The Pentagon already fought this war in 2002. It lost in ten minutes.

A $250 million war game against an unnamed Gulf adversary previewed the mines, swarms and suicide skiffs of 2026. The most dangerous pages of that playbook, and of three centuries of wars against Iran, remain unused.

By the Middle East Desk · June 2026 · 13 min read

A $250 million war game against an unnamed Persian Gulf adversary previewed the mines, swarms and suicide skiffs of 2026 with uncomfortable fidelity. The most dangerous pages of that playbook, and of three centuries of wars against Iran, remain unused.

In the summer of 2002, the United States military staged the largest and most expensive war game in its history. Millennium Challenge 2002 ran from July 24 to August 15, involved roughly 13,000 troops across seventeen simulation sites, cost $250 million, and was set in 2007 against an unnamed Persian Gulf adversary that participants understood to represent a state like Iran or Iraq.[1] The opposing "Red" force was commanded by a retired Marine three-star, Lt. Gen. Paul Van Riper, who had been given a free hand to fight as a regional adversary actually would, and who used that freedom to the full.

When the Blue fleet steamed into the Gulf, Van Riper launched a pre-emptive, massed salvo of cruise missiles timed to overwhelm the Aegis air-defense system, coordinated with swarms of small boats, some packed with explosives on ramming courses, mixed indistinguishably into civilian boat and air traffic. Sixteen warships went to the bottom of the simulated Gulf, among them an aircraft carrier, ten cruisers and five of six amphibious ships, with more than 20,000 notional sailors and Marines lost; some accounts put the figure at nineteen ships.[2][5] "The whole thing was over in five, maybe ten minutes," Van Riper later said.[1]

The events of June 2026, a downed American helicopter over the Strait of Hormuz, retaliatory strikes on Iranian coastal radars, Iranian missiles arcing toward bases in three Arab states, are not improvisations.[25][26] Nearly every move in the current war was rehearsed somewhere, in a Pentagon simulation, a think-tank crisis game, or an actual war fought against Iran or by it. The record of those rehearsals is worth reading closely, both for what the war has already taken from it and for the pages that remain unturned.

What Van Riper actually did

The tactical mechanics of Millennium Challenge repay attention, because they were designed around a specific insight: a technologically inferior adversary survives by refusing to fight on the terms American systems are built to detect.

Blue's first move was a strike on Red's communications infrastructure, after which Van Riper was expected to be deaf and blind. Instead he went silent by choice. "You're going to have to use cellphones and satellite phones now," he recalled telling his staff. "We're going to use motorcycle messengers and make announcements from the mosques." Orders moved by courier. Aircraft launched using light signals, a method out of the Second World War, so that no radio emission betrayed the preparation. American signals intelligence had nothing to intercept because nothing was transmitted.[2]

The naval blow followed the same logic. Rather than meet the fleet at sea, Red waited until it entered the confined waters of the Gulf, where reaction times compress and civilian traffic is dense, then fired everything at once from shore batteries, aircraft and small craft on multiple axes. The salvo saturated the fleet's defenses; the explosive small boats arrived with, and inside, the missile wave.[1]

What happened next made the exercise famous. The Joint Forces Command refloated the sunken fleet and resumed play under new constraints. Red was barred from using chemical weapons, forbidden to fire on V-22s and C-130s, and at one point ordered to keep its air-defense radars switched on so they could be found and destroyed.[1][5] Van Riper stopped issuing orders in protest and spent the remainder of the game as an adviser. "I stayed on to give advice, but I stopped giving orders," he said. "There was no real point any more."[2] His judgment afterward was blunt: "War-gaming is not normally corrupted, but this whole thing was prostituted; it was a sham."[1]

The command's defense was candid in its own way. Gen. Peter Pace argued the refloat was simple economics: "You kill me in the first day and I sit there for the next 13 days doing nothing, or you put me back to life and you get 13 more days' worth of experiment out of me."[2] JFCOM's own classified after-action report, released in 2010, conceded the substance: "As the exercise progressed, the OPFOR free-play was eventually constrained to the point where the end state was scripted."[1] Van Riper's original 21-page observations memo, classified Secret, was finally released in 2024 after an eleven-year Freedom of Information fight.[3]

The exercise's legend has also drawn skeptical scrutiny, and that case deserves space here too. A detailed critical reading argues that the famous sinking owed much to simulation artifacts: the fleet's air-defense model had been partially disabled because it kept engaging simulated civilian airliners, and a real-world maneuvering restriction had placed the ships' digital positions far closer to shore than a wartime commander would ever have sailed them. The same critique notes that small boats have historically fared poorly against naval airpower, as at Bubiyan in 1991, when helicopters destroyed an Iraqi small-craft flotilla wholesale.[4] Both things can be true: the exercise was rigged after the fact, and the original result flattered Red. What matters for the present is which of Van Riper's methods turned out to describe reality.

The pages the 2026 war has already used

The answer, on the evidence of the spring, is most of them.

Iran declared the Strait of Hormuz closed to U.S.- and Israel-linked shipping at the war's outset and began laying mines in early March. American mine-clearance operations did not begin until April 11, and the Pentagon told Congress that fully clearing the strait could take six months.[6] This is almost exactly the scenario Caitlin Talmadge's 2008 study "Closing Time" had assessed, concluding that Iran, by linking mines, anti-ship cruise missiles and air defenses, could halt or impede traffic in the strait for a month or more.[12] Tehran appears to have lost track of some of its own mines in the process, an echo of the 1987 Tanker War, when the reflagged supertanker Bridgeton struck an Iranian mine of 1908 Russian design on the first convoy of Operation Earnest Will.[19]

The small craft followed within days of the mines. Iran deployed explosive skiffs disguised as fishing boats, the civilian-mingling tactic from Millennium Challenge updated with remote control: a tanker was struck by an Iranian uncrewed surface vessel on March 1, two more by explosive boats on March 11.[7] In a single engagement on May 4, American forces destroyed six Iranian small boats while intercepting cruise missiles and drones.[8] Satellite-navigation spoofing and jamming disrupted commercial shipping across the Gulf, and by early May more than forty merchant vessels had been attacked or seized.

