Station Usage Data 08/09

[WatcherZero]

Published today

Notes
http://www.rail-reg.gov.uk/upload/pdf/stn_usage_report_0809.pdf

Figures
http://www.rail-reg.gov.uk/upload/xls/station_usage_0809.xls

Notably the formula now incorporates MOIRA Replacement Demand Matrix data to better estimate PTE zonal tickets which were previously uncounted by LENNON, the result is most stations in PTE areas have risen by an average of 28%.

 

[Aictos]

Thanks for that, as expected though growth was confirmed at my station - not much, but still a lot better then to be expected.

 

[Proud Salopian]

Growth at 16 of the 19 stations in Shropshire: the 3 that saw a fall are Bucknell, Hopton Heath and Knighton (all on the Heart of Wales Line - though Broome did see a rise). Telford Central almost at 1 million, and Shrewsbury now above the 1.6 million mark! Craven Arms, Shifnal and Wem have all jumped above the 100,000 mark.

 

[Ivo]

Ah, excellent. Been waiting for this for many-a-month. Thanks for the heads-up.

 

[Kareem]

Quite surprising to see quite a high number of stations in Greater London with a decline in rail users.

I suppose, stations which are served by FCC is not a surprise!

 

[Proud Salopian]

Perhaps commuter numbers are down due to the recession? This would affect some stations/lines more than others. Here in Shropshire for instance the railways are quite heavily used for leisure and other non-business/commuting travel, including by foreign tourists, so we've not been so badly affected by the recession (passenger numbers wise). The decline on the Heart of Wales Line could be explained (in part) by the engineering works affecting the line quite often. The large increase on the Shrewsbury-Wolverhampton line is largely explained by the change in the way the usage is calculated and by increased measures against fare evasion/avoidance.

 

[yorksrob]

Excellent, It's always an eye-opener to sort by the total number of entrances and exits and see how many stations with better train services and station facilities are actually lower down the list than yours!

 

[Ivo]

Is there a reason that stations seem to have identical entry and exit data, which is of course completely inaccurate? And damn them for using Calibri!

 

[deltic]

Does anyone understand the interchange numbers

Do 4.5m people really travel by train into London Victoria and catch a train back out again or does it include people who cross London and then catch a train out of another London Termini

 

[Ivo]

A quote from the report:

Most of the train tickets that are sold are inter-available – the customer has a choice of routes and operators. For example, when a customer buys a ticket to travel from Leicester to Leeds, that customer may travel on various combinations of Midland Mainline, GNER and Virgin Trains, and may interchange at Doncaster, Sheffield or Derby.

Um, hold on a sec; GNER? Midland Mainline? Virgin Trains? None of those three is still around (this Virgin would be Virgin CrossCountry)! Someone needs to check this! This file is supposed to be 2008/09, not 2003/04!

 

[royaloak]

Does anyone understand the interchange numbers

Do 4.5m people really travel by train into London Victoria and catch a train back out again or does it include people who cross London and then catch a train out of another London Termini

Surely those people will then (at some stage) return to the other London termini, cross London and use Victoria, unless they are moving house of course!

 

[WatcherZero]

Interchange is people who arrive in and change trains never exiting the station by foot, probably calculated from multipart tickets. The exact same Station entries and exits is strange. Perhaps their using some sort of doublebook accounting where they must match?

GNER since the report includes 07/08 figures as well and GNER closed December 2007, Midland Mainline Nov 2007, .

 

[Ivo]

Interchange is people who arrive in and change trains never exiting the station by foot, probably calculated from multipart tickets. The exact same Station entries and exits is strange. Perhaps their using some sort of doublebook accounting where they must match?

GNER since the report includes 07/08 figures as well and GNER closed December 2007, Midland Mainline Nov 2007, .

Okay, fair enough I guess, but that was still during the previous year. I didn't realise MM changed that recently. And VXC was about then as well I believe, so I accept this theory

Anyway, with regards to the data, even Denton has an equal number of travellers! I would have thought that most of Denton would be outsiders arriving to experience the legendary nowhere station, who would then go on by bus... A few are different, but I see a problem here. An example is Waterloo. It does have a difference between the two, but the fact that eficit of only 127,000 isn't very likely, notice the 07-08 figure of about 91m. If I remember rightly the 07-08 figure at Waterloo was noted because it was a nine figure number!

 

[WatcherZero]

All the stations that declined by more than 12% are in London or Essex (bar Sellafield where services have been withdrawn and the odd other example like Harlech and a few other ~50k Welsh stations) even after stripping out all the Osytercard changed stations.

