A world‑class poker pro wins only a bit more than half their hands—yet can earn millions. A startup founder makes all the “right” calls and still runs out of cash. In both stories, skill and luck collide. Today, we’ll step into that collision and ask: what are you really betting on?
A quote from Annie Duke, former poker pro, sets the stage: “You can make the best decision and still have a bad outcome.” Thinking in bets takes that from a slogan to a habit. It says: every decision you make today—what product to ship, which role to hire for, even whether to send that email—is a wager on how the future will unfold. Not a yes/no guess, but a probability you’re implicitly assigning. The trouble is, most of us place those wagers blindly. We act as if we’re certain, speak in absolutes, then feel blindsided when reality doesn’t cooperate. In this episode, we’ll treat your choices like explicit bets: what odds you believe, what you’re risking, and what you stand to gain. Instead of asking “Am I right?” you’ll start asking “How right do I think I am—and what if I’m wrong?”
Most people already speak the language of bets without noticing. You say there’s a “good chance” a project slips, or it’s “unlikely” a hire works out—but those phrases are fuzzy. In high‑stakes environments, that vagueness is expensive. The Good Judgment Project’s top forecasters didn’t just trust their gut; they turned “good chance” into “I’m at 65%, and here’s why,” then updated as new info arrived. The same mindset explains why companies using probabilistic planning blow fewer budgets. This episode is about dragging those implicit bets into the open so you can tune them, test them, and learn faster.
When you strip away slogans and theory, “thinking in bets” comes down to three practical moves: make your beliefs visible, attach odds to them, and size your bet to the quality of your belief.
Start with the belief. Every important choice hides a sentence like: “Launching this feature in Q2 will increase activation.” In most teams, that sentence never gets written down; it just lurks in conversation. The first upgrade is to state it explicitly, in plain language, before you act. That simple move separates what you believe from what you’re doing about it.
Next comes attaching odds—not fake precision, but a sharpened guess. Instead of “this should work,” try: “Given similar launches we’ve done, I’m roughly 60–70% confident.” You’re not trying to be perfectly calibrated on day one. You’re training a muscle: translating gut feel into something testable. Over time, those numbers become less about ego and more about coordination. A product lead at 70% and a finance lead at 30% suddenly have something concrete to align—or disagree—on.
Then there’s bet sizing: what you actually put at risk. Treating choices as graded, not binary, lets you right‑size the experiment. At 60% confidence, you might run a limited rollout or A/B test; at 80%, you might commit engineering time but still keep a rollback plan. The point isn’t to avoid being wrong; it’s to match how much you stake to how solid your information is.
This is where many people stumble on a quiet trap: outcome bias. If the launch flops, the story quickly becomes “terrible decision”; if it spikes, “genius call.” Both stories ignore the real question: did we size and structure the bet in line with what we knew then? The best poker players review hands they won and hands they lost with the same scrutiny, precisely to escape that trap.
A useful trick from finance is to think in scenarios: base case, upside, downside. Before committing, sketch: “Here’s what I think happens most of the time, here’s the plausible win, here’s the plausible pain.” You’re not predicting a single future; you’re mapping a small range and asking, “Can we live with the downside for the chance at the upside, given our odds?” Now your “yes” or “no” sits on a clear scaffold of assumptions you can later revisit, refine, or abandon.
Your calendar is already full of bets; labeling them just makes that visible. Picture a product director debating whether to prioritize performance or a new feature. They might say, “I’m leaning performance,” but that hides the structure of the bet. A clearer version sounds like: “I’m at 65% that performance wins us more revenue over 12 months than Feature X, assuming competitors don’t ship a clone first.” Now there’s a time frame, a metric, and a key condition.
One media company did this before greenlighting a new subscription tier. Instead of arguing in circles, leaders each wrote a one‑sentence “bet statement” with their own odds and conditions. Patterns emerged: they agreed on likely adoption but wildly disagreed on churn risk. That focused the next two weeks on running small tests about churn rather than endlessly re‑hashing opinions about pricing.
Your challenge this week: for three important upcoming choices, write a single, concrete sentence that starts with “I’m betting that…” Include a rough time horizon and at least one condition that would change your mind.
As AI systems flood us with percentage‑based risks—market crashes, extreme weather, health events—the real advantage shifts from having data to knowing how to “price” it. A leader who treats a 40% threat like a certainty will over‑hedge; one who shrugs at 70% will sleepwalk into avoidable crises. Thinking in bets turns those numbers into levers: adjusting plans, buffers, and timing. Over time, organizations that practice this will treat uncertainty less like a fog and more like a navigable landscape.
Over time, this mindset turns your day into a series of small experiments instead of final exams. Rather than grading yourself on wins and losses, you start noticing how well your “odds” match reality. Like tending a garden, you prune weak theses, double‑down on hardy ones, and accept that some seeds fail—even when you planted them well.
Try this experiment: For the next 3 decisions today that involve uncertainty (what to work on next, whether to speak up in a meeting, or how to respond to an email), pause and assign each option a probability out loud: “I’m 70% confident this is the best choice because X, Y, Z.” Then, before you see how it turns out, quickly jot a one-line “result check” you’d expect if you’re right (e.g., “If this was a good choice, I’ll see A, B, C by tomorrow”). When the outcome arrives, compare what actually happened to your predicted result, and adjust your confidence score for similar future decisions. This way, you’re treating each decision like a bet and using feedback to sharpen your thinking.
