Projects

Not All Those Who Wander Are Lost

Some of my ideas are good. Some of them are even worth implementing and a distinct few are worth sharing.

Jiggling Atoms

Thermodynamics is just statistics in a trenchcoat. Incredibly simple gas models (written in Python, so yes, really simple) yield surprisingly realistic behaviour, and they are great for building intuition. Entropy, osmosis, thermal expansion and why blowing on soup makes it cool faster all become intuitive when thinking of jiggling balls.

The Corner in VO2 Kinetics

If you ask a bunch of people to exercise (run or cycle) at maximal intensity for four minutes and collect their respiration data, you can do a lot of fun things. Like fitting two lines to the data and writing it up as a paper (in progress).

Sonntagsfrage Analysis

Election polls are important to powerful people; therefore, there are a lot of them. The aim of the project was to spot systematic bias and find meaningful correlations to economic indicators. Turns out that things following the same trendline strongly correlate. Who would have thought...

Ideas

  • PolyAnalytics

    I tried analysing economic and technological development over the course of a game of Polytopia. Trouble getting access to enough game IDs killed the project.

  • Poll-y-Market

    I want to compare how well calibrated Polymarket markets are against established betting odds, and how that calibration evolves over time. Work in progress.

  • Modular Flowerpots

    I designed a series of 3D-printable modular flowerpots that click together. They come in different sizes, but they are a pain to get watertight. That sank the project.

About these projects

All of this is built in my free time, mostly in Python, and aimed at understanding rather than polish. The code is on GitHub: use it, break it, and if you spot a mistake or have an idea, I'd love to hear about it.