At Aarhus university, a couple of the students from the common kitchen used to wander to the coffee lounge together to get a chat and a common break. However, as time went by, they got their own offices, and timing when to go to the coffee lounge became harder. But, as the computer scientist they are, they knew that the answer was the internet, and so the "Coffee ping chat" was born.

In this chat, one can send the coffee emoji, to symbolize that one is going to the coffee lounge. Similarly, others may respond with the coffee emoji, to symbolize that they will join and meet you at the coffee lounge. This over time leads to quite a lot of data. In this project, I have pulled this data, and analysed the fequency of the pings, and how the groups interact. This is then printed in the danish student magazine "Mads Føk".

Available at: https://www.madsfoek.dk/udgivelser/51-1.pdf#page=12

This article covers the first year of pings from 2022-2023. The frequency analysis is based on weekdays, months, as well as hours within a week and all days of the year. The group analysis defines that pings within 5 minutes is in the same group, and then analyses the group sizes.

Available at: https://www.madsfoek.dk/udgivelser/52-1.pdf#page=10

This article covers the second year of pings from 2023-2024. The frequency analysis is similar to the first article. Further, this articles dives into the emoji encoding of the coffee ping, as 3 additional unicode bytes extend each emoji. It is shown that the additional bytes are a variation selector, and using this information, a heap of missing pings from the previous year is found and discussed.

Available at: https://www.madsfoek.dk/udgivelser/52-2.pdf#page=10

This article covers the collected pings over the years 2022-2024, as well as analyzing the difference between the two periodes. This includes comparing the expected number of pings per day to a normal distribution and a cummulative plot of pings over the years, to locate when major downfalls in the number of pings is present. This article also looks deeper at a group analysis, and establishes the relationsship between each pair of (anomymized) person of the chat, with reguards to common coffee groups.

Available at: https://www.madsfoek.dk/udgivelser/53-1.pdf#page=20

This article covers the collected pings over all three years from 2022-2025, including a short discussion of the differences and similarities of the years. A new analysis on the groups is performed, where first it is noted that counting groups of different sizes directly is skewing the data. This skewness arrises from twice as many people being needed for a two person group over a single person group. The analysis is therefore altered to focuses on the number of people in total of the different group sizes. This is then used to show how the group sizes are distributed over the hours of a day. Further, a social credit system and later elo ranking is calculated for each member of the chat, to try and quantify how much each person is pulling other people along to the coffee lounge. The basis for these calculations i the ordered ping of each group.