Base Rates: The Forgotten Foundation of Better Predictions

Introduction

Base rates are the forgotten foundation of better predictions. A base rate tells you what usually happens before you decide what will happen in this particular case.

That may sound basic, but it is one of the most ignored ideas in everyday judgment. People often jump straight to the specific story in front of them. A founder has a brilliant pitch, so the startup feels likely to succeed. A job candidate speaks confidently, so they feel likely to perform well. A friend starts a demanding habit, so this time it feels like the change will stick.

The problem is not that specific evidence is useless. The problem is that specific evidence becomes dangerous when it floats without context. Base rates give you that context. They ask: what normally happens in situations like this?

If you want better predictions, start with the outside view before you personalize the case. Look at the category, the historical pattern, and the typical outcome. Then adjust carefully based on the evidence you actually have.

What Are Base Rates?

A base rate is the general frequency of an event or outcome in a broad reference class.

In plain English, it is the background odds.

For example:

• the percentage of startups that survive five years
• the average time a software project takes compared with its initial estimate
• the typical success rate of a new diet or exercise routine
• the historical return range for a type of investment
• the normal performance of people hired through a certain process

The base rate does not tell you exactly what will happen in one individual case. It gives you a starting point. From there, you can adjust up or down if you have strong evidence that the specific case is different.

This is why base rates pair naturally with probabilistic thinking. They help you move from "I feel sure" to "What is likely, given what usually happens?"

Why Base Rates Matter

Base rates matter because the human mind is easily captured by stories.

When a case feels vivid, personal, or emotionally charged, we tend to overweight the details we can see. We imagine the ambitious founder, the exceptional student, the charming candidate, or the product that feels different from everything before it. The story becomes so compelling that the background odds disappear.

This creates predictable mistakes:

• we overestimate rare success because the success stories are more visible
• we underestimate common failure because failure is less fun to study
• we confuse confidence with evidence
• we treat our case as special before proving that it is special
• we make forecasts from anecdotes instead of patterns

Base rates do not remove judgment. They discipline judgment. They force you to earn your exceptions.

If the usual outcome is poor, you can still believe your case will do better. But you should be able to explain why. What evidence changes the odds? Is it durable? Is it measurable? Has it worked in similar cases before?

Without those questions, optimism quietly becomes guesswork.

The Inside View vs the Outside View

One of the cleanest ways to understand base rates is the difference between the inside view and the outside view.

The inside view focuses on the specific case:

• our plan
• our talent
• our motivation
• our product
• our constraints
• our reasons this time is different

The outside view focuses on the reference class:

• similar projects
• similar teams
• similar markets
• similar decisions
• similar timelines
• similar outcomes in the past

Both views matter, but most people overuse the inside view. It feels richer because it contains details. It feels respectful because it treats the current situation as unique. It also flatters us because our own intentions are always visible from the inside.

The outside view is less dramatic but often more accurate. It asks how comparable situations usually turn out. That single shift can improve forecasts because it anchors you to reality before your story takes over.

A Simple Example: Project Timelines

Suppose a team estimates that a new feature will take four weeks.

The inside view says:

• the scope is clear
• the team is strong
• the design is almost done
• everyone is motivated
• the deadline feels achievable

The base-rate question is different: how long do similar features usually take this team?

Maybe the answer is eight weeks. Maybe the last five comparable features were all estimated at four to six weeks and delivered in seven to ten. That does not prove the new estimate is wrong, but it should make the team cautious.

The base rate reveals something the inside view hides: optimism may be part of the process.

A better forecast might start at eight weeks and then adjust if there are real reasons this feature is simpler than the past examples. The team may still choose a faster deadline, but now it knows it is making an aggressive bet rather than a neutral estimate.

This is a practical use of base rates. They turn vague confidence into visible assumptions.

Base Rates in Hiring

Hiring is full of base-rate neglect because interviews create vivid impressions.

A candidate may speak clearly, tell a strong story, and seem like an obvious fit. That information matters. But before you lean too hard on it, ask what usually predicts success in this role.

Useful base-rate questions include:

• How often do candidates who interview well succeed after six months?
• Which past hiring signals actually predicted performance?
• What failure patterns repeat in this role?
• How often have we mistaken confidence for competence?
• What does good performance look like after the first few months?

These questions are less exciting than a great conversation, but they improve judgment. They remind you that interviews are only one signal. Work samples, reference checks, trial projects, and past performance may have stronger predictive value.

