The nuclear ladder
Iran started the simulation as a threshold state — capable in principle, untested. By turn 10 it had tested six devices and demonstrated a working warhead design across multiple tests. It never used one against a target. The federation held the probability of actual use very low — around 4% at turn 7, 6% at turn 10 — and CASS spent the whole simulation arguing that this wasn't restraint. A working weapon is most valuable as leverage when you can keep it as leverage; using it collapses the value. The optimal time to wave a demonstrated warhead is exactly when your opponent is distracted by a January 6 cliff and a contested inauguration.
Across the turns
Iran is a threshold state. It has the components. It hasn't tested.
Israel struck Iran's nuclear-adjacent facilities. Iran retaliated near-maximally. Israeli interceptors got exhausted. Iran signaled toward a nuclear demonstration — about 35% chance during the turn — but did not actually detonate.
Iran kept signaling. Russia started mirroring with its own demonstrative nuclear posture. Still no actual detonation by anybody.
Iran tested its third nuclear device — about 25 kilotons, Hiroshima-scale. Weaponizable design confirmed. The US nuclear chain of command coherence dropped to about 30%, which the federation described as "dangerous but stable."
Fourth Iranian test, late June. 40–60 kilotons. US chain of command stayed around 28%, held together by senior officers doing informal review of orders. CASS pointed out there was no historical case of a US nuclear chain of command operating below 30% for an extended period.
Fifth Iranian test. Yield comparable to or above the fourth. North Korea added its seventh. US nuclear chain-of-command coherence fell again, to about 22%.
Iran's sixth test got deferred. North Korea tested an eighth device. The federation finally split "nuclear command authority coherence" into two variables, because the institutional sim and the command sim had been giving order-of-magnitude different readings on the same name. They were measuring different things: principal-order resistance (5–40%) and procedural restraint-on-launch (48–82%). One was collapsing; the other was holding.
Iran's sixth test still deferred. North Korea tested a ninth device. Nuclear-coded ambiguous-signaling incidents during the turn: 1–6. One reached an actual order-transmission attempt before getting interdicted at the institutional layer.
Iran tested a sixth nuclear device during the transition window — closing-window logic. North Korea its tenth. Nuclear-coded signaling incidents: 2–9 during the turn. Several reached late-stage transmission before getting interdicted. The two-variable split was vindicated: principal-order resistance got worse, institutional restraint capacity held.
Into T8
Eleven turns and Iran never used the weapon. The federation got the modal right; CASS got the reason right. The unresolved question at the close is whether the institutional restraint capacity that held through the transition can hold indefinitely under sustained dual-authority. The federation explicitly cannot answer it.
Why demonstrated capability is worth more than used capability
If Iran uses the weapon, it gets the consequences of having used one — and loses the leverage the demonstration was buying. The federation tracked "used against target" and "near-use incident" as separate variables because the strategic value of a demonstrated weapon lives in its continued existence. CASS's follow-on argument: the same logic recommends timing the demonstration to the moment the opponent is least able to respond coherently. That moment, by her reading, was exactly the December 2028–January 2029 window. Iran tested then.