please empty your brain below

Oxford Circus gets closed due to overcrowding in the evening peak pretty often, though maybe less than it did pre-pandemic. I assume that's why it's so much more popular as a destination than as an origin.
When I used to travel on Southeastern from Shortlands to Victoria, it was quite normal for many of the passengers to alight at Brixton, queue to walk down the narrow staircase and transfer to the Victoria line to avoid the regular closures at Victoria Underground due to overcrowding in the morning peak. It wasn’t worth returning to Brixton in the evening because you’d risk not being able to get on a crowded train. This was many years ago, pre-pandemic and before the reconstruction of Brixton tube but I wonder if it could still be a reason for the figures from Brixton to Oxford Circus.
It's interesting for sure but who the hell does the requesting customer think he is with that email? A request like that is going to take hours of work yet he can't even be bothered to write a nice email. FOI is the worst sometimes
My thoughts exactly, Simon G.
There ought to be some sort of filter that provides full information to proper, well written requests and brief summaries to requests that are written in crayon or miserable English.

But the law probably doesn't allow that.
London Bridge - Waterloo is intriguing, considering how long it takes to reach the Jubilee Line platforms compared with National Rail.

However, tube to tube using Oyster or contactless must mean it misses out Overground to tube and National Rail to tube, and vice versa, as well as those joining at interchanges, for example Highbury & Islington.
I think the most astonishing revelation here is that 600,000 people use the dangleway
An interesting topic for an annual visit.
When I lived in Brixton it always amazed me how quickly I could get to Oxford Circus - especially compared to trying to get anywhere from Camberwell.

According to this Mayoral Question around 10% of tube passengers use paper tickets but I suspect they are concentrated at stations like King's Cross which might be enough to change some of the rankings.
I suspect the dataset is incomplete/truncated through the improper use of the old .xls format to process/publish the data which can only handle 65536 rows.

Same issue that the UK government experienced briefly during the early months of the pandemic.

<sigh> The unlimited .xlsx format has been available in common use since 2007, and even a "bare" .csv format would not have been limited but would be easy for all to process.
As lovely and fascinating that this data is - I do feel a little sorry for the person(s) at TfL who have to answer quite so many frivolous Freedom of Information requests from nerds...
Maybe a token fee would dissuade some of them!
I wonder whether the Elizabeth Line opening will reduce Oxford Circus figures, to eliminate over-crowding.
The Waterloo and City annual journey statistics would be a good cross-check for these numbers. But it was shut for almost all of 2021.
Recent occupants of Downing Street have demonstrated that there is more to a person than just their level of education.
"A request like that is going to take hours of work". Like many responses, I suspect this may be just finding and then copying data from one place to another, but the time might be up to 18 hours, which is the FOI cost limit of £450 at a notional £25 per hour. Any longer than that, and TfL can and probably would refuse to comply with the FOI request.

The letter says the spreadsheet has the exact data requested but does not mention that some is omitted. The spreadsheet indicates that small counts are removed, but does not explain the reasons or methodology (at least someone spotted the issue). I suspect it is truncated and not all of the 31s are included - there are over 100 point-to-point routes with 32 journeys (ditto 33), but only 57 listed at 31. Perhaps we need a follow up request to asking for the data for all journeys of say 35 or below?

Yes, I think a nominal fee - say £10 - would stop most frivolous FOI requests. A bit like the prescription charge.
Regarding the 65536 rows of data. From memory this is still the limit on importing data from a CSV file with an up to date version of Excel.
So I'm suspecting that the data came straight out of a relatively simple query on the ticketing system and was imported into Excel for ease of reading.
The purpose of language is to communicate. Judging by the fragments quoted, this requestor manages that very well. Yes, it looks as if they didn't have the usual conventions for written English knocked into them at school, but so what? I find the style quite refreshing.
I dealt with a few FoIs in my TfL days. This one would have been easy.

