giovanni

A CLI tool used to calculate the likelihood of a shiny Pokemon encounter for a given number of resets, or simulate the distribution of required SRs for a given generation of Pokemon.

Encounter odds:

Compute the probability of encountering a shiny in Gen 4 with a single encounter:

$ python -m giovanni odds --gen 4 --verbose
[INFO] 2020-12-22 16:07:54 [giovanni.cli.odds] - Probability of shiny encounter after 1 soft reset: 0.012%

The probability of encountering a shiny in Gen 6 with 5000 SRs and a shiny charm equipped:

$ python -m giovanni odds --gen 6 --soft_resets 5000 --charm
[INFO] 2020-12-22 16:07:57 [giovanni.cli.odds] - Probability of shiny encounter after 5,000 soft resets: 91.301%

Montecarlo simulation:

A simulation of Gen 6 soft resets required to encounter a shiny:

$ python -m giovanni simulate --gen 6 -n 5000
simulation: 100%|██████████████████████████████████████████████████████████████████| 5000/5000 [00:13<00:00, 365.02it/s]
[INFO] 2020-12-22 15:53:43 [giovanni.cli.simulate] - Montecarlo simulation results --
Average     3980.136
Std Dev     3983.669
Max        32989.000
Min            1.000
Name: Required encounters (gen=6, num_trials=5000, swarm_size=1), dtype: float64

To simulate, e.g., SS/HG soft resets required to encounter a shiny starter use the swarm_size argument:

python -m giovanni simulate --gen 4 -n 5000 --swarm_size 3
simulation: 100%|██████████████████████████████████████████████████████████████████| 5000/5000 [00:15<00:00, 329.87it/s]
[INFO] 2020-12-22 15:51:01 [giovanni.cli.simulate] - Montecarlo simulation results --
Average     2737.669
Std Dev     2717.707
Max        22760.000
Min            1.000
Name: Required encounters (gen=4, num_trials=5000, swarm_size=3), dtype: float64

Notes

• No considerations are made around shiny locks. Make sure that you are aware whether the 'Mon you're hunting can be found in shiny form.

• This project only covers the main titles in the game series. It does not cover offshoots or mobile games in the series.

• The current implementation only takes into account base shiny rates by generation (see future work)

Future work

A list of things that would be cool to implement:

• Considerations around encounter method:

  • Wild/SR
  • Masuda
  • Chaining
  • Etc.

• Gen 8

Installing

Installing from PyPi:

giovanni is available on public PyPi:

$ pip install giovanni

Get setup for local dev:

Set up by cloning and building from Git:

$ git clone https://github.com/tgsmith61591/giovanni.git
$ cd giovanni
$ make requirements lint-requirements testing-requirements

Run a quick test to make sure everything is copacetic:

$ make test