A venture capitalist evaluates a promising startup. The founders are exceptional—Stanford PhDs who previously sold a company to Google. The product is innovative, the market is enormous, the early traction is impressive. Everything about this specific company suggests success.
The VC invests.
Here’s what the VC didn’t ask: What percentage of Series A startups return their investors’ capital? The answer is roughly 35%. What percentage return more than 3x? About 15%. What percentage become the massive winners that drive venture returns? Perhaps 5%.
The specific story was compelling. The founders really were impressive. The product really was innovative. But in venture capital, impressive founders with innovative products fail constantly. The base rate—the underlying frequency of outcomes in this class of investments—suggested a 65% chance of losing money regardless of how compelling the specific case seemed.
This is base rate neglect: the systematic tendency to ignore or underweight historical frequencies when evaluating individual cases. And it may be the most widespread cognitive error in investing.
The Inside View vs. The Outside View
Daniel Kahneman and Amos Tversky, whose research earned Kahneman the Nobel Prize in Economics, identified two fundamentally different approaches to forecasting: the inside view and the outside view.
The inside view focuses on the specific case at hand. What are the particular circumstances? What makes this situation unique? What do we know about this specific company, this specific management team, this specific market opportunity? The inside view constructs a narrative from case-specific details.
The outside view asks a different question entirely: What happened historically when people were in similar situations? Forget the specifics—what does the base rate tell us about outcomes in this reference class?
Most investors operate almost exclusively from the inside view. They build detailed financial models, analyze competitive dynamics, evaluate management quality, assess growth prospects. All case-specific. All compelling. All systematically overconfident because they ignore the outside view.
Kahneman discovered this bias in a particularly memorable way. He was part of a team developing a new curriculum for teaching judgment and decision-making in Israeli high schools. After a year of work, he asked each team member to estimate how long until they’d finish the project. Estimates ranged from 18 to 30 months.
Then Kahneman asked a different question to one team member who had experience with similar curriculum development projects: “How long did comparable projects take?” The answer was sobering—most took 7 to 10 years, and roughly 40% were never completed at all.
The inside view, focused on their specific team and specific progress, suggested 18-30 months. The outside view, based on how similar projects actually went, suggested 7-10 years or failure. (The project ultimately took 8 years.)
This pattern repeats everywhere forecasting happens. The inside view is optimistic, detailed, and wrong. The outside view is humble, statistical, and far more accurate.
Why the Inside View Dominates
If the outside view is more accurate, why do intelligent people consistently ignore it?
First, the inside view is more engaging. Stories are compelling; statistics are boring. When evaluating a company, the narrative of the founders, the elegance of the product, the size of the market—these create vivid mental pictures. The base rate is just a number.
Second, the inside view feels more informative. You’ve done careful analysis of this specific company. You know things. The base rate feels like giving up on your hard-won insights. It feels like admitting your analysis doesn’t matter.
Third, the inside view flatters our competence. If base rates determine outcomes, then analyst skill matters less. Our detailed work, our deep research, our hard-won expertise—all become less important. The inside view preserves the illusion that our individual judgment is decisive.
Fourth, the inside view is how we naturally think. Kahneman’s work demonstrates that humans are narrative creatures. We understand the world through stories, not statistics. Overriding this requires deliberate effort that most people never make.
The result: investors construct elaborate inside-view analyses while ignoring the base rates that would calibrate their confidence. They believe each investment is unique when most investments fall into well-documented patterns with well-documented odds.
Base Rates That Investors Ignore
What are the base rates that investors should know but usually don’t? The data is available but rarely consulted.
IPO Returns
The base rate for IPO investing is poor. Research by Jay Ritter, the leading academic authority on IPOs, shows that IPOs underperform comparable companies by approximately 3% annually over the five years following issuance. The median IPO underperforms even more dramatically—the average is pulled up by rare massive winners.
When you buy an IPO, you’re participating in an asset class with a negative base rate. You need specific reasons to believe this particular IPO will beat the odds—and those reasons need to be strong enough to overcome a structural disadvantage.
