Great Western Coffee Shop

Via Twitter, a passenger flow map.

Mark

https://github.com/anisotropi4/kingfisher/blob/main/station.md

Fascinating. https://github.com/anisotropi4/kingfisher/blob/main/README.md

A very interesting visualisation. Looking at PAD, you can see how some bright spark out of Tufton St might conclude that the network would turn a good profit if only they closed all the other stations on the western route. On the other hand, looking at smaller stations you can see how they give people access to longer regional journeys - which is why we need more of them!

I have taken a local look - very interesting to see the wide regional and national travel opportunities that are taken up from a simple station serving a medium sized market town

(http://www.wellho.net/pix/mkm_tick_22z.jpg)

Thanks for that link which has helped a bit. Me: not yet in a position to explain what it is the maps for individual stations show, i.e. I don't (yet) understand what it is that's been done here.

Mark

I had to scratch my head for a few minutes too

(http://www.wellho.net/pix/xpdmh.jpg)

OK - journeys ticketed to / from Dilton Marsh in the year April 2021 to May 2022. Each line on the map shows the train routing taken by the most direct passenger line for each ticket sold, and the thickness indicates the number of journeys. So - none of the individual flows broad enough to show substantial thickening and an indication of the furthest journeys sold being to Plymouth, Llanelli, etc.   Nothing beyond Brighton or Ipswich.

By contrast - lots of regional and long distance stuff from Melksham. Some odd looking gaps until you realise that shortest route to Bridgwater is via Taunton and to Weston is via Bristol.  I noted Stansted, Heathrow, Gatwick, Fishguard and Holyhead all mapped. No journeys to Rye, Llandrindod, Newquay, and a data error in that it this East West Rail is already open.  The regional stuff is significant in that it brings people in to the town

(http://www.wellho.net/pix/xpmkm.jpg)

Final example is Bath and here the numbers are high enough to thicken the lines and virtually the whole network is shown, with lines that are accessed from both ends broken in the middle!  The only places I thing are missed are a few like Ledbury and Colwall

(http://www.wellho.net/pix/xpbth.jpg)

What's the dataset underlying this?

Thanks Grahame for your thoughts..

I wonder - how did they account for season tickets? Does each ticket for the journey between two points count as "one"?

In my case, when I was doing two days a week as a consultant during my last two years, and usually travelling off-peak during the pandemic hybrid working regime (having acquired my senior rail card, to "replace" the discounts on my former Gold Card annual), I was buying a return each day I travelled rather than one annual season for the whole year, so potentially would have been contributing many more registrations between Goring and Guildford than when I was doing the journey most working days.

Something interesting to browse. Presumably over time it will be possible to look at how flows change over time. I wonder if enough historic data is available to show patterns pre- and post-Covid if someone has the inclination to do this. I note that the most recent individual station stats have included most common origin/destination for each station for the first time, so perhaps the more detailed dataset has only been made available more recently.

The broken lines show that routes indicated often aren't the actual routes taken. Still useful and interesting, but perhaps showing dots (size indicating flow volume) for each origin/destination would be better. However in dense areas the lines may well be clearer, which perhaps influenced the way it's presented.

As with the usual station stats, I wonder if more distant destinations are under-represented as a result of split ticketing, or indeed if common split-points have inflated figures, or if these are still drowned out by the number of passengers who just buy a the full-route ticket.


I assume season tickets are calculated in the same way as the individual station stats that are released annually; I would guess the answer is somewhere in the published methodology information that I  think is published the station stats. The GWR delay repay calculation figures suggest a possible methodology:
weekly ticket: 1/10 of the ticket price
flexi season ticket: 1/16 of the price
monthly ticket: 1/40 of the ticket price
annual ticket: 1/464 of the ticket price

A similar question arises over stations often accessed with travelcard/zonal type tickets - especially London suburbs. I assume there's a standard methodology for guestimating this but there may be some under-representation going on.

Some info here:
https://github.com/anisotropi4/kingfisher/blame/main/README.md

Data set methodology for the annual returns for station entrance and exit estimates may be found at:
https://dataportal.orr.gov.uk/media/1917/station-usage-quality-and-methodology-report.pdf
and see the base of page 16 and top of page 17 for season ticket figures.  I would expect the same base data to have been used to extract the data flows, especially as ORR has added the most common destination from each station this year.    This report also directs you on to:
https://dataportal.orr.gov.uk/media/1904/station-usage-steer-methodology-report.pdf