Download a Spreadsheet of All Restaurant Inspections

I FOIAed a complete list of restaurant inspections for the C’ville area for the past year. It’s interesting reading! Who can crunch some interesting numbers with this?  #

14 Responses to “Download a Spreadsheet of All Restaurant Inspections”


  • Waldo . . . this is awesome!!

    Amazing information!!

    Someone should give a copy of this to Chef Ralph. A few restaurants on this list really might need his sanitation help (do a sort based on the “NumCritical” violations, which I am assuming means “Number of Critical Violations”, and you’ll see some very interesting restaurants some up).

  • I’m trying to figure out how to best serve up this data. I’d like to provide some sort of a rating system, to provide an at-a-glance understanding of a restaurant’s inspection history. And I think I’m going to lay an API on top of this, so that third-party websites can query and display restaurants’ health inspection ratings alongside reviews, etc. It’d be great, when looking at Google Maps or Yelp, to see that a restaurant might get four stars, but it’s in the bottom quartile of area restaurants in terms of health inspections.

  • Absolutely useless information for foodies. NumCritical, NumNonCritical are the only 2 pieces of oh so revealing data. So, China King Buffet (nasty) has 1 critical and Tokyo Rose (best Sushi in Cville) gets 2? Red Lobster where you can’t come out of there feeling anything but about-to-die (0 critical) compared to Rhett’s that’s mostly good eats (4)? What a joke!

  • Waldo: I don’t quite know what you are talking out in terms of the IT in your posting (I’m just not at the same point you are technologically). However, I’d love to explore this concept with you further. I really think you are on to something that could really be helpful on many levels [highlighting good establishments that are doing what they should be doing, and encouraging (or perhaps “forcing”) poor performing establishments to do better].

    Christian: I don’t understand why this information is “Absolutely useless . . . for foodies”. As a “foodie” (by anyone’s definition – happy to demonstrate my “foodie” resume anytime), I want to support those establishments that a.) provide worthwhile food; AND b.) have a realistic chance of staying in business.

    The simple fact is that establishments that make people sick do not stay in business in the long-term. The higher the number of critical violations, the more likely that someone will get sick . . . eventually.

    Don’t take the lack of critical violations to mean anything from a culinary standpoint; it is just one of the many aspects of a foodservice establishment. However, one’s culinary ability means absolutely nothing if their business goes out of business.

    One can be the best chef on the planet; however, if they can’t make their business work, it really doesn’t matter. Proper sanitation is just one of the concepts a successful establishment must master in order to be successful.

  • Frank – If “you’re the best Chef on the Planet”, you know food better than those bureaucrats in the DoH. Statistics, especially those based on ignorance, and measurements derived thereof, are worthless.

    There are of course norms for ‘Food Safety’. You don’t put raw chicken on the top rack. You clean your deep fryer frequently. Food cannot stay out of temperature for more than 15 minutes. But placing brined or pickled food in glass jars is better than plastic even though DoH doesn’t think so. Adding heapings of salt in your preparation doesn’t raise a flag but it can make you sick. Soy dressing is said to not require refrigeration, but it does. If there’s no bleached bucket near your prep station, it’s a critical violation, but you can spread the nasties when wiping the dinner tables because they don’t check.

    The list goes on and on. In my experience – which is considerable in this domain – that list is not only invalid, it is misleading. So, you can lead the sheeple to the spanking new Red Lobster (zero critical!!!) while I continue to patronize the real Chefs in town. For as long as they’re in business. Fair enough?

  • PS. Growing veggies of your own can lead to a critical violation, but buying GMO crap, serving Listeria fruits or sprinkling Salmonella spinach are just fine. Then again, the courageous DoH inspector doesn’t get to see Domino’s super-trained college kids wipe their **** when boxing up their pizza.

    My advice: use your common sense and don’t depend on those phony ratings.

  • Waldo–

    If you are interested in providing a web service (as described) then perhaps you can simply choose not to build “quantizing” or “binning” functions yourself. That might also take some of the pressure off of you to get them perfect. {grin} If you choose to position yourself as simply a data republisher, you put these other choices onto whomever is using the data. (Sort of a “common carrier” approach to web services.)

    It bugs me enormously to see bad ideas of relevance and ranking spread about on the web with direct impacts on behavior. The possible effect of scaling that you mention in the blog post linked from your first comment would be a perfect example. Admittedly, people like to see “star” ratings, no matter how simplistic or misleading they may be. And with finite time and a lot of decisions to make, they are useful.

    Maybe the problem is one of moving the manner in which such rankings are produced a little further into the consciousness of the people who use them?

