Someone asks you a question in a meeting. You have no idea. In the half-second before you speak, your brain offers you a deal: fake a confident answer now, deal with the consequences never.
Take that deal often enough, and you build a reputation that looks solid from a distance and collapses the moment anyone leans on it. There is a better trade available, and the research behind it is stranger than you would expect.
A hospital study that proved the opposite of its own hypothesis
In the 1990s, Amy Edmondson was a doctoral student at Harvard Business School studying medication errors in hospital nursing teams. Her hypothesis was the obvious one: better teams should make fewer mistakes. She ran the numbers and got the reverse. As the AAMC reported on her Learn Serve Lead talk, the teams that scored highest on a team diagnostic survey reported higher error rates, not lower ones.
Rather than conclude her hypothesis was simply wrong and move on, she sat with the number and landed on a different explanation: "Better teams probably don't make more mistakes. They were the teams willing to talk about them.”
The reinterpretation became her 1999 paper in Administrative Science Quarterly and the concept of psychological safety. Which means the entire field started with someone deciding that the number in front of her was measuring something other than what she assumed.
Your error rate is not what people are judging
Nobody around you can see how often you are actually wrong. They can only see how often you say so. Silence gets read as competence right up until reality shows up, and then it gets read backwards as concealment.
So the thing you are protecting by bluffing is not your reputation for being right. What you are actually spending down is your reputation for being findable when things go wrong.
Ask yourself which colleague you trust with a genuinely uncertain project. It's almost never the one who has an instant answer for everything. Usually, it's the one whose confidence you have learned to read accurately because they told you where the edges were.
The forecasters who beat the CIA never said 100%
There is a second body of evidence worth knowing. Between 2011 and 2015, the US intelligence agency IARPA ran a forecasting tournament. Philip Tetlock and Barbara Mellers at the University of Pennsylvania entered the Good Judgment Project, staffed by ordinary volunteers. Their top forecasters, the "superforecasters," outperformed professional intelligence analysts with access to classified information by roughly 30%.
Their method was not certainty. They assigned explicit probabilities, kept score, and updated when they were wrong. Confidence expressed as a number turns out to beat confidence expressed as a tone of voice.
The three-part answer that replaces the bluff
When you do not know, run these three moves in order. It takes eight seconds.
- Anchor what you do know. Never open with the gap. "Signups are up 12% month over month, and the conversion piece is where it gets murkier."
- Name the gap precisely. Vague admissions sound like general incompetence. Specific ones sound like expertise with edges. "I don't know how much of that came from the paid channel versus organic."
- Attach a name and a date. "Pulling the channel breakdown and sending it to you by 3pm tomorrow."
Compare that to "Um, I'm not sure; let me check." Same honesty, completely different signal. One of them has a delivery date on it.
Put a number on your confidence
For the times you have a view but not proof, say the number out loud.
"I'm about 70% on this. What's driving that is the last two quarters of retention data. What would move me off it is if the churn were concentrated in one segment."
Three things happen.
- Your boss learns how much weight to put on it.
- You have pre-registered what would change your mind, which is the hardest thing to do after the fact.
- And when you turn out to be wrong, you were never wrong; you were 70%.
Vary the number honestly. A person who is 95% on everything has a broken instrument, and the people listening figure that out faster than you think. Keep a running note of the calls you numbered and how they landed. After a month you will know whether your 70% is really a 70%, which is more self-knowledge than most people at your level have.
The harder rep: "I was wrong"
Different move with a higher payoff. When a call you made did not work, say it in one flat sentence, with no defense attached and no explanation until someone asks. "The vendor I pushed for was the wrong pick.” Then move to the correction: "Going forward I'm running two quotes before recommending one."
Skip the softeners. "In hindsight" and "to be fair" are both ways of taking it back while appearing to own it.
Two habits that quietly cost you credibility
Apologizing for not knowing turns a neutral fact into a confession. Drop "sorry, I don't know" for "let me check that." Nothing has happened that requires an apology.
Then there is the fluent non-answer, the one where you talk for forty seconds around a question you did not answer. Everyone in the room who knows the topic can hear it, and those are exactly the people whose opinion of you matters most.
Your move in the next meeting you attend
Pick one meeting this week and commit in advance to using the three-part answer the first time you get a question you cannot fully answer. Write the shape on your notepad before it starts: know, gap, date.
Then do the harder half. Before Friday, send one message to a colleague or manager on something you got wrong recently. Two lines: "The call I made on [X] was wrong. Here's what I'm changing."
Watch what comes back. It will be far less painful than the version you are imagining right now, and it buys you something the confident bluff never will.
Until next time, Future Leader.
