Finding the Most Unhygienic Food in the UK | Hacker News Reader

Finding the Most Unhygienic Food in the UK

blog.wolfram.com

61 pointslelf20 comments

The trouble with analysis like these is the lack of domain specific knowledge. For instance the author states that it is suspicious that some authorities inspect 100% of food businesses, speculating that new ones must take a while to appear on their radar. In fact new food businesses must register 30 days before they open. I opened a coffee shop in Oxford (the authors city by coincidence) back in 2010. I was inspected the day I opened for business, and many other authorities try to do the same.

Open data is great, but there is always the temptation to assume it includes everything you need to know.

> I was inspected the day I opened for business

Is this the correct way to inspect though? I mean if you've just opened everything is going to hopefully be shiny and new. If you left it a month before inspecting you might get a better picture.

It will still find problems like incorrect food storage temperatures and cross contamination. These are probably more dangerous than general dirtiness, so it might be a good tradeoff to catch these early at the cost of missing lesser problems.

your intuition is wrong. first of all, inspections are recurring, usually unannounced.

second, new restaurants buy used equipment. if they didn't inspect upon opening, what do you think would happen?

third, there are such things as turn-key restaurants, meaning a new operator simply takes over a kitchen and re-opens it under a new name/concept. nothing is shiny or new.

Inspections are not a one-time thing, they'll be back. And you probably wouldn't want to let someplace operate for a month with grotesque health code violations, before shutting it down.

>everything is going to hopefully be shiny and new

"Hopefully." When it isn't, you'll be glad that's the time they chose to inspect.

At least some of the rating is to do with the facilities so they are making sure that the building and equipment are up to spec.

I don't see how lacking that knowledge is detrimental. It didn't affect the analysis, it just caught the author's attention as an anomaly to investigate. That's part of the process.

Cool! Which coffee shop, just out of interest? (I'm in Oxford these days too)

Unfortunately not one that's still around :)

What a neat visualization, but falls on its face by implying mean score correlates to poor hygiene. Certainly one would want to avoid low-scoring restaurants, but if a region has a high mean score it could mean lax enforcement and lower regional mean could indicate stricter enforcement.

I'd love to see this analysis linked to (say) Yelp reviews complaining about food-poisoning (ala https://www.washingtonpost.com/news/to-your-health/wp/2014/0...). That would make it easier to spot when the health agency ratings don't match actual consumer health experience.

Are you allowed to complain about food poisoning on Yelp?

My town, which coincidentally happens to be Oxford like in the article, has its own local site "Daily info" which includes user-submitted restaurant reviews, and the review form asks you not to make allegations of food poisoning (presumably for legal reasons) and to submit possible food poisoning reports to the local authorities for investigation instead.

I note the Chinese takeaway I got food from last week shows up on this map as a bright red blob...

  > Are you allowed to complain about food poisoning on Yelp?

In the US, sure. UK libel laws could be used to suppress 'food poisoning' without a doctor's note, but I think you'd be pretty safe to say 'sick'/'vomit'/'diarrhea'.

Interesting, The graphs seem to correlate very highly with population density from just looking. I wonder if that is a cause, coincidence, or something in the calculations?

Awesome to see Craigavon as the most hygienic place, I love very close and used to go to school there.

I'd want to see how this correlates to say, the revenue the hotel makes.

Last I checked, Scotland was still part of the UK, yet this data for the 'whole country' doesn't even reach as far North as Newcastle.

Scotland use a Pass/Fail rating, which doesn't suit his algorithms.

>Unfortunately, this excludes Scotland, which runs a different scheme that provides only a pass-or-fail-type conclusion.

From TFA...

Reply on news.ycombinator.com