When Gen II/Johto Region species were introduced into Pokemon GO, a Nest Migration occurred – despite not occurring on the standard Migration schedule (at 12am UTC every other Thursday).
Perhaps more interestingly, the nesting species were observed to follow a pattern: each species Pokedex # was “remapped” in order, such that the species Pokedex # always increased in it’s post-Migration result.
With the introduction of Gen III (Hoenn Region) species, a similarly irregular Nest Migration schedule occurred. However, reports have been mixed in regards to whether the Migration followed a similar (or any) pattern.
Fortunately, the Silph Road has an extremely relevant data source in the Global Nest Atlas! Thanks to the hard work of our travelers around the world, we have a dataset that can help answer this question.
Visualizing The Data
In examining only those Nests which had a verified species in Migration #36 (the Migration preceding the Hoenn launch) and have a verified species in the days following the Hoenn Migration, we are able to visualize any potential patterns rather effectively. The result?
Employing a parallel coordinates plot, we can visually inspect the species in Migration #36 vs the Hoenn Migration. In the plot, we can select all Nests with a pre-Migration species (the left axis by species id) and see what species those Nests resulted in (the right axis by species id).
It is apparent that every pre-Migration species has been mapped to a significantly diverse spread of post-Migration species. This is in stark contrast to the Johto migration, where each species id # only increased in the Migration.
The conclusion evident in the data is: the Hoenn Migration did not follow a similar pattern to the Johto Migration.
You can play with this data visualization yourself below! This visualization is only manipulable on desktop (so hop on a laptop to give it a try if you’re on mobile). Simply:
- To Filter Species: Click + drag along either axis to apply a filter (be patient with your CPU!)
- To Remove a Filter: click (and don’t drag) to remove an axis’ filter
A level of inaccuracy is inherent in all citizen-science data gathering, like the Nest Atlas employs. However, a majority of the data collected on the Nest Atlas is sufficient to properly visualize the patterns we’re searching for.
With the Johto Migration pattern, for example, a parallel coordinates plot showed an overwhelmingly clear pattern (notice Migration #14):
Compare this to the Hoenn Migration, and the difference is striking!