Strategic decisions are bets — commitments of resources made under uncertainty with the expectation of a favorable outcome.
The question isn't whether your decisions involve uncertainty. They always do. The question is whether you'll be explicit about that uncertainty or pretend it doesn't exist.
Most organizations choose pretense. They say "we're confident this will work" without quantifying what "confident" means. They avoid probability language because it feels like admitting weakness. But refusing to measure uncertainty doesn’t eliminate it. It makes it harder to compare stated confidence with outcomes and learn from the pattern.
What happens when you treat decisions as bets
When you assign probabilities to outcomes, even rough ones, several patterns emerge.
Assumptions surface. Saying "I'm 65% confident this will work" forces you to articulate what would have to be true for success. You can't hide behind vague optimism.
Learning accelerates. When you track predictions against outcomes, you see patterns in your judgment. You discover which types of decisions you're systematically overconfident about, which risks you underestimate, which assumptions consistently prove wrong.
Resource allocation gets clearer. Comparing bets by their expected value (probability × impact) makes trade-offs explicit. Comparing a 30% chance of transformational impact with an 80% chance of incremental improvement makes the trade-off clearer than calling both “important initiatives.”
Organizations that avoid probability language leave their confidence unmeasured. Assigning a probability makes that confidence explicit and testable.
The connection to "what would have to be true?"
Roger Martin's strategic choice framework asks: what would have to be true for this option to be the right choice?
For a new program to succeed, it would have to be true that participants value the outcome enough to commit significant time. It would have to be true that you can deliver quality at the scale required. It would have to be true that funders will support something without an established track record.
Each of those is a bet. Once you make them explicit, you can estimate odds. How confident are you that participants will commit that time? 70%? 40%? The number matters less than the discipline of assigning one.
Then you can ask: What evidence would change my confidence? What could I test quickly? Is this bet worth making at these odds?
Martin's framework surfaces the assumptions. Probability thinking forces you to evaluate them honestly.
The discomfort is the point
People resist this because assigning probabilities feels like false precision when information is incomplete. But you have enough information to make the decision, which means you have enough information to estimate odds.
Strategic decisions require judgment under uncertainty. Quantifying that uncertainty, even roughly, is an important discipline: it makes the confidence behind a decision explicit and gives later review something concrete to test.
Precision and calibration are different things. A rough probability makes a judgment explicit; it does not establish that the judgment is calibrated. Calibration is assessed across repeated forecasts by comparing stated probabilities with observed outcomes. Among outcomes assigned a 65% probability, about 65% should occur over a sufficiently large set. Recording estimates and reviewing the results makes it possible to check and improve calibration.
Strategy is gambling: committing resources before outcomes are known. The question is whether you'll be a disciplined gambler (tracking odds, learning from outcomes, improving over time) or an undisciplined one (making bets without acknowledging them as bets).
Calibration source: Hubbard Decision Research, Calibrated Probability Assessments: An Introduction
Connects to: Value of Information.