Two-Sided Conviction Test
Argue both sides before committing capital
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 96%
The test treats analysis as a structured argument rather than a search for evidence supporting an initial preference. The decision-maker first builds the best case for an investment or strategy, then constructs and defends the strongest opposing case. A partner can challenge the original thesis, after which both participants deliberately switch positions. This exposes hidden assumptions, reveals which evidence genuinely changes the conclusion, and prevents confidence from hardening into identity. The final judgment is rarely either polar extreme; it is a calibrated position that reflects evidence from both sides. Repeating the exercise as new facts emerge creates a continuing feedback loop and helps distinguish durable conviction from mere attachment to an idea.
Origin
Tom Gayner described arguing ideas in both directions with Steve Markel as a technique for reaching a more robust understanding.
Core principles
- 01Every investment thesis has a credible opposing case.
- 02Switching sides exposes assumptions hidden by advocacy.
- 03Robust conclusions usually sit between polar extremes.
- 04Conviction should follow challenge rather than precede it.
- 05Analysis remains an ongoing practice, not a final answer.
How to run it
- 1
State the initial thesis
Describe the proposed decision, its causal logic, and the evidence required for it to succeed.
Pro tip Write the thesis before researching objections so its assumptions remain visible.
Watch out A vague thesis cannot be tested meaningfully.
- 2
Construct the opposing case
Develop the strongest plausible argument against the thesis rather than a weak objection that is easy to dismiss.
Pro tip Use evidence and incentives that a thoughtful opponent would recognize.
Watch out A straw-man bear case creates false confidence.
- 3
Switch positions
Defend the side you initially opposed and require your counterpart to defend your original view.
Pro tip Continue until each person can explain the other side fairly.
Watch out Do not use the exercise merely to rehearse a predetermined conclusion.
- 4
Find the calibrated conclusion
Identify which claims survived challenge and choose an exposure or action proportional to the remaining uncertainty.
Pro tip Record what evidence would cause another change of view.
Watch out The answer need not be a binary yes or no.
- 5
Reopen with new evidence
Return to both cases when facts change, treating the decision as something to keep working rather than permanently solved.
Pro tip Schedule reviews around meaningful evidence, not market noise.
Watch out Constant reconsideration without new information can become indecision.
In the wild
An analyst believes a platform will benefit from network effects. She first documents the growth case, then argues that customer concentration, regulation, and switching behavior could prevent those effects. A colleague takes the bullish side while she defends the bearish one. They retain the investment but reduce its size and define evidence required before adding capital.
→ The team reaches calibrated conviction instead of an untested bullish position.
Common mistakes
Building a weak opposing case
Refuting superficial objections does not test the assumptions that actually determine the outcome.
Treating disagreement as disloyalty
Personalizing the challenge prevents participants from changing positions and learning from the exercise.
Forcing a polar conclusion
Evidence often supports a conditional or sized position rather than complete certainty on either side.
Is it for you?
Best for
It is best for investors and leaders making uncertain decisions with credible evidence on both sides.
Not ideal for
It is not ideal for emergencies that demand immediate action or choices governed by an unambiguous rule.
From the transcript
“any pancake, no matter how thin, has two sides.”
“And that was just the technique to get more robust understanding of what might in fact be the case, and usually it's not the polar…”
“you shouldn't be bullish on anything unless you understand the bear case, and vice versa”
From the episode
Thomas S. Gayner on Things That Matter in Markets (Podcast)
Thomas S. Gayner