Counterfactual Control Test¶
Reasoning test — instantiates Outcome Responsibility Attribution Calibration
Asks whether an agent actually had a feasible, foreseeable alternative that would likely have changed the outcome — the line between real control and mere presence at a result driven by luck.
Counterfactual Control Test probes one agent's responsibility with a counterfactual: given what they knew and could foresee at the time, did they have a feasible alternative that would probably have changed the outcome? Its defining move is to isolate control from luck. Being present at a bad outcome — or even causally upstream of it — is not the same as having been able to prevent it. The test separates the agent who could have acted differently, and so bears responsibility, from the one carried along by circumstances they could not have altered. Crucially, it fixes the agent's knowledge as it stood then, not with the clarity of hindsight.
Example¶
A fund drops 30% in a quarter and the reflex is to fire the manager. The Counterfactual Control Test asks: among the decisions actually open to them, on the information they had, was there a feasible alternative that would likely have avoided the loss? If the fall was a broad market crash that took down every comparable fund, no feasible, foreseeable alternative existed — the loss was largely luck, and blaming the manager is outcome bias. If instead they had concentrated 60% of the fund in a single name against their own mandate, a feasible alternative — staying within mandate — plainly existed and was foreseeably safer, so control was real. Same 30% loss; opposite attributions, decided by control rather than by the size of the number.
How it works¶
- Fix the decision point and the information available then — not what is known now.
- Construct the alternatives actually open to the agent at that point.
- Ask whether any of them would likely (not merely possibly) have changed the outcome.
- Separate "no feasible alternative" (luck, no control) from "feasible alternative not taken" (control, and responsibility).
Tuning parameters¶
- Feasibility bar — how realistic an alternative must be to count; set generously and you conjure options no real person would have seen (hindsight); set strictly and everyone escapes.
- Foreseeability window — what the agent should have anticipated given their role; widening it raises the standard of care.
- Likelihood threshold — how probable the better outcome must be to establish control — "would have" averted versus merely "might have."
- Role calibration — the same alternative is feasible for an expert and not for a novice; the bar tracks the competence the role demanded.
When it helps, and when it misleads¶
Its strength is that it is the sharpest instrument against outcome bias and moral luck[n1] — the habit of grading agents by results they did not control — and it cuts both ways, against over-blaming the unlucky and over-crediting the lucky. Its failure mode is that counterfactuals are speculative: the "feasible alternative" is imagined, and imagination bends easily under hindsight to make any outcome look avoidable ("they should have seen it coming"). The classic misuse is constructing, after the fact, an alternative that is obvious only because we now know what happened. The discipline is to build the alternative from what was knowable at the decision point, test it against a competent peer's likely foresight rather than an omniscient one, and treat a strained counterfactual as weak evidence of control.
How it implements the components¶
counterfactual_avoidability_test— the core move: whether a feasible alternative would likely have averted the outcome.agency_control_boundary— the test's product is exactly this line, between what was and was not within the agent's control.
It weighs one agent's control, not the whole picture: it does not lay out the causal chain it works within (outcome_event_record, causal_contribution_map — Causal Contribution Timeline), record what the agent knew or fix the role's standard (knowledge_foreseeability_record, role_duty_standard — Omission–Commission Parity Check and Role–Duty Mapping), or combine agents into shares (attribution_weighting_rule — Blame–Credit Apportionment Table). The same avoidability test is applied specifically to failures-to-act by Omission–Commission Parity Check.
Related¶
- Instantiates: Outcome Responsibility Attribution Calibration — it is the per-agent control judgment the appraisal turns on.
- Consumes: Causal Contribution Timeline (to locate the agent's decision point in the chain) and Role–Duty Mapping (the foreseeability standard for the role).
- Sibling mechanisms: Omission–Commission Parity Check · Causal Contribution Timeline · Blame–Credit Apportionment Table · Just Culture Review · Responsibility Attribution Matrix · Role–Duty Mapping · Attribution Uncertainty Label · Scapegoat Screening Review · Responsibility Diffusion Check · Outcome Responsibility Review Panel · Credit Contribution Register
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Counterfactual Control Test operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it asks whether an agent actually had a feasible, foreseeable alternative that would likely have changed the outcome — the line between real control and mere presence at a result driven by luck.
Independent corroboration: The frozen evidence defines Counterfactual Control Test as 'Asks whether an agent actually had a feasible, foreseeable alternative that would likely have changed the outcome — the line between real control and mere presence at a result driven by luck', so its operative form is Assessment, Review & Assurance.
Nearest alternative: Analysis, Modeling & Optimization — It evaluates a particular agent's feasible alternatives to produce a control finding, using counterfactual analysis as evidence.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Philosophy
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Moral philosophy established responsibility tests that separate controllable, foreseeable alternatives from outcomes fixed by luck.
Related originating lineages:
- Law & Governance — Causation and culpability doctrine operationalizes feasible alternatives, foreseeability, and but-for influence.
- Psychology — Hindsight and outcome-bias research explains why perceived control is inflated after results are known.
Review resolution: The control principle, alternative possibilities, and moral luck are philosophical responsibility concepts. Law turns them into institutional tests and psychology guards against retrospective distortion; the feasible-and-foreseeable checklist is a practical synthesis.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- Stanford Encyclopedia of Philosophy: Moral Responsibility and Alternative Possibilities
- Stanford Encyclopedia of Philosophy: Moral Luck
- Cornell Legal Information Institute: But-for test
Notes¶
[n1] Moral luck — the problem, developed by Bernard Williams and Thomas Nagel, that we praise and blame agents for outcomes partly fixed by factors outside their control. The control test is the operational response: hold the agent to what they could feasibly and foreseeably have done, not to how the dice fell. ↩