Most IPO buyers don’t know this. They focus on the story—the exciting company, the growth prospects, the buzz. The inside view dominates while the outside view suggests caution.
Turnaround Investments
How often do struggling companies successfully turn around? The data suggests roughly 30% of the time, varying by industry and severity of distress. That means when you identify a company “trading at a discount because of fixable problems,” there’s a 70% historical probability that the problems won’t get fixed.
Value investors are particularly susceptible to turnaround optimism. Cheap stocks feel like opportunities. The analysis shows how things could improve. But base rates suggest that most cheap stocks are cheap for good reason and stay cheap—or get cheaper.
This doesn’t mean you should never invest in turnarounds. It means you should demand a much larger margin of safety and a much higher potential return than you would for a company without turnaround requirements. The base rate tells you how much compensation you need for the elevated risk of failure.
Growth Projections
How often do high-growth companies sustain their growth? Michael Mauboussin’s research at Credit Suisse and Morgan Stanley provides the sobering data: Companies growing earnings faster than 15% annually sustain that growth for five years only 25% of the time. Companies growing faster than 25% annually sustain it even less often.
When you buy a high-growth company at a high multiple, you’re implicitly betting the company will sustain exceptional growth. The base rate says most won’t. Your specific analysis may give you reasons for confidence, but those reasons need to overcome a 75% historical failure rate for the growth assumption embedded in the price.
Acquisition Synergies
How often do acquisitions create the value management promises? Studies consistently show that 60-80% of acquisitions fail to deliver projected synergies. The acquiring company’s stock typically underperforms following acquisition announcements, especially for large deals.
When management announces an acquisition and explains the strategic rationale—the cost synergies, the revenue opportunities, the competitive positioning—the inside view sounds compelling. But the outside view says most acquisitions destroy value for the acquirer’s shareholders. Your base rate expectation should be skepticism, adjusted upward only if specific factors suggest this deal is exceptional.
Analyst Forecasts
How accurate are professional earnings forecasts? Not very. Research shows analyst estimates are systematically optimistic and systematically miss the degree of deviation from expectations. More importantly, the dispersion of actual results far exceeds the dispersion of forecasts—analysts cluster around similar estimates while actual outcomes vary widely.
When your investment thesis depends on a company hitting analyst expectations, you’re building on a foundation of historically unreliable predictions. The base rate for earnings surprises (positive or negative) is high enough that “in line with expectations” should be viewed as one outcome among many, not the default assumption.
Reference Class Forecasting
Knowing base rates exist is useful. Systematically incorporating them into decisions requires a structured approach: reference class forecasting.
The method has three steps:
Step 1: Identify the reference class. What category does this investment belong to? Be specific but not artificially narrow. A biotech company awaiting FDA approval belongs to the reference class of “biotech companies awaiting FDA approval,” not “pharmaceutical companies” (too broad) or “biotech companies founded in Boston by MIT PhDs awaiting approval for rare disease treatments” (too narrow—you won’t have enough cases for a meaningful base rate).
Step 2: Research the base rate. What percentage of investments in this reference class have succeeded by whatever definition matters? Academic research, industry data, and historical analysis provide the outside view. This step requires work that most investors skip.
Step 3: Adjust based on specific factors. Now, and only now, consider what makes this particular case better or worse than average. The inside view information doesn’t disappear—it adjusts your estimate away from the base rate. But it adjusts rather than replaces.
The critical insight: start from the base rate, then move. Don’t start from your inside-view estimate and check it against the base rate. The anchoring effect matters—wherever you start biases your final estimate. Starting from the base rate ensures statistical reality anchors your judgment.
An Example in Practice
Consider evaluating a retail company implementing a digital transformation strategy. Management is optimistic, the plan seems sensible, early results are promising.
Step 1: Reference class. “Traditional retailers attempting digital transformation.” Not all retailers (too broad), not this specific company’s situation (too narrow).
