Last week I had an idea and decided to see where it would lead.
For the past three days, I've been working on a new interactive dashboard to visualize the journey of Europa series plates across the continent, and I'm excited to share it with you:
https://shortokapi.github.io/plate-dispersion-mapThe Problem with Raw Data:
As we all know, raw EBT data is biased. If we simply plot where notes are found, the map may show us where the most active users live, rather than where the notes actually traveled.
The Solution:
This tool cross-references EBT database with Guy Sohier's print run estimates, and applies a statistical algorithm called Location Quotient (LQ).
In simple terms: it calculates a "baseline" for your country's tracking power, and then measures if a specific plate is over-represented or under-represented in your area compared to that baseline. This balances the map, allowing us to see the true organic migration of the notes!
It also features an Anomaly Detector: if a plate is captured at a rate wildly above the European average (e.g., 10x what would be normal), the system flags it as "Anomalous Data" (this is what's happening with the famous Trichet Europa,
Z002).I hope you enjoy exploring the data as much as I enjoyed building this. If you are curious about the math and methodology behind it, I've written a detailed explanation on the project's GitHub page.
Let me know your thoughts and if you find any interesting plate behaviors!
As a little teaser, here is an assortment of screenies from the dashboard.
- 5M003.png (362.74 KiB) Viewed 446 times
- 10N012.png (340.28 KiB) Viewed 446 times
- 20R030.png (338.96 KiB) Viewed 446 times
- 50S037.png (335.36 KiB) Viewed 446 times
- 100W005.png (330.29 KiB) Viewed 446 times
- 200E005.png (352.34 KiB) Viewed 446 times
