When no action covers every candidate state, one possible rule is to maximize expected success. That rule needs both outcome predictions and weights for possible states.
The mechanism behind the question
The teaching score is Σ p(state) × success(action, state). It is a finite calculation. It does not explain where reliable probabilities come from, and it ignores action costs.
Make an observation
Make state 1 much heavier than state 3, then reverse the weights. Observe which preparation has the highest score.
What must you know to choose well?
| Preparation | State 1 | State 2 | State 3 | Expected success |
|---|---|---|---|---|
| A | Success | Success | Fail | 66.7% |
| B | Fail | Success | Success | 66.7% |
| C | Fail | Fail | Fail | 0.0% |
Best expected score: 66.7%. Candidate(s): A, B.
Weights are normalized to sum to one. Scores use this explicit finite table. No observation model, fitted probability, action cost, or E-series experiment is reproduced. A tie is retained, not silently broken.
What to notice
The optimal choice depends on the assumed distribution. If one action costs more time or energy, success alone is not a complete objective.
Check your understanding
Follow this idea into the research
Textbook revision 1 · Evidence snapshot: 05 Nov 2026 · Changes & corrections ↗