The difference in the top two flows will likely be a combo of Oxford Circus evening closures as previously mentioned but also the attractiveness of a seat from Brixton. I used to live between Brixton and Clapham and travel to Oxford Circus. I would always walk to Brixton in the morning but get the tube back to Clapham North in the evening.
Interesting figures indeed for us geeks, but TfL really shouldn't be legally obliged to spend time and money producing them just because someone wants to know. FOI should be about revealing inefficiencies, discrimination and bad practice in public bodies, with a reason having to be given for the request and an ombudsman to adjudicate if the request is turned down. Anything else should be at the subject body's discretion, and if the requester explains the background (maybe "I'm writing an article") then it might be answered anyway as good customer service.
This shows how badly London needed the Victoria Line.
As others have mentioned the data is easily extractable and I think available already on TfL's data site which is mainly aimed at app-developers.

I'm very much of the opinion that this sort of data should be published and not need a FOI. But at least TfL are a beacon of openness compared to DfT and National Rail where everything is clouded in secrecy.

I was surprised by the numbers for North Greenwich to Canary Wharf, due to a lot of bus interchange?
It's a shame the FOI applicant's requested format wasn't for origin stations in the columns, and destinations in the rows. That would have avoided issues with defunct Excel version row limits and made station pairs easier to view, at the cost of making it a bit harder to find the top/bottom ranking journeys.
Shame that the link to the request has now broken at TfL's end (I suppose). A fact, along with "Freedom of Interest", that made me wonder whether this was a DG joke.
The FoI has mysteriously disappeared (but the spreadsheet link still works).
I know that roadworks in Edgware town centre have been pretty bad over the last couple of years, but I find it extremely surprising that 38 people took the tube from Edgware to Canons Park.
You can access entry and exit data for each Tube station from 1 July 2020 to (as it stands, updated weekly) 13 November 2022 here.

For example, you can see entry at Westminster approximately doubled from 15 to 18 September 2022 (up to over 40k from around 20k normally) and then halved on Monday 19th (10k). But exits did not change so much over that period.
Northwood to Mill Hill East - may relate to people with corporate hospitality tickets going to watch Saracens play rugby.
14591 travelled Covent Garden - Leicester Square and 9387 did the return.

3017 went from Charing Cross to Embankment.

35 people apparently travelled from Kenton to Northwick Park!
The FoI request is back on the TfL website: now with a CSV file, which goes down to include the 288 journeys only made once (which include Kings Cross LU to Heathrow Terminal 4 - !??? - which makes me wonder as to what was happening...) Note: Libre Office will pull it all in.

Two other things to note: it only gives figures from one LU station to another: journeys starting or finishing at a non LU station (or both) are not included; and there are some stations (and hence flows) which might be expected to be there as LU which aren't - Wimbledon and Richmond are two I've noticed.
Well spotted, thanks!

The complete database doesn't include the combinations with zero annual journeys, which I see include Croxley to Ruislip Gardens and Theydon Bois to Ickenham (in both directions).

Heathrow T4 station was closed throughout 2021, so it's amazing there are any journeys to it at all.
This is certainly only a subset of the real data available, and thus your conclusion, through no fault of your own, is in danger of being flawed, unless your source can guarantee that this is the top portion of a database sorted by journey numbers. Garbage in - Garbage out.

The actual number of possible journeys has be the factorial (mathematicians please confirm) of the total number of Tube stations, and I very much doubt that number is exactly 64 x 1024 (64K).

It has to be Micro$oft source; you can just smell it.
It's the top portion of a database sorted by journey numbers, that's precisely how they published it.

Also, do read to the end of the post.
If Colliers Wood and Wimbledon Park were connected by a line, or even just a connection from the former's line to Wimbledon (e.g. the District extending over the Sutton line and interchanging the Nothern at Morden), the journey would definitely be used by more than 1 person.
I think your spreadsheet is incomplete. As a nerd, the number 65536 is special, since its one more than the max that can be held in a 16 bit number (and we start at line 1).
I think its cutoff at that point.

dg writes: do read to the end of the post.
Having been playing around, probably the most surprising figure is the 91763 who travelled from Canary Wharf to Canary Wharf, along with the 5037 who went from Kings Cross to Kings Cross, the 32685 from Stratford to Stratford and the 26209 from Waterloo to Waterloo










TridentScan | Privacy Policy