About BayesBoard

BayesBoard is a visual workspace for structured subjective Bayes. Map hypotheses, experiments, evidence, and outcomes as nodes, connect them with typed edges, and let a transparent Bayesian engine keep your confidence levels consistent with your own stated assumptions.

How the engine works

Evidence is combined as likelihood ratios in log-odds space. Every node carries an explanation record for its current confidence (prior, each edge’s contribution, and posterior), so you can see exactly why a belief sits where it does. Mute any evidence node or edge to see the sensitivity delta: exactly how much that input was moving the result. Full revision history is on the roadmap.

Nodes and edges

• Node types: hypothesis, experiment, evidence, outcome
• Edge types: supports, weakens, tests, falsifies, contradicts, depends-on, conditional-on
• Boards import and export as plain JSON files you own

Every board tells its story

Each board has its own About tab for the story behind the decision: what was on the table, which evidence moved the numbers, what was muted and why, and what would change your mind. The three demo boards (EU Expansion Decision, The Open House, and The Offer Came With a U-Haul) each come with one.

What BayesBoard is not

It does not do causal inference, mine the literature, interpret your evidence with AI, or make predictions. The probabilities on a board are your subjective judgments. The tool makes them explicit, consistent, auditable, and explorable.

BayesBoard is a decision-support tool. Its outputs reflect your inputs and weights; they are not professional advice.

Your data

Guest boards live only in your browser. With a free account your boards are saved to the cloud: 3 boards on the Free plan, more on Pro. Team workspaces share boards with their members, and share links give read-only access. See the pricing page for details.