Monkey Genealogy

Every trading day the worst-performing tenth of the population is culled, and the best-performing tenth breed to replace them. Children inherit a blend of both parents' strategy genes, plus the occasional mutation. This is the record of who lived, who died, who they were descended from, and whether any of it actually helped.

Does selection actually work?

Read this before believing the chart. Culling losers makes any population look brilliant in hindsight — the survivors are winners because the losers were deleted, not because anything improved. That is survivorship bias, and it is the single most common way a backtest lies. So twenty of the hundred monkeys are a control group: they never die, never breed, and never change. If the evolved population isn't beating them, then breeding accomplished precisely nothing and the gap you see elsewhere on this site is an artefact of deleting the evidence.

Strategy drift across generations

What the population is made of, generation by generation. Strategies that lose money get culled out; strategies that survive spread. Note that nothing ever goes extinct permanently — every monkey carries dormant genes for all twelve strategies, so a family of momentum traders can quietly pass down a talent for breakouts for generations before a mutation switches it back on.

Bloodlines

Generations

Obituaries