In principle, with enough data on each spawn point, we can get a good estimate of the spawn table distribution and thus cluster the spawnpoints. This has been done before for Boston: https://www.reddit.com/r/TheSilphRoad/comments/5etwz9/analysis_identification_of_potential_biomes_by/ It seems feasible if we can automatize the process and dynamically cluster the spawnpoints once in a while.
In principle, with enough data on each spawn point, we can get a good estimate of the spawn table distribution and thus cluster the spawnpoints. This has been done before for Boston: https://www.reddit.com/r/TheSilphRoad/comments/5etwz9/analysis_identification_of_potential_biomes_by/
It seems feasible if we can automatize the process and dynamically cluster the spawnpoints once in a while.