Skip to content
Faster Than Normal
Mathematics & Probability

Variance

Model #0794Category: Mathematics & ProbabilityDepth to apply:
4 min read

Contents

  1. 1. Core Idea
  2. 2. How to See It
  3. 3. How to Use It
  4. 4. Founders & Leaders
  5. 5. Connected Models
  6. 6. One Key Quote
  7. 7. Summary & Further Reading
Mathematics & Probability
Section 1

Core Idea

Variance measures how far a set of outcomes spreads from the average. High variance means results swing widely; low variance means they cluster tight. The number itself is the average of the squared deviations from the mean — squaring so that over- and under-shoots don't cancel out. For founders, variance is the difference between "this works on average" and "this works reliably." A strategy with great expected value but enormous variance can still kill you before the average materialises.

Get Faster Than Normal by email

Ideas from founders and companies.

Free newsletter. Unsubscribe anytime.

Or open the full subscribe page.

Section 2

How to See It

Revenue & Forecasting
You're seeing Variance when monthly revenue swings wildly even though the annual average looks fine. The average masks the risk — high variance means any single month could be a crisis.
Hiring & Performance
You're seeing Variance when a team's output is brilliant one sprint and mediocre the next. The average output may be acceptable, but the spread signals process or talent inconsistency.
Section 3

How to Use It

Never look at an average without asking about the spread. When evaluating strategies, channels, or hires, compare both expected value and variance. Prefer lower-variance options when survival matters; accept higher variance when you can absorb the downside and the upside is asymmetric. Reduce variance through diversification, process standardisation, and smaller batch sizes.
Decision filter
"What's the spread around this average? If the worst-case swing would hurt us badly, the average doesn't matter — reduce the variance first."
As a founder
Track the variance of your key metrics, not just the mean. When variance is high, invest in systems that narrow the spread — better processes, smaller experiments, diversified channels — before optimising for a higher average.
Section 5

Founders & Leaders

Jim SimonsFounder, Renaissance Technologies
Simons built Renaissance Technologies on the principle that understanding variance is more important than chasing returns. His Medallion Fund succeeded not by making the biggest bets but by making thousands of small, statistically validated bets where the variance of each was well understood and controlled. Simons treated unexplained variance as the enemy — every anomaly was investigated, every model stress-tested against distributional assumptions. Founders can adopt this by measuring the spread of their key outcomes, not just the average. When variance is high and unexplained, don't scale — investigate. Reduce the variance first, then compound. Simons showed that disciplined variance management turns modest edges into extraordinary, reliable performance over time.
Section 7

Connected Models

Reinforces
Standard Deviation & Normal Distribution
Standard deviation is the square root of variance — the same concept in more intuitive units. Together they describe the shape and spread of outcomes. Use both to assess whether an average is trustworthy or misleading.
Reinforces
Regression to the Mean
Extreme outcomes tend to move back toward the average over time. High variance makes extreme results more likely in any single period, but regression says don't overreact — the next period will likely be closer to the mean.
Tension
Risk-Reward Ratio
Higher expected reward often comes with higher variance. The tension: you want upside but need to survive the spread. Use risk-reward ratio to decide if the variance is worth it given your ability to absorb losses.
Section 8

One Key Quote

"Never cross a river that is on average four feet deep."
— Nassim Nicholas Taleb
Section 11

Summary & Further Reading

Variance measures the spread of outcomes around the average. High variance means unreliable results regardless of a good mean. Track both the average and the spread, reduce variance when survival matters, and accept it only when you can absorb the downside. Don't trust an average without knowing how wide the distribution really is.

Why this matters next

mental modelsRegression to the Mean

Variance applied the Regression to the Mean mental model

mental modelsRisk-Reward Ratio

Variance applied the Risk-Reward Ratio mental model

mental modelsScale

Variance applied the Scale mental model

mental modelsVariance

Variance applied the Variance mental model

mental modelsDistribution

Variance applied the Distribution mental model

mental modelsTrust

Variance applied the Trust mental model

Frequently asked questions

What is Variance?

Variance is a mental model used for better thinking and decision-making.

How do you apply Variance?

To apply Variance, identify situations where this framework is relevant, then use it as a lens to evaluate your options and decisions. The model is most useful when combined with other complementary mental models.

What category does Variance fall under?

Variance falls under the Mathematics & Probability category of mental models. Other models in this category can be found on the Mathematics & Probability hub page.

Why is Variance important?

Variance is important because it provides a structured way to think about problems that would otherwise be approached with intuition alone. Understanding this model helps you avoid common reasoning errors and make better decisions.

Continue exploring

BT

Mental model

Bayes Theorem

A mathematical framework for updating beliefs based on new evidence, proportiona

BT

Mental model

Black Swan Theory

Rare, unpredictable events with extreme impact that are retrospectively rational

CO

Mental model

Compounding

Small consistent gains accumulate exponentially over time — the most powerful fo

CC

Mental model

Correlation vs Causation

The critical distinction between two variables that move together and one actual

ER

Mental model

Ergodicity

The distinction between ensemble averages and time averages — what works across

EG

Mental model

Exponential Growth

Growth that accelerates proportionally to its current size, producing deceptivel

Get Faster Than Normal by email

Ideas from founders and companies.

Free newsletter. Unsubscribe anytime.

Or open the full subscribe page.

Popular Mental Models

First Principles ThinkingOccam's RazorCircle of CompetenceInversionConfirmation BiasSecond-Order ThinkingDunning-Kruger EffectSurvivorship BiasPareto PrincipleOpportunity Cost