which isn't enough.

Unless you're saying we should go Swiss and significantly lower the minimum connection times? Which I would say were a great idea in theory but in practice we can be confident all that will happen is a huge increase in missed connections. They ain't going to be waiting like they would do in Switzerland.

I'm saying timetabling should be more rigorous, with trains actively timed to reflect minimum interchange times. Journeys like Manchester - Bristol should be half-hourly. If a complex network of everywhere-to-everywhere direct trains can't provide it, then run a simpler network where passengers have half-hourly consistent interchange opportunities.

It's generally not. It's usually hourly or with multiple hourly services that dovetail nicely into a very rough half-hourly (or indeed which don't dovetail at all and leave 50 min gaps). Either way, whether it's a base of 4tph, 2tph, 1tph or 0.5tph the point I was making still stands.

In Cross Country land (at least the core part of the network, which is Bristol - Sheffield and Manchester - Reading), the vast vast majority of service groups run at 2tph, including most of WMT, most of GWR, and all of EMR intercity. Passengers should get the full advantage of 2tph offerings - and this is done through making both halves of the hour exactly the same, like they do in the Netherlands.

 

[Starmill]

In Cross Country land (at least the core part of the network, which is Bristol - Sheffield and Manchester - Reading), the vast vast majority of service groups run at 2tph, including most of WMT, most of GWR, and all of EMR intercity. Passengers should get the full advantage of 2tph offerings - and this is done through making both halves of the hour exactly the same, like they do in the Netherlands.

They're inevitably not the same though, that's the whole point, they're two hourly services which are interspersed with one another as best they can.

== Doublepost prevention - post automatically merged: 4 Feb 2026 ==

I'm saying timetabling should be more rigorous, with trains actively timed to reflect minimum interchange times. Journeys like Manchester - Bristol should be half-hourly. If a complex network of everywhere-to-everywhere direct trains can't provide it, then run a simpler network where passengers have half-hourly consistent interchange opportunities.

That would probably work if you had double the number of platforms at Birmingham New Street, like they would in Switzerland, yes.

 

[NCT]

They're inevitably not the same though, that's the whole point, they're two hourly services which are interspersed with one another as best they can.

When you have 2 services 28/32 minutes apart with 3 minutes of journey time difference, you have a lazy half-hourly service, not two hourly services that happen to work together. The passenger perception and expectation is half-hourly - that's the hygiene expectation, not bonus.

That would probably work if you had double the number of platforms at Birmingham New Street, like they would in Switzerland, yes.

No.

 

[Bletchleyite]

When you have 2 services 28/32 minutes apart with 3 minutes of journey time difference, you have a lazy half-hourly service, not two hourly services that happen to work together. The passenger perception and expectation is half-hourly - that's the hygiene expectation, not bonus.

I think that was true when almost all tickets were sold walk-up with no reference to the timetable. It's less so now with almost everyone outside of metro services like Merseyrail using journey planners. There are still advantages to clock face patterns, but them being precisely half hourly is a lot less important and you gain benefits by them not quite being such as serving more stations rather than, for example, determining that only Milton Keynes Central is an InterCity station and serving it with all the Manchesters, rather than having one that does MKC, one Rugby and one whatever the other one is (Nuneaton?)

 

[Starmill]

When you have 2 services 28/32 minutes apart with 3 minutes of journey time difference, you have a lazy half-hourly service, not two hourly services that happen to work together. The passenger perception and expectation is half-hourly - that's the hygiene expectation, not bonus.

That's not how anyone sees it. Most people don't see Derby to Sheffield as hourly half-hourly or anything similar, they look at the times because there are 4tph but jumbled up together. Same for Derby - Birmingham. Same for Birmingham - Cheltenham. And so on.

Realistically you will never get these to half-hourly with all the same run times and stops. It's just not going to happen.

== Doublepost prevention - post automatically merged: 4 Feb 2026 ==

No

Then I'm pleased you agree the current approach is the best use of scarce resources. Unless someone's willing to pay for all of these additional tracks and platforms it's the system we're stuck with.

 

[MatthewHutton]

When you have 2 services 28/32 minutes apart with 3 minutes of journey time difference, you have a lazy half-hourly service, not two hourly services that happen to work together. The passenger perception and expectation is half-hourly - that's the hygiene expectation, not bonus.

