usage: review_game.py [-h]
                      [--performance-bins PERFORMANCE_BINS [PERFORMANCE_BINS ...]]
                      [--rating-bins RATING_BINS [RATING_BINS ...]]
                      [--sample-sizes SAMPLE_SIZES [SAMPLE_SIZES ...]]
                      [--population POPULATION]
                      [--num-repetitions NUM_REPETITIONS]
                      [--production PRODUCTION [PRODUCTION ...]]

Demonstrates the effect of proper sample size usage in the context of a game
with cost and payoff

optional arguments:
  -h, --help            show this help message and exit
  --performance-bins PERFORMANCE_BINS [PERFORMANCE_BINS ...]
                        The true distribution of the population's performance
  --rating-bins RATING_BINS [RATING_BINS ...]
                        The distribution the stack ranking policy assumes
  --sample-sizes SAMPLE_SIZES [SAMPLE_SIZES ...]
                        The sizes of stack ranking groups to test
  --population POPULATION
                        The total size of the organization being stack ranked
  --num-repetitions NUM_REPETITIONS
                        The number of Monte Carlo runs to use
  --production PRODUCTION [PRODUCTION ...]
                        The true production of employees in each performance
                        bin

pip install -r requirements.txt
python review_game.py --sample-sizes 8 16 32 64 128 # script that generates the data
python plot_population.py # script that creates population boxes

5,0.5,1.2,1,228.53114999999968,50.21,112.42,37.37 # Stats when stack ranking groups of 5
10,0.5,1.2,1,253.08149999999958,19.05,136.6,44.35
20,0.5,1.2,1,253.97154999999958,24.32,153.63,22.05
40,0.5,1.2,1,260.3488999999994,18.34,165.8,15.86
80,0.5,1.2,1,264.6153999999994,14.3,173.9,11.8
100,0.5,1.2,1,266.9551999999995,12.08,178.33,9.59
200,0.5,1.2,1,270.44119999999947,8.78,184.93,6.29 # Stats when stack ranking groups of 200