Now its hard to compare with the regions because pretty much every station is up over 28% due to accounting for PTE tickets now (everyone in Greater Manchester at least bar Altrincham which had a minor decline). Now is this the economy and London has seen a big fall in travel while the rest of the country has seen an increase, or because theyve been overestimating London users for years and the ~30% across the board decline with the new model reflects that?

 

[Solent&Wessex]

I notice that Norton Bridge, Barlaston and Wedgewood - all of which are still open (I don't recall them ever being closed) - but do not have a train service (they have been served by temporary rail replacement buses for a good few years now) don't feature in this list.

Closure by stealth??

 

[Ivo]

All the stations that declined by more than 12% are in London or Essex (bar Sellafield where services have been withdrawn and the odd other example like Harlech and a few other ~50k Welsh stations) even after stripping out all the Osytercard changed stations.

Now its hard to compare with the regions because pretty much every station is up over 28% due to accounting for PTE tickets now (everyone in Greater Manchester at least bar Altrincham which had a minor decline). Now is this the economy and London has seen a big fall in travel while the rest of the country has seen an increase, or because theyve been overestimating London users for years and the ~30% across the board decline with the new model reflects that?

Whilst I would say that including Essex is a bit unreasonable (the only one beyond the M25 is Cressing, which is one of the county's smallest stations), the London note is an interesting one. The cyncis will claim that it is recession-related; whilst they may have a point it is worth noting that in reality fewer journeys are made between two set points in London with each passing year anyway, as more switch to Oyster/Travelcard. But do look at this example. Sort the list by the "Check" column, and then A-Z second, with "Check" in Descending order. Scroll down to Cricklewood, Harold Wood and Upminster (obviously not at the same time!). Starting with Cricklewood, the entire list as far as Greenford is in the London Zones (17 in all); with Harold Wood the list continues until Leytonstone High Road (20 stations)! Also, from Upminster, this trend continues for 16 stations, including Overground stations in Watford (not technically London). It does say a lot, considering there are only so many National Rail stations in London (333 in the Travelcard Zones according to the list).
--- old post above --- --- new post below ---

I notice that Norton Bridge, Barlaston and Wedgewood - all of which are still open (I don't recall them ever being closed) - but do not have a train service (they have been served by temporary rail replacement buses for a good few years now) don't feature in this list.

Closure by stealth??

Norton Bridge may as well be closed; frankly Denton is better looked after than it is! The entire platform seems to be overgrown!

 

[Solent&Wessex]

Norton Bridge may as well be closed; frankly Denton is better looked after than it is! The entire platform seems to be overgrown!

I know it might as well be closed as it is only served by buses, and you can't actually get to the station now as the footbridge which leads to the island platform has been pulled down - so there is no way to access the platform. A quick look on Google Streetview shows that all evidence of the station has gone at street level. There are no signs or anything to indicate that there is a station there.

BUT, the station is still open, You can sell tickets to and from it and it features in the timetable database, so it should still be counted. Unless they are trying to hide it and hope that people forget about it?

As an aside, I cannot see Stone either - and that is definitely still open as it is served by some Virgin services.

 

[DaveNewcastle]

Is there a reason that stations seem to have identical entry and exit data, which is of course completely inaccurate?

I'm finding that particularly difficult to comprehend.

The previous dataset indicated very clear differences, and in some cases we'd not be surprised by the big difference between Entry and Exit. But in this dataset nearly every station shows EXACTLY the same numbers. Can't be right?

 

[Mvann]

I'm finding that particularly difficult to comprehend.

The previous dataset indicated very clear differences, and in some cases we'd not be surprised by the big difference between Entry and Exit. But in this dataset nearly every station shows EXACTLY the same numbers. Can't be right?

I think note 9 might have some of the answers to that. A lot of the data is taken from ticket sales. So if a journey works out to be cheaper by buying a return instead of a single ticket, I'd buy the return ticket and not use the return portion, but the return portion is included in the stats. Also, various season tickets are worked out on a set formula and don't take peoples actual uses into consideration, e.g I used to travel from Leicester to datchet, but might come back only as far as either mkt harborough or Kettering for various reasons.

 

[wintonian]

Back when you could be sure local tickets this end of the South West Main line would only cost 10p more for a return I would buy a return if I only wanted to go one way because thats what I normaly bought and I didn't care about 10p incase I needed to get back even if that was unlikley, though now your more likley to pay 25 or 30p more.

But thats probably just me.

