[sei]

[the genius filter]

How to Make Decisions When You Can’t See the Future

Most decision-making advice assumes you can assign probabilities.

Run the numbers. Weigh the odds. But the choices that matter most arrive without data sheets or historical patterns. Career pivots. Market shifts. Crises nobody saw coming.

Frank Knight, an American economist, identified this problem in 1921. In Risk, Uncertainty, and Profit, he drew a line that most people still ignore: risk can be measured; uncertainty cannot.

Risk is the casino, where the house knows all the odds. Uncertainty is the fog beyond the edge of the map, where probabilities don't exist at all. Confuse the two, and you risk it all.

So when the map runs out, the greatest decision-makers stop pretending they have one. Instead, they adapt, and they endure.

In this issue, we’re breaking down Knightian Uncertainty and mapping out how to move forward when you can't know the way.

[the spark]

What to Do When the Numbers Stop Working

Knight pointed out what most economists would rather ignore: You can’t hedge against the unknown.

You can buy insurance, buff up positions, and calculate expected value, but true uncertainty has no price because there's nothing to price. The probabilities just don't exist.

Knight’s insight came from watching entrepreneurs jump into the fray. The ones that came out the other side weren’t necessarily the ones that jumped in with the best map in hand.

Because at some point during the journey, everyone would reach the edge of their map, that point just beyond the margins, and be faced with a paralyzing choice: Turn back, or march on into uncertainty. And the only way out is through.

When a founder launches into an untested market, there's no historical data to analyze. No comparable cases. No odds to weigh. But decisions have to happen anyway.

Knight identified three strategies that work when prediction fails.

1. Heuristics are rules forged from experience. A venture capitalist doesn't run projections on a pre-revenue startup. He looks for patterns: founder grit, market timing, product clarity. Mental shortcuts that survive contact with the unknowable.
2. Redundancy absorbs shocks. Safety nets. Cash reserves. Backup plans. It's the boring stuff that seems wasteful, until Plan A fails. Then it's the difference between adapting and collapsing.
3. Optionality preserves momentum. Sign the contract with an exit. Build the skill that transfers. The future might be unknowable, but the way you jump in determines how you land.

Knight didn't solve the problem of uncertainty, but he did show that surviving it requires different tools than managing risk. When comfy, predictable probabilities vanish, instinct and adaptability take over.

[the science]

Heuristics > complexity.

In 2011, psychologists Gerd Gigerenzer and Wolfgang Gaissmaier published research showing that simple heuristics outperform complex statistical models when facing uncertainty.

Their team studied retail managers who needed to predict which customers would purchase again. Sophisticated academics assumed the answer required Bayesian analysis or regression models.

The managers ignored those tools (and probably hadn’t even heard of them). Instead, they relied on a single rule: if a customer hadn't purchased anything within a set number of months, classify them as inactive. That's it. One cue. No equations.

The hiatus heuristic, as the researchers came to call the managers’ method, correctly classified 83% of customers at an apparel retailer. The complex statistical model, using plenty more information and extensive computation, classified only 75%. The heuristic beat out the nerd math at an airline and an online business, too.

Gigerenzer and his colleagues tested heuristics across industries and found the same pattern. Under conditions of true uncertainty, where probabilities can't be calculated, and the future can't be modeled, simple rules consistently beat optimization strategies.

When the map runs out, knowledge from experience is your best bet for making it through the fog.

[the takeaways]

1) Know When You've Left the Map​
​When you can't assign probabilities, you're facing true uncertainty, not risk. Recognizing this distinction stops you from pretending models will save you.

2) Build on Pattern Recognition​
​Build decisions on rules forged from experience, not spreadsheets. Simple heuristics outperform complex models when probabilities don't exist.

3) Create Backup Systems​
​Build reserves and redundancies before you need them. What looks wasteful during the calm becomes survival equipment when the storm arrives.

4) Protect Your Pivot
​Optionality is priceless insurance. Always keep an eye on the exits, and position yourself to adapt to whichever future arrives.

5) Question The Numbers​
​Statistical confidence breaks down outside its domain. Trust your instincts to know when to stop calculating and start relying on what you've learned through experience.