The Swiss do the splitting thing as well by the way.

And yes they probably have all services running exactly 30 minutes apart but with lower speeds etc.

 

[Snow1964]

That would probably work if you had double the number of platforms at Birmingham New Street, like they would in Switzerland, yes.

Back in 1970s the cross country timetable managed to have cross platform interchange at Birmingham New Street, the NE-SW and South-NW stopped alongside each other for easy changes. The timetable was built around these easy cross platform changes.

But of course have to have Swiss style connections, need an Operator that normally runs trains on time, but XC doesn't have staff that can manage that everyday.
Most recent 3 periods, just a feeble 34.6%, 44.5%, 51.6% on time
fairer to use within 3 mins, that is just 55.6%, 65.7%, 72.4%

And that is if XC service actually operates
Cancellations for the 3 recent periods 10.7%, 13.1%, 7.7%

Cannot expect passengers to accept connections (rather than direct trains) when 28-45% of trains are likely to put connection in jeopardy, and upto 1 in 8 chance of either train not operating.

Key Business Performance Indicators | CrossCountry Trains

Discover CrossCountry's train performance, including delays and cancellation rates, passenger satisfaction, and complaints.

www.crosscountrytrains.co.uk

== Doublepost prevention - post automatically merged: 5 Feb 2026 ==

I think that was true when almost all tickets were sold walk-up with no reference to the timetable. It's less so now with almost everyone outside of metro services like Merseyrail using journey planners. There are still advantages to clock face patterns, but them being precisely half hourly is a lot less important and you gain benefits by them not quite being such as serving more stations rather than, for example, determining that only Milton Keynes Central is an InterCity station and serving it with all the Manchesters, rather than having one that does MKC, one Rugby and one whatever the other one is (Nuneaton?)

No good using journey planners and buying advance tickets for specific trains when cancellations are short forms are high. Advances for specific trains need that train to operate, and have the seats they sell.

XC cancellations 3 most recent periods 10.7%, 13.1%, 7.7%

short forms 2.4%, 3.6%, 4.3% (trend going up)

If got to take 2 trains (due to no direct trains) for a return, there is a over 50% chance one of the 4 trains is cancelled when 13%, and of 4 trains got a unreasonable chance of your seat not being there if train actually runs

Key Business Performance Indicators | CrossCountry Trains

Discover CrossCountry's train performance, including delays and cancellation rates, passenger satisfaction, and complaints.

www.crosscountrytrains.co.uk

 

[JonathanH]

Back in 1970s the cross country timetable managed to have cross platform interchange at Birmingham New Street, the NE-SW and South-NW stopped alongside each other for easy changes. The timetable was built around these easy cross platform changes.

Which platform island did that use?

How many trains an hour were using New Street at the time?

How long was the dwell for the trains to facilitate cross platform interchange?
 

 

[MatthewHutton]

Which platform island did that use?

How many trains an hour were using New Street at the time?

How long was the dwell for the trains to facilitate cross platform interchange?
 

And how often in reality could you make the connection?

 

[NCT]

That's not how anyone sees it. Most people don't see Derby to Sheffield as hourly half-hourly or anything similar, they look at the times because there are 4tph but jumbled up together. Same for Derby - Birmingham. Same for Birmingham - Cheltenham. And so on.

You are so lost in the semantics you've missed the context completely. Passengers perceive a frequent service on those corridors. The problem is effective frequency drops off a cliff once an interchange is introduced into a journey, because high frequencies on one side don't interchange well with high frequencies on the other. Every 'clever' direct train you introduce to the system by eroding an otherwise half-hourly takt, the more interchanges you break.

Realistically you will never get these to half-hourly with all the same run times and stops. It's just not going to happen.

GB timetable is a whisker away from Dutch quality. It just needs a small number of things being treated more systematically and consistently with discipline. The Southern region has largely done it. There are basic things like symmetry. Derby - Sheffield is a classic example - the timetable is asymmetrical.

Then I'm pleased you agree the current approach is the best use of scarce resources. Unless someone's willing to pay for all of these additional tracks and platforms it's the system we're stuck with.

