FACTS · Is faster better?
The film states no real-world facts. Every number on screen and in the narration is the output of one run of the toy
simulation in filmsim.js (seed 11), recomputed on every build. It describes this simulation, not real animals.
The rules (filmsim.js, FILM_P)
- 160 Pips in generation 1, speeds spread evenly over 0.6–1.9; board ±14 units; 230 beans every day; Pips land at random each morning.
- A Pip heads for the nearest bean within 3.2 units, otherwise wanders; it spends k × speed² energy per second (k = 0.7, starting energy 1) and stops when it runs out.
- 0 beans → gone; 1 bean → lives to the next day; 2 beans → lives and leaves a copy whose speed is mutated (Gaussian, σ 0.09, floor 0.3).
Numbers and how they come out of the run
| On screen / narration | Value | Notes |
|---|---|---|
| Pips on day 1 | 160 | |
| Day 1 verdict | 55 gone, 51 one bean, 54 two beans → 159 | the head count rolls from 160 to 159 |
| "three times as many children" | slow (< 0.9) 19, fast (> 1.5) 6 | parent-speed groups of 35 and 41 Pips; per Pip the gap is larger |
| The fast Pip, the slow Pip | speeds 1.69 and 0.72 | ids 149 and 126 in the run |
| Generation 9 close-up | 20 fast (> 1.5) Pips left; generation 1 had 41 | |
| Mean speed | 1.25 (generation 1) → 0.73 (generation 20) | 208 Pips in generation 20 |
| "≤ 0.5" bin | 44 Pips | 20 of them sit at the simulation's speed floor of 0.3 |
The energy constant k = 0.7 was chosen so that the effect shows within 20 generations; at k = 0.5 the speed distribution barely moves in that time. The rules are otherwise unchanged by that choice.