Methodology
No black box. Here is exactly how LifeOdds turns your assumptions into a recommendation — and exactly where it stops being reliable.
Distributions, not point estimates
Every uncertain input is a probability distribution — Normal, LogNormal, Beta, Gamma, Student-t, Bernoulli or Uniform — so the model reasons about a range of futures instead of one confident guess.
Monte Carlo simulation
Each run draws one sample from every variable and plays the decision forward over your time horizon. Doing this thousands of times builds up the full distribution of outcomes for each option.
Net present value
Cash flows in each future are discounted back to the present and summed into a net present value, so money later is worth less than money now.
Risk-adjusted utility
Options are scored by their certainty-equivalent — the mean NPV minus a penalty on volatility — alongside the probability of loss, the share of futures where the option ends up behind.
Conviction
The option with the best risk-adjusted score wins, and LifeOdds reports a conviction score for how decisively it won across the simulated futures.
Honest limitations
Garbage in, garbage out — the model explores your assumptions and cannot know what you left out. Variables are sampled independently, the wellbeing-to-money conversion is a fixed heuristic, and none of this is financial, legal or career advice.