The deal was signed on June 17 and dead within twenty days. Two things killed it: Israel's war in Lebanon, which no deal could stop, and Iran's Revolutionary Guards, who never meant to keep it. What followed is not chaos. It is the state the war keeps returning to - a low, steady burn that none of the powerful players actually wants to end. The hardest part to accept is about the United States: it is fighting this war, and it cannot end it.
- June 17: the US and Iran sign the deal, a 14-point memorandum of understanding (P085 confirmed).
- June 19: the first working meeting in Switzerland is cancelled. Iran refuses to attend while Israel is still bombing Lebanon (47+ killed June 18). US Vice President Vance drops out.
- June 23: talks in Qatar stall over three things - letting UN nuclear inspectors back in (Iran denies it ever agreed), $12 billion in frozen Iranian funds (the US wants the money limited to buying food), and Iran's stockpile of near-weapons-grade uranium.
- Late June: the Strait of Hormuz reopens faster than expected. So much oil flows that the price falls to about $70 a barrel.
- July 4-9: Iran finally buries Ali Khamenei, the supreme leader killed on the war's first day in February. Millions mourn. His successor, Mojtaba, is still not seen in public.
- July 6-7: Iran's Revolutionary Guards attack three oil tankers in the Strait of Hormuz.
- July 7: Trump cancels Iran's permission to sell oil and declares the deal "over." Oil jumps about 6% to $76.
- July 7-9: the US strikes 80+ targets in Iran (air defenses, radar, missile sites, 60+ Guard patrol boats). Iran fires back at 85 US targets in Bahrain and Kuwait, most shot down. One Guardsman is killed. No troops move in on the ground.
20 days
from signed deal (June 17) to Trump calling it 'over' (July 7)
80 / 85
US strikes on Iran vs Iranian strikes on US bases in the July clash - a matched exchange
~$76
the oil price once the fighting resumed - a contained clash, not a full war
0
of the powerful players actually want peace
Findings
Why the deal died
The deal was signed on June 17 and dead within twenty days. Both reasons were built in from the start. Iran's Revolutionary Guards attacked oil tankers in the Strait of Hormuz, which gave Trump his reason to walk away. Israel kept bombing Lebanon, which kept Iran from ever showing up to the talks in Switzerland. Neither the Guards nor Israel had signed the deal, so neither was bound by it. The two forces most able to wreck it were the two the deal could not touch.
Three powers all want the war frozen
The war stays stuck because three of the strongest players each want it stuck, for their own reasons. China profits: cheap Iranian oil keeps flowing, and the US stays tied down far from Asia. The Gulf states prefer a frozen war to any ending - an Iranian victory next door would be dangerous, and a US-led victory that plants an American client on their border would be too - so they quietly refuse to host the strikes that would force a finish. And Israel is winning slowly: the steady burn grinds Iran down while Israel keeps every option open, so it has no reason to stop. Three different motives, one shared result - nobody strong enough to end the war actually wants to. The full argument is in
The war's equilibrium is not a path to peace.
America is too weak to end it
The strangest part is that the superpower fighting the war cannot end it. It cannot invade: the ground force needed to seize Iran's nuclear site is not there. Its navy can patrol the Strait of Hormuz but cannot close it. The Gulf states will not hand over the airbases a knockout blow would need. And the one tool that could end the war - a signed deal - is the tool Israel and the Guards keep breaking. Even if Trump wants out, wanting was never the problem. Power is. Washington is a player in this stalemate, not its master.
Scored
Only one prediction resolved this session, and it landed on the least important thing forecast. P085 (the deal would be signed by July 5) came true - it was signed June 17 - but the signing was never what mattered. The deal died over the two things its follow-on predictions track: whether anyone in Iran lets nuclear inspectors in (P087), and whether Israel pulls back from Lebanon (P088). Both are now live, and both look likely to confirm. The new predictions move away from guessing what leaders will decide, which is hard to read, and toward what the fighting on the ground can and cannot do (P090-092), which is easier to call.
New Predictions
Standalone Predictions
Endgame Scenarios
Down from 12%. The MoU signed June 17 and died within 20 days; the clean victory-withdrawal frame died with it. A managed US exit is still reachable, but only through a re-signed deal Israel and the IRGC would break again. The kinetic route stays closed - ground assault infeasible, Gulf basing denied, US confined to standoff strikes.