Step 2: Base rate. Research suggests roughly 25-30% of traditional retailers successfully execute digital transformation measured by sustained revenue growth and margin improvement. Most fail to achieve the stated goals.
Step 3: Adjust. What makes this case better or worse than average? Stronger balance sheet than peers: adjust upward. Management team with no digital experience: adjust downward. Early results genuinely better than typical transformation: adjust upward. Highly competitive market with well-funded digital natives: adjust downward.
Your final estimate should be somewhere in the range of 20-40%, anchored on the base rate and adjusted by specific factors. If your inside-view analysis suggested 70% probability of success before considering base rates, you’ve identified a large gap that should trigger serious reconsideration.
The Superforecasters
Philip Tetlock’s research, published in Superforecasting, provides the best evidence for base rate thinking’s power.
Tetlock ran forecasting tournaments where thousands of participants predicted geopolitical and economic events. Most forecasters performed barely better than chance. But a small group—the “superforecasters”—consistently outperformed, even beating professional intelligence analysts with access to classified information.
What distinguished superforecasters? Several habits, but base rate thinking was central. When asked to predict an event, superforecasters started with the outside view. “How often has this type of event occurred historically?” preceded “What are the specific factors in this case?”
The superforecasters also updated their base rate estimates continuously as new information arrived. They treated forecasting as a process of adjustment from a statistical starting point, not as a one-time judgment based on case-specific analysis.
Importantly, superforecasters maintained calibration—when they said something had a 30% probability, it happened roughly 30% of the time. Most forecasters are overconfident: their 80% predictions happen perhaps 60% of the time. This miscalibration stems from overweighting the inside view and underweighting base rates.
Lessons for Investors
The superforecaster research suggests specific practices for investors:
Keep prediction records. You can’t improve calibration without feedback. Track your probability estimates and outcomes. Over time, analyze whether your 60% predictions happen 60% of the time.
Seek the base rate before analyzing the case. Make it a procedural requirement. Before building your model or writing your memo, document the base rate for this category of investment. What does the outside view say?
Express estimates as probabilities, not certainties. “This will work” is less useful than “This has a 65% chance of working.” Probabilistic thinking forces engagement with base rates and alternative outcomes.
Update continuously. Your base rate estimate is a starting point, not a fixed conclusion. As new information arrives, adjust. But always maintain connection to the statistical foundation.
Embrace uncertainty. The best forecasters know they don’t know. Overconfidence correlates with poor performance. Humility about your ability to predict—which base rate awareness reinforces—correlates with better actual predictions.
Why Base Rates Require Margin of Safety
Base rate thinking connects directly to margin of safety: the discount to intrinsic value that protects against being wrong.
If you know the base rate for an investment category is 30% success, you’re acknowledging a 70% probability of being wrong even with excellent analysis. That 70% demands protection. A price that offers no margin for error is inappropriate for an investment where error is the most likely outcome.
The math becomes concrete. If 70% of turnarounds fail, and failed turnarounds lose 50% of investment value on average, you need the successful outcomes to gain far more than 50% just to break even in expectation. The margin of safety must be large enough that your expected value is positive despite the high base rate of failure.
Base rates also tell you how much analysis matters. In categories where base rates are extreme—say, 5% success for pre-revenue biotech—even exceptional analysis barely moves the needle. Your job isn’t to find the winners; your job is to pay a price that compensates for the overwhelming odds of failure. No amount of inside-view analysis turns a 5% base rate into a 50% probability for a specific investment.
In categories where base rates are closer to 50-50, inside-view analysis matters more. Your specific insights can meaningfully shift probability in either direction. Here, the margin of safety can be smaller because your analysis provides real information.
The principle: base rates determine how much margin you need and how much your analysis matters.
Base Rates and Circle of Competence
Your circle of competence should include knowing the base rates in your areas of focus.
If you invest in banks, you should know: What percentage of banks with this capital ratio faced regulatory action? What percentage of loan categories experience elevated losses in recessions? What’s the historical range of net interest margins in different rate environments?
