Examples
Examples
Try the starting example
- State probabilities adding to 1
- 0.4,0.6
- Actions: name, payoff in each state
- A,100,0 B,50,50
- Best expected payoff without perfect information
- 50
- Expected payoff with perfect information
- 70
- Value of perfect information
- 20
- Best action(s) without information
- B
- Expected payoff by action
- A: 40; B: 50
Try another scenario
- State probabilities adding to 1
- 0.8,0.2
- Actions: name, payoff in each state
- A,100,0 B,50,50
- Best expected payoff without perfect information
- 80
- Expected payoff with perfect information
- 90
- Value of perfect information
- 10
- Best action(s) without information
- A
- Expected payoff by action
- A: 80; B: 50
Calculation method
Calculation method
Best expected payoff without perfect information
Maximum across each action’s probability-weighted payoff.
Expected payoff with perfect information
For each state choose its best payoff, then probability-weight those state maxima.
Value of perfect information
With-information value − without-information value; a small negative roundoff is clamped to zero.
Best action(s) without information
List exact best-expected-value ties.
Expected payoff by action
Sum each action payoff×state probability.
Symbols and units
- a
- State probabilities adding to 1
- b
- Actions: name, payoff in each state
Use the units shown. A list keeps the order in which you enter its values. The symbol ^ means “raised to the power of”.
Assumptions & limits
Assumptions & limits
- An upper bound on the value of knowing the state perfectly under your supplied payoffs and probabilities. It is not the value of an imperfect forecast or a recommendation to buy information.
Are these inputs private?
These calculations run in your browser. A shared link contains the values you choose to share. Only questions submitted to the separate AI finder are sent to OpenAI.
Further reading
MIT: value of perfect information ↗
Background reading; this does not imply endorsement or independent review of SumFizz.