Selection error means a successful available action was missed. Candidate failure means the tested action set contains no success for that state and task. The remedies are different.
The mechanism behind the question
In E7-D, 90.7% of its recorded computation failures were attributed to neither of the two candidates succeeding. That percentage is among failures, not among all queries.
Make an observation
Enable the no-success scenario. Change the state weights and goal; inspect the best 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
No reweighting rescues a table whose entries are all failures. A new action, a different deadline or a changed task would require a new test.
Check your understanding
Follow this idea into the research
Textbook revision 1 · Evidence snapshot: 05 Nov 2026 · Changes & corrections ↗