If you invest in technology, you should know: What percentage of SaaS companies growing at this rate sustain it? What percentage of market leaders maintain leadership for ten years? What’s the historical success rate of platform transitions?
If you invest in retailers, you should know: What percentage of store expansion programs achieve target returns? What’s the base rate of successful international expansion? How often do discount formats take share from traditional competitors?
This knowledge isn’t trivia—it’s essential context for evaluating any specific opportunity in your circle. You can’t know whether an individual case is above or below average if you don’t know the average.
Building this base rate knowledge requires deliberate study: academic research, industry data, historical analysis of outcomes. It’s the unglamorous work that separates sophisticated investors from those who merely traffic in stories.
Common Mistakes
Base rate thinking is simple in principle but difficult in practice. Several common errors undermine its power.
Choosing Narrow Reference Classes
When the broad reference class has an unfavorable base rate, there’s temptation to define a narrower class where base rates seem better. “Sure, most turnarounds fail, but this is a turnaround with a new CEO from a successful company, which succeeds 60% of the time.”
Sometimes narrow reference classes are appropriate—if you genuinely have enough cases to establish a reliable base rate. But often, the narrow class is a way of ignoring inconvenient statistics. If your narrow reference class has fewer than 30-50 cases, the base rate estimate is unreliable. You’re probably better served by the broader class with more data.
Adjusting Too Far from Base Rates
The inside view pulls powerfully. After extensive analysis, investors often adjust so far from base rates that the base rate becomes irrelevant.
“The base rate for IPOs is underperformance, but this particular IPO has exceptional management, huge market opportunity, and proprietary technology—I estimate 80% probability of outperformance.”
Maybe. But how many IPO investors thought the same about their IPOs? Almost all of them. And most of those IPOs underperformed anyway. The inside view felt compelling in every case. Adjusting to 80% probability means believing your analysis is dramatically better than the analysis of investors in the 65% of IPOs that underperformed.
A useful discipline: limit your adjustment from base rates to 20-30 percentage points except in extraordinary circumstances. If the base rate is 30%, your estimate should probably fall between 10% and 60% unless you have truly unusual information.
Ignoring Base Rates for “Unique” Situations
Every investment feels unique. That’s the nature of narrative thinking—the details are always specific, never generic. But this feeling of uniqueness is itself a bias. Most “unique” situations fit into reference classes with well-established base rates.
The question isn’t whether this situation has unique elements—it always does. The question is whether those unique elements are likely to dramatically shift outcomes away from historical frequencies. Usually, they aren’t. The factors that drive outcomes are more similar across cases than our inside-view analysis suggests.
Confusing Base Rates with Predictions
A base rate is not a prediction. Saying “30% of turnarounds succeed” doesn’t mean this turnaround has a 30% probability of success. It means 30% is your starting point before considering case-specific information.
The base rate is a prior probability in Bayesian terms. Your analysis provides evidence that updates that prior. The posterior probability—your actual estimate—incorporates both. Skipping the prior and going straight to an inside-view estimate produces systematically overconfident forecasts.
The Practice
Incorporating base rate thinking into your investment process requires deliberate habits.
Exercise 1: The Base Rate First Procedure
For your next investment analysis, write down the base rate before doing any other work. What is the reference class? What is the historical success rate? Only after documenting the outside view should you build your model, evaluate management, or analyze competitive dynamics.
Notice how this changes your analysis. With the base rate established, your inside-view work becomes calibration: How does this compare to the average case? What factors suggest above or below base rate probability?
Exercise 2: Building Your Base Rate Library
For your circle of competence, research and document the key base rates. What are the historical frequencies that matter for evaluating investments in your area?
Create a reference document you can consult when evaluating new opportunities. Include sources so you can update as new research becomes available. This library becomes a competitive advantage—most investors don’t have it.
Exercise 3: Tracking Calibration
For the next year, express all investment theses as probability estimates. “60% probability this stock outperforms over three years.” “75% probability the acquisition closes.” “40% probability earnings beat estimates.”
