The Alignment Lab

A four-role, live-then-async decision simulation, run once as a group and then audited alone — the semester's central assessment.

Due: 3 May 2027 at 12:00 pm AEST · Weight: 45%

The brief#

Play one of four roles in a decision you do not control the outcome of, then go back alone and work out what actually happened.

The full process — the four roles, the pre-meeting contact window, the live meeting, and the individual audit that follows it — is described in full on The Alignment Lab page, including the fairness and safety rules that apply before anyone is assigned a role. Read that page before this one; this page marks the work, that page runs it.

What you submit#

• Initial judgment (individual, before role cards are distributed): a short, dated statement of what you think should happen, based only on the public materials.
• Strategy memo (individual, before contact opens): my responsibilities, my evidence, what I don’t know, who I plan to contact, and my ethical boundaries — the same five fields as the strategy memo template.
• Negotiation log (individual): a record of every contact you made or received — its purpose, any new information, any commitments, and whether your judgment changed — or a note that you chose not to make contact, and why.
• Decision record (one per live group; produced individually under the async make-up or alternative pathway — see The Alignment Lab page’s pathway breakdown): the outcome, in the format on the Templates page.
• Decision autopsy (individual, after the reveal): 500–700 words reaching one of three conclusions about what actually decided the outcome, versus what the decision record says decided it — a specific distortion the record and log support, no significant distortion evident, or indeterminate on what you have (naming the specific evidence that’s missing, the alternative explanation you could not rule out without it, and what would change your judgment if it turned up). Indeterminate is not a shrug: it is argued to the same standard of evidence as the other two.
• Partial rerun (individual): a short re-run of one specific juncture from the negotiation or meeting, testing whether one named process change would have produced a different outcome.

How it is marked#

Everything above is marked per person, not per group, with one exception: two people in the same room can submit very different strategy memos, logs and audits, and are marked on their own work throughout. The decision record is the one artefact a live group produces together, and is marked once and applied to everyone named as a participant in it. None of the five criteria mark negotiation win/loss, performance skill, how often you spoke, or whether your role’s stated goal was achieved — they mark how you used evidence, adapted to what you learned, and reasoned about the outcome afterwards.

Missing the live meeting, or missing role assignment and the contact window entirely, does not change these five criteria or their weights — it changes which pathway supplies the evidence for them. The Alignment Lab page’s “What you submit” section sets out, criterion by criterion, exactly what the live, async make-up, and alternative / structured-async pathways each submit, and states explicitly why each is held to the same evidentiary standard rather than a lighter one.