Estimate how many celebrities you should list to be interesting for roughly the top 20% of your audience (the "influencers"), using a simple log-normal model.

Run popular_calc.py and answer three short questions: the size of the group you care about, how many celebrities people usually know at the low end, and how many they usually know at the high end. The tool assumes the low/high numbers describe the middle ~50% of people, fits a log-normal spread from that, and reports how many celebrities you'd need to include to engage roughly the top 20% (plus how many people that is). The calculator uses only the Python standard library.

Take K known celebrities, look at log10(K), and model that as normal. Treat your "low" and "high" answers as the 25th and 75th percentiles to pin down mu and sigma in log space. Read off the 80th percentile (top 20%) from that distribution to get the target number of celebrities to list. In this context:

• Top 20% ("influencers"): people at or above the 80th percentile of "celebrities known" in your group.
• "Popular" celebrity count: the estimated minimum number of celebrities someone would need to recognize to sit in that top 20% band, based on your low/high inputs.

Example 1 - top-20% target (N_total = 1,000,000; low = 100; high = 500):

=== Quick popularity check (log-normal model) ===

Among what group of people (amount of people, for example in your city)?: 77000
How many celebrities do people usually know at the low end (minimum): 500
Maybe more? (upper usual count, the high end people usually know): 1000

--- RESULT ---
Assuming the middle ~50% of people know between 500 and 1000 celebrities.
To be considered "popular" among the top 20% (influencers), you likely need to be known by:
+---------------+
| ~1,090 people |
+---------------+

Run python popular_calc.py and press Enter to keep the suggested numbers or type your own.