At year end, analyze calibration. Did your 60% predictions happen 60% of the time? Most investors discover significant overconfidence. This feedback is invaluable—it reveals how much you’re overweighting the inside view.
Exercise 4: The Pre-Mortem Base Rate Check
Before finalizing any investment, conduct a base rate pre-mortem. Imagine you’re looking back from three years in the future and the investment failed. What’s the base rate probability that would happen? Write down the most likely reasons for failure drawn from historical patterns in the reference class.
This exercise forces engagement with the outside view at the crucial moment of decision. It’s easy to ignore base rates during the excitement of discovery; it’s harder when you’ve explicitly written down the historical failure modes.
Exercise 5: Reference Class Discussions
When discussing investments with others, start by agreeing on the reference class and base rate. “Before we debate the merits of this particular turnaround, let’s agree: what’s the base rate for turnarounds in this industry? What does that suggest about how much evidence we need before being confident?”
This frames discussion around calibration rather than narrative persuasion. It’s harder to argue your way to overconfidence when the base rate anchors the conversation.
Base Rates and Other Frameworks
Base rate thinking reinforces and is reinforced by the other essential frameworks.
Margin of Safety Calibration
Margin of safety answers “how much protection do I need against being wrong?” Base rates answer “how often am I likely to be wrong?” Together they calibrate position sizing and discount requirements.
A 70% base rate of failure demands larger margin. A 50% base rate demands less. Your discount to intrinsic value should be proportional to the probability of being wrong—and base rates provide the best estimate of that probability.
Circle of Competence Depth
Inside your circle of competence, you should know base rates cold. Knowing base rates is part of what defines competence. If you invest in banks but don’t know the historical frequency of dividend cuts during recessions, you’re not as competent as you think.
Conversely, lacking base rate knowledge is a signal you’re operating outside your circle. The inside view feels compelling even when you don’t know the statistical foundations. If you can’t cite relevant base rates, you can’t properly calibrate your confidence.
Resulting Prevention
Resulting—judging decisions by outcomes—is amplified by base rate neglect. When you don’t know the base rate, every outcome feels meaningful. Success suggests good analysis; failure suggests bad analysis.
With base rates, outcomes calibrate differently. If the base rate was 30% and your investment failed, that’s the most likely outcome—no evidence of analytical failure. If the base rate was 30% and your investment succeeded, that’s against the odds—maybe you were genuinely right, or maybe you were lucky. Base rates contextualize outcomes and prevent false learning.
Asymmetric Returns Structuring
Asymmetric returns thinking asks: What positions have limited downside and large upside? Base rates inform both sides of that calculation.
If the base rate of success is 20%, you need the successful outcome to pay 5x or more just to break even in expectation. If the base rate is 50%, the payoff requirements are less extreme. Base rates determine how asymmetric a position needs to be to compensate for its probability of failure.
The Deep Insight
Base rate thinking is fundamentally about epistemic humility—acknowledging that your individual analysis operates within statistical regularities you don’t get to escape.
The inside view flatters our competence. It suggests our careful analysis provides meaningful edge, that we can see what others miss, that our specific situation differs from the boring averages. The outside view is humbling. It suggests most of our confident predictions will be wrong, that our situations are less unique than they feel, that the crowd’s outcomes predict ours.
This humility is painful but profitable. The investors who consistently outperform are not those with the most compelling inside-view narratives. They’re those who calibrate confidence to evidence, who start with statistical reality before indulging analytical creativity, who know the odds they’re fighting.
Kahneman summarized it perfectly: “People who have information about an individual case rarely feel the need to know the statistics of the class to which the case belongs.”
That rare need—cultivated into habit, enforced by procedure—separates calibrated investors from the confident and wrong.
Base rates provide the statistical foundation for every investment decision. They tell you how often investments like yours succeed, calibrating confidence before analysis begins. For understanding how base rates determine required margin, see margin of safety. For building the competence to know base rates in your areas, explore circle of competence. For using base rates to structure favorable risk-reward, see asymmetric returns. And for preventing outcomes from corrupting your learning, revisit resulting.