Team Probability Test
Separate confidence in today's idea from confidence in the team's ability to adapt
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 95%
Before leaving RBC, Katsuyama and three colleagues faced a mismatch between low confidence in their first IEX concept and higher confidence in one another. He asked each person to write down the probability that the current idea would work. Katsuyama's estimate was only ten percent, and the others were lower. The crucial second estimate was different: they agreed there was a seventy percent chance the four of them would figure out a viable answer. That distinction changed the decision from a single-product bet into a team-and-problem bet. Their original smart-order-router concept did fail, but the team pivoted toward building a trading venue. The reusable test separates idea risk from adaptation capacity and requires independent estimates, explicit comparison, and evidence from prior joint execution.
Origin
Katsuyama created the test after his wife asked how likely the planned business was to work and prompted him to confirm that assumption with his prospective co-founders.
Core principles
- 01Confidence in an initial idea is not confidence in the venture
- 02Independent estimates reveal hidden disagreement
- 03A strong team can justify action despite a weak first concept
- 04Past problem-solving ability is evidence for adaptability
- 05The decision still requires a problem worth attempting
How to run it
- 1
State the present idea
Describe the exact business model or solution being evaluated now. Prevent confidence in the broader mission from being mistaken for confidence in this implementation.
Pro tip Include the customer behavior the idea depends on.
Watch out A vague idea produces meaningless probability estimates.
- 2
Estimate independently
Have every core teammate write down the chance that the current idea succeeds before discussing it. Reveal the numbers together.
Pro tip Use written estimates to reduce anchoring on the most senior person's view.
Watch out A group estimate can conceal major private doubts.
- 3
Score team adaptability
Separately estimate the chance that this team can find another workable solution if the present idea fails. Do not let this score overwrite the first one.
Pro tip Ask what the team would still possess after the first concept fails.
Watch out Optimism about friendship is not evidence of execution ability.
- 4
Review the evidence
Examine the team's record of solving related problems, combining expertise, and staying together under pressure. Use demonstrated behavior rather than aspiration.
Pro tip List concrete products, discoveries, or crises the team handled jointly.
Watch out A strong individual résumé does not prove the group works well together.
- 5
Make the two-level bet
Proceed only if the problem matters, the downside is acceptable, and confidence in adaptation justifies the weak initial idea. Record likely pivot triggers.
Pro tip Treat the current concept as a hypothesis rather than a promise.
Watch out Team confidence does not remove financial or personal downside.
In the wild
Katsuyama estimated only a ten percent chance that the team's initial plan—routing other banks' orders—would work; his colleagues estimated even less. They nevertheless agreed on a seventy percent chance that the four of them could solve the broader problem. The first model proved too invasive for banks that already owned routers, so the team pivoted to creating a venue.
→ Low confidence in the first idea did not prevent a proven team from pursuing and adapting around the larger problem.
Common mistakes
Conflating the idea and the team
A weak first idea does not automatically imply a weak team, and a strong idea does not prove the team can execute it.
Estimating aloud as a group
Discussion before scoring invites anchoring and conformity. Independent written probabilities expose the real spread of belief.
Calling confidence evidence
The IEX team had already diagnosed market structure and built Thor. Adaptability should be supported by demonstrated joint work.
Is it for you?
Best for
It is best for proven teams considering a high-uncertainty venture whose initial product or model may need to change.
Not ideal for
It is not ideal for an untested team with no evidence of complementary execution or willingness to pivot.
From the transcript
“guys, write on a piece of paper, what percent chance you think this idea?”
“I was the highest at everyone else was lower”
“what we all agreed was that there was a seventy chance that the four of us would figure this out.”
From the episode
IEX Group Brad Katsuyama: Masters of Business (Audio)
Brad Katsuyama