    Christian raises the point that there are many ways to judge a restaurant, and I would add that some ways are privileged, for reasons good and bad. If you (Waldo) republish the VDoH data with no algorithmic additions then someone else (several someone elses, even) could mash it into rankings that might include other viewpoints (how much of the foodstuffs bought into a restaurant are from local suppliers, whether fresh produce is the norm, etc.). If you add your own “star” rating, that can certainly still happen, but I know that I would probably take your proffered rating to use in calculation (instead of the more textured and nuanced raw data). There are good reasons I would do that (it’s easier, I trust you, and I think that my calculation on this data probably wouldn’t evolve very differently than yours). Still, it reminds me that every time a data provider offers a nontrivial aggregation or categorical measure packaged with data, it’s worth thinking a little about what it is that’s being offered (whose viewpoint, whose values, etc.).

    Republishing this data as unaugmented web services would still be a tremendous step forward. I would be happy to help with that work.

  • Absolutely useless information for foodies. NumCritical, NumNonCritical are the only 2 pieces of oh so revealing data.

    Right—you’re agreeing with me.

    So, China King Buffet (nasty) has 1 critical and Tokyo Rose (best Sushi in Cville) gets 2?

    Well, yeah. These ratings have nothing to do with the quality of food. They simply reflect health inspection data. If I were providing data about (for instance) how close one can park to a restaurant, you would have the same complaint. But parking has nothing to do with the tastiness of the food, any more than health inspection records generally do.

    Your gripe is with the very fundaments of health inspections. I can’t help you there. You have a perspective that is shared by only the tiniest segment of the population (that is, restaurant owners). But just about all of the rest of us find it useful. And if the source data is unfair, then disinfect it with a little sunlight! The more that people see these records and think “really, they got in trouble for growing their own basil?”, the more demand that there will be for change.

  • If you are interested in providing a web service (as described) then perhaps you can simply choose not to build “quantizing” or “binning” functions yourself. That might also take some of the pressure off of you to get them perfect.

    I am absolutely starting with this premise, no doubt. My goal is to make every bit of this data available via a RESTful API. I’m just trying think another step ahead, because I’d hate for somebody to take this data and do a thoughtless star rating based on lousy metrics. If I can bake in some sort of a boiled-down rating system, based on an advertised, peer-reviewed algorithm, I’d like to hope that it would reduce the chances of somebody inventing their own terrible, unfair one.

    Republishing this data as unaugmented web services would still be a tremendous step forward. I would be happy to help with that work.

    W00t! I could use the help. :) Theoretically, this is as easy as bulk loading the CSV into MySQL and writing a quick script to encapsulate the data in JSON and provide a few methods to call it. I think it’d just be a few hours. My hope is that a proof-of-concept for our district will warrant getting a dump of the data for the whole state 4x/year, to set up an API for the whole of Virginia.

  • This data is mostly useless to me. If making this more readable and usable cost some money don’t forget those of us who have no idea what the software issues are but would like to help. I know Waldo has asked for some some amount of cash for good works in the past and if you need some bucks to make this so, let us non programmers types help.

    I would welcome better use of public info that could actually be used by the public. If what results is now unfair to Tokyo Rose (who is nowhere near as good as Now and Zen ;)btw ) then work to make the ratings more releavant and fair.

  • I know Waldo has asked for some some amount of cash for good works in the past and if you need some bucks to make this so, let us non programmers types help.

    In this case, currently, money wouldn’t make any difference. (But thanks! :) It is possible, I think, that the Virginia Department of Health will get my FOIA request for this data for the entire state and say “whoa—this is going to take some time and, therefore money,” and at that point a financial barn-raising might be necessary to pay VDH to write the software to be able to dump this data on a regular basis. If that proves to be the case, I won’t be shy about asking for help!

  • So, you want to social network invalid data in the hope said data will improve by popular demand? You may have a business idea, but I doubt the data will improve in a positive way: the general public is ignorant about food because there’s no real food tradition within that constituency. TV dinners in the 60’s and 70’s, then 2 income households and finally mega-agro business and fast foods giants obliterated what meager traditions existed. So on one side you’ve got utter incompetence from the DoH and on the other you’ve got a majority of folks that have fried food almost every day and think heapings of cheap mozzarella is what makes a great pizza. Sorry, no, it’s better to have no fake data at all. As for real foodies, they know the score already…

    P.S. I’ll have to try ‘Now and Zen’ and judge for myself!

  • “and on the other you’ve got a majority of folks that have fried food almost every day and think heapings of cheap mozzarella is what makes a great pizza”

    and on the other other side you have folks who think that barbecued chicken on top makes a great pizza…

    As for real foodies, they go out of town for pizza.

  • So, you want to social network invalid data in the hope said data will improve by popular demand?

    Invalid? Nobody said it was invalid. You said that they’re measuring the wrong metrics, and wish that they’d measure different things. But the fact is that they’re measuring what they’re measuring, and the griping of bitter restauranteurs will not change that. If you are correct, then exposing this bad data to a much, much wider audience should result in popular demand for change. If you are wrong, then people will be quite satisfied. Unless you fear that you are wrong, I don’t see why you should object.

    You may have a business idea

    No, I don’t.

Comments are currently closed.

Sideblog