The model to emulate here isn't Swiss, but Dutch - operate Birmingham New Street with the operational simplicity of Utrecht Centraal. The Dutch principle (at least in the Ranstad area) is simplicity - they don't do try to offer clever everywhere-to-everywhere direct trains - they prioritise simplicity. If a journey is possible in one half hour, then (with few exception) the exact same journey is possible in the opposite half hour. That's what we should have for journeys like Manchester - Bristol.

 

[MatthewHutton]

The model to emulate here isn't Swiss, but Dutch - operate Birmingham New Street with the operational simplicity of Utrecht Centraal. The Dutch principle (at least in the Ranstad area) is simplicity - they don't do try to offer clever everywhere-to-everywhere direct trains - they prioritise simplicity. If a journey is possible in one half hour, then (with few exception) the exact same journey is possible in the opposite half hour.

Yeah but there are plenty of times that doesn’t work.

Marylebone to Birmingham is asymmetric because otherwise you couldn’t fit the High Wycombe stopping service in.

That's what we should have for journeys like Manchester - Bristol.

Hourly is OK for a pair with 75k trips between them in both directions. And it’s not like other pairs are particularly big. The only other relevant ones are Cheltenham which is 30k and Bristol Parkway with 23k.

Really if we took it to extremes Bristol wouldn’t get any service at all and they’d all go to Reading as between it and the other stops south of Birmingham in the Cherwell valley there are a lot more passengers.

Birmingham international is 40k, Coventry is 75k, Leamington is 30k, Banbury is 18k, Oxford is 60k, Reading is 50k, Southampton is 21k.

 

[NCT]

Yeah but there are plenty of times that doesn’t work.

Marylebone to Birmingham is asymmetric because otherwise you couldn’t fit the High Wycombe stopping service in.

I'm not sure you understand what symmetry means in the timetabling context.

Taking an example relevant to this thread. In the May 26 timetable, Sheffield - Birmingham services leave Sheffield at xx26, 56, and arrive at Birmingham at xx06 and xx28 (give or take). A symmetrical timetable would have Birmingham - Sheffield services leaving Birmingham at xx54 and xx32, and arriving at xx04 and xx34. Instead the northbound services are actually dep Birmingham xx03 and xx27 and arr Sheffield xx19 and xx28

 

[MatthewHutton]

I'm not sure you understand what symmetry means in the timetabling context.

Taking an example relevant to this thread. In the May 26 timetable, Sheffield - Birmingham services leave Sheffield at xx26, 56, and arrive at Birmingham at xx06 and xx28 (give or take). A symmetrical timetable would have Birmingham - Sheffield services leaving Birmingham at xx54 and xx32, and arriving at xx04 and xx34. Instead the northbound services are actually dep Birmingham xx03 and xx27 and arr Sheffield xx19 and xx28

Ok but some of that is because only half the Reading services can go via Coventry because of capacity so the other half arrive in Birmingham very early. Should the xx:15 arrival from Reading really sit around until xx:33 or something?

And they’ve also got to hit their slot on the east coast mainline that probably isn’t symmetrical as it has an hourly takt.

 

[NCT]

Ok but some of that is because only half the Reading services can go via Coventry because of capacity so the other half arrive in Birmingham very early. Should the xx:15 arrival from Reading really sit around until xx:33 or something?

And they’ve also got to hit their slot on the east coast mainline that probably isn’t symmetrical as it has an hourly takt.

Still not getting it.

The Birmingham - Reading section isn't symmetrical either. The problem is different parts of the national timetable aren't operating to a common symmetry.

ECML isn't symmetrical - but you can (and should) still have symmetry with an hourly takt. At a station, if the arrival pattern of services A, B, C, and D are xx05, 13, 24, 46, then the departure pattern of the same services in the other direction should be xx55, 47, 36, 14.

With a symmetrical timetable, you would not have problems in one direction only.

 

[MatthewHutton]

Still not getting it.

The Birmingham - Reading section isn't symmetrical either. The problem is different parts of the national timetable aren't operating to a common symmetry.

ECML isn't symmetrical - but you can (and should) still have symmetry with an hourly takt. At a station, if the arrival pattern of services A, B, C, and D are xx05, 13, 24, 46, then the departure pattern of the same services in the other direction should be xx55, 47, 36, 14.

With a symmetrical timetable, you would not have problems in one direction only.

I think it’s starting to make sense!