Survivorship Bias Detector

Overview

This pillar identifies survivorship bias, a common logical error where data focuses only on successful outcomes. It provides a crucial reality check, preventing users from making overly optimistic predictions based on incomplete evidence.

What It Does

The Survivorship Bias Detector scans the underlying data of a market to see if it systematically excludes failures. It looks for datasets that only track winners, like top-performing stocks or successful startups, while ignoring the vast majority that failed. The pillar then estimates the potential impact of these missing data points to provide a more realistic baseline probability.

Why It Matters

Most analysis is skewed towards success stories, creating an inflated sense of potential and leading to poor predictions. This pillar corrects for that over-optimism by accounting for the silent evidence of failure. It helps you avoid common traps and make more robust decisions.

How It Works

First, the pillar analyzes the data selection criteria for a given market, looking for keywords like 'top-performing' or 'active'. Second, it cross-references this with broader historical records to identify entities that have failed or dropped out. Finally, it quantifies the bias by estimating the historical failure rate and adjusting the perceived probability of success.

Methodology

The analysis uses cohort tracking to compare an initial population size to the final 'survivor' group. It backtests against historical datasets with known high attrition rates, such as startup funding rounds or clinical trials. A statistical model estimates a 'silent failure rate' by comparing the market's dataset to more comprehensive, unfiltered sources.

Edge & Advantage

This provides a powerful contrarian edge by revealing when market consensus is built on an incomplete picture. It allows you to spot overpriced assets and bet against popular but statistically flawed narratives.

Key Indicators

• Dataset Filtering Logic

  high

  Detects if the data source explicitly filters for success, such as 'top funds' or 'active companies'.

• Attrition Rate Estimation

  high

  The estimated percentage of entities that dropped out of the dataset over its timeframe.

• Silent Evidence Score

  medium

  A composite score indicating the likely impact of missing failure data on market probabilities.

Data Sources

• Crunchbase / PitchBook

  Provides data on both funded and failed startups, useful for checking bias in tech markets.

• SEC Filings (EDGAR)

  Can reveal delisted companies or failed funds, contrasting with popular 'top performer' lists.

• ClinicalTrials.gov

  A registry of clinical studies, including those with negative or terminated results which often go unpublished.

Example Questions This Pillar Answers

• → Will this new tech startup reach a $1 billion valuation?
• → Will the S&P 500 index provide a positive return next year?
• → Will a movie from this acclaimed director be a box office hit?

Tags