B. Negotiated ceasefire
65%
Up from 62%, but the internal split inverts. B1 (genuine settlement) collapses from 0.30 to ~0.07 - the June deal was the live test of a genuine settlement and it failed at R012's two named preconditions (IRGC non-enforcement, Israeli sabotage). B2 (frozen-conflict extension) rises from 0.32 to ~0.58 and is no longer an 'extension' but the destination: the low-simmer equilibrium R017 (not a path to peace) names, over-determined by China, the Gulf, and Israel each preferring the freeze independently. The July 7-9 US-IRGC exchange is B2's kinetic face, not a new war.
C. Grinding attrition / ground ops
3%
Down from 4%. Near-dead. The ground assault that would take Isfahan is beyond the deployed force, and Gulf basing denial stands. The US response to the July tanker attacks was standoff-only, confirming the constraint.
Regime collapse. Down 1 point. Khamenei's long-delayed funeral (buried July 9) drew millions; the succession to Mojtaba, though he remains publicly absent, proceeded without visible fracture. R005 holds - the IRGC is the regime's coercive substrate and the war strengthens its case.
Wider regional war. Flat. The July 7-9 base-on-base exchange (US 80+ targets; IRGC 85 targets in Bahrain and Kuwait, mostly intercepted, one death) tested the rung again and was absorbed. No driver loaded onto wider war specifically; the Israel drag-in risk is real but unrealized. Held per scenario-weight discipline.
H. Ceasefire collapse / Phase 2
9%
Up 1 point. Trump declared the MoU 'over' July 7 - a harder break than prior TACO folds - and strikes resumed. But the exchange is contained (symmetric, oil ~$76, no ground), which reads as B2's kinetic face rather than Phase 2. H now carries the fat tail: the thermostat breaking upward if a mass-casualty event - a burned tanker crew, an un-intercepted base strike killing Americans - forces a government off the two-track fiction.
The war found its equilibrium. The June deal (signed June 17) ran R012's (The No-Lose Architecture) named falsification test and failed at the two load-bearing preconditions: the IRGC did not enforce it (the July 6-7 Hormuz tanker attacks) and Israel sabotaged it (June 18 Lebanon strikes killed 47+ and collapsed the June 19 Switzerland round). What the deal died into is Branch Four of that architecture - the freeze that weakens Iran while preserving options - now running in the open. The reweight records the structural finding: B1 (genuine settlement) collapses because it was tested and failed; B2 stops being a 'frozen-conflict extension' and becomes the destination. R017 (not a path to peace) names the equilibrium as over-determined - China (discounted crude, US pinned to the Gulf), the Gulf (revenue with basing-denial control), and Israel (the simmer is already a win, a high boil even better) each prefer the freeze independently - and names the apex: the United States is too weak to end the war it is fighting, even if Trump wants out. Ground assault impossible, naval campaign only manages the strait, Gulf forbids the basing, and the one instrument that ends the war (a deal) is the one Israel and the IRGC are built to break. Framework confidence holds at 0.55.
Board State: the obvious next move
For the US, the simplest move is the one it is already making: answer the tanker attacks with strikes from ships and aircraft (no Gulf airbases needed, so the Gulf's refusal does not stop it), wear down Iran's boats and missiles, and leave the door open for Iran to return to the deal. Punishment, not invasion.
For Iran's Guards: keep the threat to Gulf shipping alive with occasional attacks, fire back when hit (mostly for show, since the missiles get shot down), but avoid hitting Gulf states' oil or cities, which would turn the Gulf against Iran.
For Israel: keep bombing Lebanon at its own pace, and use the deal's collapse to resume the strikes on Iran that the ceasefire had held back - limited only by how much its own army can sustain.
For the Gulf states: no change. Deny the US its airbases, press both sides to keep it small, and keep the war frozen, because either way it could end is worse for them.
What to Watch
- Does the fighting stay small (P090), or does a disaster - a tanker crew killed, a US base hit - drag a government into a bigger war?
- Do tankers keep moving through Hormuz (P091), or does one large incident scare shipping off even without a full blockade?
- Oil: does the price stay under $95 (P092), or does a real blockade or a strike on Gulf oil send it higher?
- Do talks restart? Trump left himself a way back in - he said negotiations "might" continue and gave the oil sanctions a ten-day wind-down instead of cutting them off at once.
- Does Israel start hitting Iran directly again now that the ceasefire is "over"?
- Mojtaba: Iran's new leader still has not appeared in public. His first appearance would change who can actually deliver a deal.