Changing by doing

I got the Eurostar back from Paris yesterday. So I made a new thing.

This time I’ve combined data from the excellent new Energy Performance Certificate API with council tax bands and historic individual property prices to try to work out which nearby properties might be in the wrong council tax band.

It’s total nonsense. But nonsense made by AI!

Anyway, have a play at OverBanded.com

What did I learn this time?

1. Matching UK addresses across datasets is boring and hard. I’m grateful to Ministry of Justice UK for open sourcing their address-matching logic, but UK-wide address data should be core UK digital public infrastructure. This won’t ever happen unless institutional incentives change radically.

2. There is loads of interesting data in MHCLG Digital’s new EPC API. Like, the height of ceilings, number of rooms and number of extensions. The quality of all this lovely data seems somewhat variable.

3. More public organisations should publish their information via a proper API that follows Government Digital Service API standards. Looking at you, Valuation Office Agency.

4. HM Land Registry blocks API calls from Replit, the vibe coding platform I use. I wish I’d known this when I left Paris as I could have finished the thing by London.

5. Council Tax banding is a total joke. But you knew that already. The waiting list for a VOA council tax band challenge is currently 12 months, and only heading in the wrong direction.

6.I would *strongly* advise against challenging your band – and certainly don’t do so on the basis of using this stupid toy! You’re just as likely to see VOA decide to increase all your neighbour’s bands as decrease yours. Awks.

Leave a comment