A 13-module decision-making dashboard. Decision pyramid, pre-mortem, deep decision log, decision tree, opportunity cost and expected value calculators, regret minimisation, Bayesian thinking, mental models library. Single HTML file. Runs offline. Owned forever.
À la carte value at catalogue rates: $400+. Yours for $77.
Not all decisions deserve the same process. Match the rigour to the stakes. Most operators apply too much process to small decisions and too little to big ones.
Imagine the decision has been made and 12 months have passed. It failed catastrophically. Tell the story of how it failed — in detail. Most decisions fail in predictable ways; the pre-mortem surfaces them while you can still adjust.
Every Tier 1 and Tier 2 decision goes in here. Future-you reads it. The pattern of how you decided 6 months ago tells you more about how to improve than any retrospective.
For complex sequential decisions. Each node is a decision point. Each branch is an option. Each leaf is an outcome. Drawing the tree alone often reveals options you didn't know existed.
Every "yes" is a "no" to the next-best alternative. Quantify what you're giving up. The hidden cost is what you stopped doing — usually invisible, always real.
EV = sum of (probability × outcome). The math version of "is this worth it?" Useful when outcomes are probabilistic. Less useful when probabilities are guesses — but even rough EV beats no EV.
| Scenario | Probability (0-1) | Outcome ($) | EV contribution |
|---|---|---|---|
| — | |||
| — | |||
| — |
Bezos's framework. Project to age 80. Imagine looking back. What would you regret more — doing this and it not working, or never trying it? The framework is biased toward action; that bias is usually correct for ambitious decisions.
Your beliefs should change when evidence changes — proportional to the strength of the evidence. Most people update too slowly (anchored to first impression) or too fast (overweight recent data). The discipline is conscious updating.
Standard pros/cons treats every factor equal. Weighted version forces you to assign importance — exposes which factor is actually driving the decision.
| Pro | Importance (1-10) | Weighted |
|---|---|---|
| 5 | ||
| 5 | ||
| 5 | ||
| 5 |
| Con | Importance (1-10) | Weighted |
|---|---|---|
| 5 | ||
| 5 | ||
| 5 | ||
| 5 |
Bezos: "Type 1 decisions are one-way doors. Type 2 are two-way doors." Most decisions are Type 2 — reversible. We treat too many decisions as Type 1, slowing us down on things that don't matter.
10 high-leverage models. Reach for them when you're stuck. They cut decision space dramatically.
Charlie Munger's tool. Don't ask "how do I succeed?" Ask "how do I fail?" Then avoid that. Often easier to identify and act on.
Strip the problem to its physics. Don't say "this is a marketplace, marketplaces work like X." Say "what does this specific situation actually require?"
"If I do X, what happens?" That's first-order. "And then what?" That's second-order. Most bad decisions look great at first order.
Take bets where the worst case is bounded but the best case is huge. The arithmetic favours you over many trials.
Most "they're attacking me" situations are actually "they didn't think about me." Cheaper to assume incompetence first.
The right question is "given where I am now, what's the best move forward?" Not "I've already spent $X, I have to keep going."
"It will succeed" is wrong. "It has a 60% chance of breakeven, 30% upside, 10% loss" is right. Acting on point estimates is acting on fictions.
Bouncing between strategies looks decisive. It's actually cumulative loss. Stay with a working strategy until evidence forces change — not just because the new one is shinier.
Weight advice by who pays the cost when wrong. Free advice from someone with no exposure is the cheapest because it's also the lowest-quality.
Most decisions don't matter at the 10-year horizon. Some do. The rule sorts which is which.
Quarterly review of the decision log. The single highest-leverage operator practice. You see your patterns — what you decide well, what you decide poorly, where you predict accurately, where you don't.
Seven personal decision rules. Read them quarterly.
Yours forever. Print individual modules from your browser.