Base rates do not mean you should ignore the person in front of you. They mean you should interpret the person through a more honest lens.

Base Rates in Startups and Business

Startup stories are especially vulnerable to base-rate neglect.

Every founder has reasons the company will beat the odds. The market is changing. The product is better. The timing is right. The team is hungrier. The incumbents are slow. Sometimes those reasons are true. But the starting point should still be the base rate: most startups fail or produce modest outcomes.

That does not mean nobody should start a company. It means the decision should respect the odds.

Base-rate thinking helps founders and investors ask sharper questions:

• What usually kills companies in this category?
• How many similar products reached distribution?
• What customer behavior has to change?
• What evidence shows this team can overcome the normal failure points?
• Are we seeing traction, or only enthusiasm?

This model also protects against the most seductive business phrase: "This time is different." Sometimes it is. Usually, the burden of proof is higher than the story suggests.

Base Rates in Personal Decisions

Base rates are not only for statistics, finance, or strategy. They are useful in ordinary life.

Imagine you want to build a new habit. The inside view says this time feels different. You bought the equipment, cleared the calendar, felt the frustration, and made a promise to yourself.

The base-rate question is: how often have similar habits lasted under similar conditions?

If the honest answer is "rarely," that is not a reason to give up. It is a reason to redesign the system. Maybe the plan is too ambitious. Maybe the trigger is unclear. Maybe you need social accountability, a smaller starting point, or a lower-friction environment.

Base rates turn self-judgment into design work. Instead of saying "I failed because I lack discipline," you can say "My plan looked like past plans that did not survive contact with normal life."

That is a much more useful conclusion.

Common Mistakes With Base Rates

Base rates are powerful, but they can be used badly.

Mistake 1: Treating the base rate as destiny

A base rate is a starting point, not a sentence. Some people and systems really are different. The question is whether the evidence is strong enough to justify moving away from the baseline.

Mistake 2: Choosing the wrong reference class

The quality of the base rate depends on the category you choose. "Businesses" is often too broad. "Local service businesses with low fixed costs" may be more useful. "Software projects" is broad. "New features built by this team with unclear requirements" is better.

If the reference class is too loose, the prediction becomes blurry.

Mistake 3: Ignoring new evidence

Base-rate thinking should not make you rigid. If strong evidence arrives, update. A product with real retention data deserves a different forecast than a product with only a pitch deck. A candidate with a strong work sample deserves more confidence than a candidate who only interviews well.

Mistake 4: Using base rates to sound cynical

Some people use base rates as a way to dismiss ambition. That misses the point. The model is not anti-ambition. It is anti-delusion. It helps you see the hill clearly before you decide to climb it.

How to Apply Base Rates in Real Decisions

You can use base rates with a simple sequence.

1. Define the prediction

Be specific about what you are trying to forecast. "Will this work?" is too vague. Ask something like, "Will this project ship within six weeks?" or "Will this habit still be active after three months?"

2. Find the reference class

Look for similar cases. The reference class should be broad enough to provide evidence and narrow enough to be relevant.

For a product launch, you might look at similar launches in the same market. For a personal habit, you might look at your own history with similar routines.

3. Identify the base rate

Ask what usually happened in that reference class. What percentage succeeded? How long did it usually take? What failure modes appeared most often?

You do not always need perfect data. Even a rough historical pattern is often better than starting from emotion.

4. Adjust for specific evidence

Now bring in the inside view. What is genuinely different about this case? Is the team more experienced? Is the scope smaller? Is the market stronger? Is there proof, or only a feeling?

Adjust gradually. Strong evidence can move the forecast a lot. Weak evidence should move it only a little.

5. Size the decision to the uncertainty

If the base rate is poor and your evidence is early, make a smaller bet. Run a test. Keep options open. Build a margin of safety. If the base rate is favorable and the evidence is strong, you can justify more commitment.

This is where base rates become practical. They do not just make predictions more accurate. They help you decide how much to risk.

Final Thoughts

Base rates improve predictions because they make you start with what usually happens. That one habit can prevent a lot of avoidable mistakes. It slows down vivid stories, weakens overconfidence, and gives your judgment a realistic anchor.

The best use of base rates is not cold pessimism. It is disciplined optimism. You can still believe in the unusual case, but you should know the usual pattern first.

If you want a deeper framework for using mental models in everyday decisions, 100 Mental Models expands on these ideas in a broader and more practical way.