How should team news change my view? — Statometrics Academy calculator

Try it · Bayes updater

Chance he's out (prior)25%

Says out when he IS out90%

Says out when he's fit20%

Out of 1,000 games he’s out in 250. The source says ‘out’ 375 times. It’s right in 225 of them: 60.0%.

Prior

25%

Likelihood ratio

4.50

Posterior

60.0%

Old odds 0.33 : 1 × LR 4.50 = new odds 1.50 : 1

Second, independent source

What it does to the price

Team wins WITH him50%

Team wins WITHOUT him44%

Win chance before

48.5%

Fair price before

2.062

Win chance after

46.4%

Fair price after

2.155

The news should move you from 25% to 60.0% that he’s out. The team’s fair price goes from 2.06 to 2.16 (it drifts).

Push “says out when he’s fit” down and watch the posterior climb: false alarms matter more than hits.

Filled dots are the times the source says 'out'. Orange dots are games where he really is out; blue, games where he's fit. Count the filled ones and ask how many are orange: that's the posterior.