Interactive explainer

Interactive Bayesian update

How does new evidence change an estimated success rate? Choose a prior, add successes and failures, and compare the posterior, 95% interval and probability of exceeding your threshold. Free, with no registration; your inputs stay in your browser.

Interactive

Live Bayesian update: prior, likelihood and posterior

This explainer requires JavaScript. Its calculation: a prior Beta⁡(α,β)\Beta(\alpha, \beta) and kk successes in nn trials yield posterior Beta⁡(α+k,β+n−k)\Beta(\alpha + k, \beta + n - k).

For a conversion or quality rate, first define what counts as success and which observations are comparable. This calculator shows uncertainty about one rate; use the Bayesian A/B test calculator to compare two variants.

How the explainer works

The unknown quantity is a success rate θ\theta, such as a page’s conversion rate or a method’s hit rate. The dashed curve is the prior: your assumption before data is available. Every observation is a success or a failure. The orange curve shows the data’s likelihood, normalised to area 1 for easier comparison. The filled curve is the posterior: the product of the two, normalised again.

Because a beta prior is conjugate to the binomial likelihood, the update has an exact expression:

Beta⁡(α,β)  →  k successes in n trials    Beta⁡(α+k,  β+n−k)\Beta(\alpha, \beta) \;\xrightarrow{\;k \text{ successes in } n \text{ trials}\;}\; \Beta(\alpha + k,\; \beta + n - k)

The shaded band is the central 95% credible interval; the metrics use the exact beta distribution, not an approximation.

Three things to try

  1. Little data, strong prior: choose “Informed” and add five successes. The posterior barely moves: prior knowledge weighs as much as 40 observations.
  2. Lots of data: simulate 50 observations several times. Whatever the starting prior, the posterior concentrates around the true θ\theta.
  3. Threshold: move the threshold and watch P(θ>threshold)P(\theta > \text{threshold}) change. This is exactly the quantity needed for many decisions.

From NyxAI

smartlytics is the learning hub of NyxAI GmbH.

Explore the methods we work with, explained openly, with sources and calculations you can reproduce.

Decision Core

NyxAI is developing a Decision Core for Decision Intelligence that combines optimisation, simulation, statistics and reinforcement learning: compare options, make uncertainty visible and learn from outcomes, with people in control.

Use case: explain reporting variances with AI and evidence

DeepSquare

Explainable chess AI with human playing styles: NyxAI’s testing ground for understandable decisions under time pressure.