Three-Model Triangulation
Apply three different mental models to a problem; agreement across them sharply raises your odds of being right.
- Difficulty
- Moderate
- Time to result
- ~days to results
- Steps
- 3
- Confidence
- 85%
Every mental model works by reducing a high-dimensional reality down to a few dimensions — which risks discarding a crucial one. Applying multiple models that simplify along different dimensions guards against that: if three different reductions converge on the same answer, your confidence should rise steeply, and even a second model helps enormously.
Origin
Söderström attributes this to Charlie Munger via 'The Complete Investor'; he frames it as the book's key takeaway about mental models.
Core principles
- 01A model's value is dimensionality reduction — infinite dimensions down to three or four
- 02Any single reduction risks dropping a vital dimension (e.g. pandemic risk)
- 03Models that reduce along different dimensions cross-check each other
- 04Even the second independent model vastly increases your chance of being right
How to run it
- 1
Frame the problem through one model
Pick a mental model and reason to a conclusion, accepting that this model has collapsed the problem down to a handful of dimensions.
Watch out A single model may have quietly discarded the one dimension that actually decides the outcome.
- 2
Re-analyze with a model that reduces differently
Choose a second model whose simplification keeps different dimensions than the first, and work the problem again.
Pro tip Pick models that are genuinely different in what they emphasize, not variations that drop the same dimensions.
- 3
Add a third and look for convergence
Apply a third differently-structured model. When independent reductions agree, treat that convergence as strong evidence you're right; when they diverge, investigate the dimension responsible.
Pro tip Even reaching just two agreeing models materially increases your odds — you don't always need to stop at exactly three.
In the wild
Söderström describes 'The Complete Investor' as ostensibly about investment but really a toolkit of mental models, whose central lesson is to attack any problem with three different models because each reduces the world's dimensions differently.
→ Convergent conclusions across differing models give far higher confidence than any single model's answer.
Common mistakes
Relying on a single mental model
One model reduces reality along one set of dimensions and may have thrown away the decisive one, so its conclusion can be confidently wrong.
Using three models that reduce the same way
If the models drop the same dimensions, their agreement is illusory — they share the same blind spot and don't actually cross-check each other.
Is it for you?
Best for
Decision-makers facing complex, high-stakes calls who want to stress-test a conclusion before committing.
Not ideal for
Fast, low-stakes, reversible decisions where the overhead of multi-model analysis isn't worth it.
From the transcript
“you should always apply three different models to it because what models do is they they um simplify and and reduce dimensionality”
“even the second model you apply vastly increases your chances that you're right.”
“the world has probably infinite dimensions and it reduces to maybe three or four”
From the episode
Lessons from scaling Spotify: The science of product, taking risky bets, and how AI is already impacting the future of music
Gustav Söderström (Co-President, CPO, and CTO at Spotify)