Great Western Coffee Shop

We all sorta know that long distance trains will be busier on Friday evening, and it's common for passenger counts to be done on Thursdays (in our parts, anyway) to avoid day-of-week distortions.   But just how to passenger numbers vary across the week?   

I'm looking to get some sort of measure of how I would expect services to differ in loadings through the week - starting off with a wild guess that Monday evening and Friday morning on a commuter line are going to be the quietest peaks. 

Searching for figures, the nearest I came up with was
http://www.omninerd.com/articles/How_To_Beat_Traffic_Mathematically
which is a LONG way from what I want because
* It's a different country
* It's a different mode of transport
* It's measured by traffic delays rather than numbers of people
but it does seem to confirm that Friday morning is quietest

Has anyone got / seen / know of a data source that gives loading comparisons for UK trains across days of the week and within that times of day?

I think you are probably right in your assumptions - this morning was no exception either (but probably flawed with last night's poor effort in Brazil)

I'm not sure anything is published with that amount of detail - and I'd expect those doing these counts are probably looking for the busiest times, and may avoid the quietest for that reason?

Thanks, Chris ... perhaps I can get some figures via the CRP's non-community members if I really need them; it was more our of interest and looking to consider some factors in our individual train counts.

We are very conscious to count each train and on a selection of days.   I'm very much aware that if you have 2 trains, 1 carrying 2 passengers and one having just 2 seats empty, that almost anyone you talk to will tell you the train's inevitably ('"packed and with people standing"), but in practise you only have 50% occupancy, and we're trying very hard to avoid falling into the "it's always busy" trap on the TransWilts.

I think you'd be entitled to count Tues-Thur, as Mon/Fri are always known to be non-average days.

If it helps, Weds/Thus are the busiest days on Chiltern Mainline - Thursday the busiest overall (except for weekenders return on Monday morning, which our User Group always exclude)

Friday mornings is definitely the quietest rush hour. Several reasons for this in my opinion (noting that by the evening, "weekender traffic" compensates on longer distance routes, although maybe not on local lines):-

Part time workers usually seem to avoid Fridays, maybe not surprisingly. With the rise in flexible working and regulations that put more onus on employers to allow part time working, numbers have increased. So the "part time effect" is probably increasing.

I suspect the distribution of single days off is not even across the week, but proportionately more on Friday (and maybe to a lesser extent Monday), to facilitate a long weekend.

There are fewer business meetings on Friday, so less travel, eg to London. Even with internal meetings my diary is always much quieter on a Friday.

Finally, and this is a bit more supposition, maybe of people who have the opportunity to work from home on an occasional basis do so on Friday. I can certainly think of a couple at my work who usually wfh on Friday.

My guess is the total effect would be at least 15%, and judging by the TW figures posted today, that would be borne out. There were only 85 pax on the two northbound commuter services north of Melksham today, yet on other days the number is between 100 and 110.

Whilst looking for something else, I uncovered this document which has some commentary about how and when the data was acquired for passenger counts and how reliable it is, which may be of some use to you. Not sure if the link will work, but if you google it you'll probably find it:

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/252516/rail-passengers-crowding-2012-revised.pdf (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/252516/rail-passengers-crowding-2012-revised.pdf)

I like the acronym they have come up with for Passengers in Excess Capacity - PiXC...makes it sound like such a nice, fairytale type experience!!  ::)

I'm often very cautious about the results of a survey. As Sir Humphrey would tell you, the problem with statistics is that they could well be skewed in favour of whoever commissioned the survey. i.e. by the choice of questions in a verbal / written questionaire or the method of data collection for facts it is possible to get different outcomes.

Thanks for all of those inputs / thoughts.  That DfT document (the link does work) is very interesting - it updates my "when is peak" data which was getting rather stale, and confirms that a drop in numbers on a Friday morning peak is explicable.  Nothing, I'm afraid, puts numeric values on that - it's not necessarily something we need though - just need to be aware.

The point about statistics being skewed is well made.  We've taken steps to make our data as rugged as possible - you'll see notes elsewhere on counting methods, for example, and we're carefully erring against what we want to prove (takings a pessimistic view, if you like) where we identify the possibility of skew.  It's also good to do multiple analyses / exercises and see if they come to the same conclusions ...

The "days of week" question, by the way, looks to prove nothing.  It's asked to help understand and look behind what appeared to be a bit of a trend.