Great Western Coffee Shop

Trains are not as reliable as we would like.  Is that because we are unreasonalble and would like every single one to run and on time, or are there real problems?   

We hear problem reports, it seems, from far and wide.  The BBC article below picks out Huddersfield.  How much worse is that than Bath?  How do trains in Weston-super-mare compare to trains in Southport?  What about your local line or station?

https://www.bbc.co.uk/news/uk-england-66664323


That article refers to On Time Trains (https://www.ontimetrains.co.uk) which I've not explored before - a wealth of comparative data.  To start a look at that comparison of areas, I'm sharing some of their maps with you - looking at the Severnside, London, Birmingham, Transpenine and Scottish Lowlands on the same scales.  If you "open image in new window" you can see more detail; if you visit the site there is a huge amount of data to drill down into

(http://www.wellho.net/pix/ont_20230904_01.jpg)

(http://www.wellho.net/pix/ont_20230904_02.jpg)

(http://www.wellho.net/pix/ont_20230904_03.jpg)

(http://www.wellho.net/pix/ont_20230904_04.jpg)

(http://www.wellho.net/pix/ont_20230904_05.jpg)

Edit to add - I have added a specific station example at http://www.passenger.chat/27850

I think they need to explain their methodology - is this data, for instance, measured against arrival time or departure time in the public timetable or the working timetable? That's potentially four different times two in each type of timetable. When measuring that some services are 1-2 minutes late, it matters.

There is some explanation on the site but, Chris, in my view you are looking at the same small details for every station and the comparison will be pretty accurate. As regards cancellation rates, in my view a cancellation is a cancellation regardless of whether the times are working of public tables, arrivals, departures or averages thereof.  That would only make a difference on the hour by hour charts, and then a minimal one.

Having said that, the situation is comparative and some station or another has to be worst or best.  I would not mind my local station being last if it never the less achieved 98% running, and there was another service less than 30 minutes later anyway if one was cancelled.

Comparisons though are pointless - who gives a t*ss whether their station had more or less cancellations than a.n.other station elsewhere (anywhere?) in the UK?

I am interested in my stations numbers - in other words, how it performed over a period. So my concerns above do matter. AS long as these stats are all measured the same way is a good thing statistically, but how can they be? Terminals can be arrival times or departure times, but then understanding which has been used can be interpreted with causes completely different depending on which *was* used - so we need to know for it to be doubly-as-useful.

Those starred rankings on the station pages are useless too. No benchmarking info & one cancellation/later train can mean 1.5 stars difference!

I agree with your concerns especially if you take out the data smoothing.   We could have a very long and complex discussion!!

I noted that for the single day last Friday (1st September 2023) Melksham got a green and came out 391st out of 1696.    Six hundred stations had no service at all that day due to industrial action (or nothing scheduled anyway such as Pilning!), and GWR had cut the timetable to just 4 calls in the day rather that 18 on the industry feeds ahead of time.   As far as Joe Public is concerned, that was 13 services cancelled out of 17. 

http://trains.im/ppmhistorical/ For those that haven’t heard of it, is a useful resource with easy to interpret graphs.  You can view the national picture and also drill down to each sector within an operator- GWR has 14 of them.

Playing (I admit) ... here are half a dozen interesting reports which show that a single ranking number doesn't really help make the grade:

(http://www.wellho.net/pix/ot923_1.jpg)(http://www.wellho.net/pix/ot923_2.jpg)(http://www.wellho.net/pix/ot923_3.jpg)
(http://www.wellho.net/pix/ot923_4.jpg)(http://www.wellho.net/pix/ot923_5.jpg)(http://www.wellho.net/pix/ot923_6.jpg)

My personal view is that I would give a much higher priority to the number of cancellations and the gap to the next service (i.e. frequency) than seems to be the case

September-to-date figures for Cotswold Line services look a bit sad:

On-time 65.2%

Late: 21.4%

Very late/cancelled: 13.5%

I guess the trains that get wiped off the timetable completely in advance of strike days haven't been included in the cancellations/very late figure