« /home/rc



Munich's housing market is crazy, I had to find accommodation for my Master of Science in Informatik at TUM.de. I looked up a lot of websites, such as housinganywhere.com, roomlessrent.com, wg-suche.de and most importantly wg-gesucht.de. Receiving an answer from a landlord is almost impossible, due to the high demand and the low offer (this is, of course, also why prices are so high). So I decided to write a python bot1 to find new ads and contact landlords.

Eventually, after contacting 1008 ads, I found accommodation, I quickly threw up a simple script to get some insights:

number of messages per chat with landlord

wg-gesucht.de offers an internal chat to contact landlords, this is an histogram with the number of chats with a certain ammount of messages, of course the chats with only one message are the ones i never even received an answer for. That single message was written by me and it's :

Hi, I am Roberto, a 22y.o Italian Computer Science student, I am coming to Muenchen in October for an MSc in Informatik and I am looking for accommodation for that period, about 2 years. Is the room still available?
Hope to hear back from you soon.

Chats with two messages are related to ads I received a single answer for, and which I decided to refuse, either because too expensive, too far from campus or beacause they looked like a scam. Dear landlords I am sorry for this, think of it as a "uno reverse card".

In case you are asking: yes, scale is logarithmic, reality is much more like this:

It is what it is, I have a success rate of 0.002976190476190476, I refused 2 ads and choose the third among my options.

type of accommodations

This is the distribution of accommodation type of my ads:


This is the distribution of prices for a month of accommodation. Yea I know, it's crazy.

What's even crazier is the distribution of sizes (in squared meters) of the accommodations:

The average price for an accommodation is 718.83 euros per month, while the average size is 21.28 squared meters. The average price for squared meter is 37.15 euros.


My campus is located in Garching bei Muenchen, it's not in Muenchen but it has a dedicated metro stop, distances from campus are however a little bit crazy, the average is 47 minutes by pubblic transport. This is the distribution:

I calculated the distance from campus by searching for coordinates from an address like M√ľnchen Maxvorstadt using the reverse-geocoding feature of openstreetmap offered by nominatim and then supplying them to mvg_api.get_route().

This is a map of the locations:

  1. yawgbot is much more than a simple bot, it's intended to be a customizable platform, you can write a plugin to scrape your own platform.