 

[Mvann]

As I lived in a village between Leicester and mkt harborough, it depended on how I was getting home as to which station I got off at. Also, as my grandparents lived in Kettering, I sometimes stopped there instead.

 

[deltic]

Interchange is people who arrive in and change trains never exiting the station by foot, probably calculated from multipart tickets. .

Where are the 1.7m people going who interchange at Charing Cross - they come in on a train that will have stopped at London Bridge and Waterloo East and go back out on a train that also stops at these stations (with a few minor exceptions)

As for the 20,000+ who change at the two Windsor stations are they just travelling in and back out again!

 

[DaveNewcastle]

I think note 9 might have some of the answers to that. A lot of the data is taken from ticket sales. So if a journey works out to be cheaper by buying a return instead of a single ticket, I'd buy the return ticket and not use the return portion, but the return portion is included in the stats.

I'm sure you're right in some cases, in fact in very many cases. But not ALL cases
Look again, and try to find a difference of even just one or two.
Some examples: Journey one way by as a car passenger and return by rail.
A recreational walking or cycling trip between 2 stations on different lines.
An uphill commute to work in the morning but a leisurely walk home after work.
I've done it myself several times in the year. To get a difference of exactly zero means that exactly the same number of people will have done the opposite of me in the year. Really?

Previously, the famous walking stop of Corrour (there's nowt else there) had a difference of about 1000 (more exits than entrances). Now that difference is exactly zero.

 

[Mvann]

I'm sure you're right in some cases, in fact in very many cases. But not ALL cases
Look again, and try to find a difference of even just one or two.
Some examples: Journey one way by as a car passenger and return by rail.
A recreational walking or cycling trip between 2 stations on different lines.
An uphill commute to work in the morning but a leisurely walk home after work.
I've done it myself several times in the year. To get a difference of exactly zero means that exactly the same number of people will have done the opposite of me in the year. Really?

Previously, the famous walking stop of Corrour (there's nowt else there) had a difference of about 1000 (more exits than entrances). Now that difference is exactly zero.

I hadn't looked at the numbers bit. Have enough of that at work. I just looked at the parameters of the data to see if there was any clues as to why.

 

[Greenback]

There is something in note 5.2 that I can;t understand on a Friday evening, but it may have a bearing on exits and entries. I wish i could be bothered to try and work it all out, but my brain is full!

 

[DaveNewcastle]

There is something in note 5.2 . . .

Yes.
Here it is (my bold) :

In 2008/09 further improvements have been made: Via the integration with the process that creates the MOIRA Replacement Demand Matrix, PTE ticket sales are now included, in addition to TfL sold Travelcards, and some airport link tickets.

The method for estimating passenger journeys from ticket sales has changed. This is a result of using the MOIRA Replacement Demand Matrix as a starting point. The MOIRA Replacement Demand Matrix does not disaggregate single journeys, and so when estimating passenger journeys all ticket sales have been split equally into the two directions of travel. This will only have an impact on the O-D Matrix if there is more travel on single tickets away from a station compared to travel to the station, which is not likely to be material. Therefore in the Station Usage file, entries are the same as exits.

Silly of me not to get it. Clearly, if you have two different quantities, then before you publish them, you introduce "an improvement" to make them the same.
Similarly, if we have dry weather for 5 days and rain for 3, we "improve" the figures by making them both 4.
I'm even beginning to feel the improvement already.

 

[Mvann]

Case of lies, damned lies and statistics.

 

[Greenback]

Yes.
Here it is (my bold) :

Silly of me not to get it. Clearly, if you have two different quantities, then before you publish them, you introduce "an improvement" to make them the same.
Similarly, if we have dry weather for 5 days and rain for 3, we "improve" the figures by making them both 4.
I'm even beginning to feel the improvement already.

My head hurts

 

[WatcherZero]

Well next year their promising to plug in Oyster journeys which would make London at least more accurate with exact journeys including interchanges. I guess for the rest of the country its a case of using MOIRA data thats more accurate than LENNON but less precise, if you understand what I mean. Unless next year they also use LENNON to break down journey imbalance and get a rough ratio then apply that to the MOIRA data to estimate unequal flows.

 

[Ivo]

Well, thank you to Dave (again!) for clarifying that one, but I couldn't have worded a response better than that of Mvann:

Case of lies, damned lies and statistics.

Another situation I would suggest is the following. Suppose Centro revised calls at Aston so that all traffic through it runs one-way. Unreasonable maybe, but it goes without saying that people would only bother with the train for the direct route. There are plenty of buses between central Birmingham and Aston which would cover the indirect route.