Superforecasting Principles

Superforecasting is a set of learnable habits for making more accurate predictions. Research shows the most crucial trait isn't deep expertise but active open-mindedness — the willingness to constantly seek out new evidence and update your beliefs. It replaces gut feelings with a structured, evidence-based process, improving decisions in areas from business strategy to personal finance. For example, a project manager, instead of just declaring a launch will succeed, breaks the problem down, seeks out disconfirming data from sales, and revises their forecast from 90% to 60% confidence based on new market research.

Ask this: "What new information would cause me to change my mind about this prediction?"

Superforecasting is not a magical ability but a disciplined, systematic process for thinking about the future. It emphasizes expressing forecasts in precise probabilities (e.g., "a 70% chance") rather than vague words like "probably." This forces clarity and makes it easier to track accuracy over time.

Several core principles emerged from studies of elite forecasters:

• Find the "Outside View": Before focusing on the specifics of a situation, start with the base rate. To predict if a startup will succeed, don't start with their brilliant pitch deck; start with the fact that the majority of startups fail within their first few years.

• Update Frequently: Treat your beliefs as temporary hypotheses. Superforecasters make small, frequent adjustments to their predictions as new information arrives, rather than waiting for a dramatic event. They are skilled at separating signal from noise.

• Synthesize Diverse Views: Actively seek out perspectives that challenge your own. An investment analyst applying this principle wouldn't just read their own firm's reports; they would study the arguments of the most respected critics of their position to find flaws in their own reasoning.

This approach is most effective for questions with a specific, measurable outcome and a clear timeframe (e.g., "Will inflation in the UK be below 3% by the end of the year?"). It helps guard against common cognitive biases like overconfidence and confirmation bias, leading to better-calibrated judgments.

The principles of superforecasting were empirically identified by the Good Judgment Project (GJP), a large-scale forecasting tournament sponsored by the U.S. intelligence community. Researchers Philip Tetlock and Barbara Mellers discovered that a small percentage of volunteer forecasters were consistently more accurate than professional intelligence analysts with access to classified data. Their success was attributed not to who they were, but to how they thought—their cognitive style and methods.

• Tetlock & Gardner (2015): Superforecasters are defined by their cognitive style, including active open-mindedness, grit, and a growth mindset, which are more important than domain-specific expertise or raw intelligence. [1]
• Mellers et al. (2014): Gains came from both individual habits and powerful aggregation methods. Light training in probabilistic reasoning improved accuracy, but aggregating forecasts—especially when weighting more historically accurate forecasters—boosted prediction accuracy by over 20% compared to simple averages, demonstrating a synergy between individual skill and collective intelligence. [2]
• Tetlock (2005): In his earlier work, Expert Political Judgment, Tetlock found that the average expert was often no more accurate than chance, and that deep specialists were frequently less accurate than generalists due to overconfidence and ideological bias. [3]

Superforecasting is most effective for predicting events on a timescale from several weeks up to approximately 1-2 years, with clearly defined resolution criteria; performance typically declines as the time horizon grows. It is not designed for predicting long-term, complex systemic changes or unprecedented "black swan" events. The principles also require significant cognitive effort, discipline, and a willingness to engage with ambiguity, which can be difficult to maintain without structured practice and feedback.