Kent classic lines usage

[Sad Sprinter]

Was looking through the excellent Railway Data site and come across a couple of anomalies:
According to the site, over 39,000 people travelled by train from Dover to Victoria last year. This seems an awful lot considering it's nigh on 2 hours to Victoria. If people wanted cheaper fares, you would have thought they'd take the ex-SER route to London Bridge. but only 1883 people did so.

On the over hand, only over 1000 travelled from Ramsgate to Victoria, whilst twice the number did so to London Bridge. Again, I would have thought the Victoria service would have been more popular than the service via London Bridge. Besides, why are so many people travelling from Dover to Victoria but not Ramsgate to Victoria?

I've seen a couple of stations on Railway Data showing Kent passengers are returning to the classic lines pre-COVID. What is the cause of this?

Ramsgate: https://www.railwaydata.co.uk/odm/gbr/?dest=all&string_out=Ramsgate+[5023]

Dover: https://www.railwaydata.co.uk/odm/gbr/?dest=all&string_out=Dover+Priory+[5033]

 

[ponponjo]

Worth noting that if you're booking Advance fares, there can be a (sometimes) significant price differential between going to Victoria and going via London Bridge.

On the other hand, any figures for London Terminals tickets can only be estimates - there's also the complication of weekend engineering works meaning trains via Ashford getting diverted to Victoria or trains via the Medway Towns getting diverted to Cannon St via London Bridge that potentially enter the mix.

 

[Sad Sprinter]

Thats quite interesting. Why is it higher going to LBG? To make room for people coming in from Sevenoaks? Interesting that its virtually the same time either way from Victoria or London Bridge but I'm guessing even longer to get to Charing Cross.

 

[nwales58]

The split between London stations could be very inaccurate. It is estimated by a model where tickets are inter-available.

The total across London stations will be much more accurate.

So the cause is likely the distribution model. It uses something like journey times constrained to on-train counts (which have no origin and destination) applied to the ticket data with assumptions about season usage and probably bodge factors to get rid of the most obvious inconsistencies. That all causes inconsistencies in places you have not or cannot validate.

 

[zwk500]

Have a look at this document: https://dataportal.orr.gov.uk/media...x-steer-historical-methodological-changes.pdf which explains all the various adjustments made to the data to reach the station usage estimates.