The saturation logic of Van Riper's salvo appeared against land targets rather than ships: by the fourth day of the war Iran had fired more than five hundred ballistic missiles and two thousand drones, and by April it had run ninety-five distinct strike waves. The verdict of post-campaign analysis was that the air war had exposed the limits of what strike power can achieve against a mine-and-swarm naval denial strategy in confined waters.[9]

The pages not yet turned

What has not happened is the centerpiece. The signature event of Millennium Challenge, a coordinated, pre-emptive, multi-axis missile-and-swarm strike against an American capital ship in confined waters, has no confirmed real-world counterpart. Iran claimed a cruise-missile hit on the carrier USS Abraham Lincoln in late March; the Navy denied the missiles came close, and no imagery or independent confirmation ever surfaced, though some form of engagement likely occurred.[11] No American warship has been confirmed sunk or seriously damaged in the war. The June 9 Apache is the first confirmed American airframe lost to Iranian action.[25]

Part of the reason is that the United States moved first against the instrument. In the war's opening week, American strikes destroyed more than 120 Iranian naval vessels at their moorings in Bandar Abbas, Chabahar and Konarak, gutting the conventional fleet pierside before it could sortie.[10] That operation has its own clear ancestor: Operation Praying Mantis, April 18, 1988, when the U.S. Navy, retaliating for the mining of USS Samuel B. Roberts, sank the frigate Sahand, the missile boat Joshan and a clutch of Boghammar speedboats, and crippled the frigate Sabalan, in a single day, the Navy's largest surface engagement since the Second World War.[18] The 2026 version simply arrived before the Iranian fleet left port. But the asymmetric layer, the skiffs, the shore batteries, the mines, the drones, survived the pierside destruction, and it is that layer Van Riper's salvo was built from. The pieces of his opening move all exist and have all been used separately. They have not yet been used together, against a warship, at a moment of Iran's choosing.

The quieter pages are harder to see by definition. Whether Iran's surviving command structure has adopted Red's communications discipline, couriers, signals, silence, is unknowable from open sources, though three months of partially successful decapitation strikes give Tehran every incentive Van Riper articulated in 2002.

What the older games said about the wider war

Millennium Challenge was a tactical rehearsal. The games that examined strategy have aged just as pointedly.

In March 2012, U.S. Central Command ran Internal Look, a two-week classified command-post exercise simulating the aftermath of an Israeli strike on Iran's nuclear program. As reported by The New York Times, the game forecast that the strike would pull the United States into a wider regional war, with Iranian missiles striking a Navy warship in the Gulf and killing roughly two hundred Americans, while the strikes themselves set Iran's nuclear program back only about a year, and follow-on American strikes added no more than two.[13][14] Fourteen years later, the structural prediction has tracked closely: an Israeli-initiated campaign drew America in, Iran's underground nuclear infrastructure substantially survived, and the war widened across the region. The specific trigger, two hundred dead sailors on a burning warship, has not occurred. It remains the threshold event every Gulf commander has been primed to expect since 2012.

A Brookings crisis simulation that September added a finding about Iranian decision-making that reads like a brief for this June. The team playing Iran reportedly concluded that having spent years publicly threatening to act in the Strait of Hormuz if attacked, it could not back down once an actual American attack came; its own declared red lines had become a commitment trap.[16] An Israeli simulation at the Institute for National Security Studies the same year found the opposite dynamic in Iran's proxies: Hezbollah and Hamas, played by Israeli analysts, proved reluctant to do anything that would provoke massive Israeli retaliation, and escalation stayed bounded.[17] Both findings have held up. Iran has answered every strike on its territory; its surviving proxies have been notably more careful.

The oldest lesson comes from a game that never mentioned the Gulf. Proud Prophet, played at the National War College in 1983 with more than two hundred participants and real war plans, tested the era's theories of limited and controlled escalation against a free-playing adversary. Every variant ended in catastrophe; the careful, signaling-based strategies collapsed once both sides actually played them.[15] The June rekindling is a miniature of the finding: a single downed helicopter, with no fatalities, produced strikes across four countries within forty-eight hours.[26]

What the real wars add

Before there were war games there were wars, and the ones fought against Iran and Persia form a short, consistent syllabus.

The 1856–57 Anglo-Persian War is the template for everything the United States and Israel have done since February. Britain, confronting Persia over Herat, landed expeditionary forces near Bushehr, bombarded and occupied the port, won a sharp field battle at Khushab, opened a second littoral theater at Mohammerah, and never marched up-country. Persia conceded British aims within five months.[21] Coercion from the coast, with strictly limited objectives and no attempt on the interior, worked.

The August 1941 Anglo-Soviet invasion, the only successful conquest of Iran in the modern era, shows what the alternative requires. It took two great powers attacking simultaneously on multiple axes, the Soviets from the north, British and Indian divisions from Iraq and the Gulf, against an army that chose not to fight in depth, and the campaign was over in days. The objectives, Khuzestan's oil and the Trans-Iranian Railway that became the Lend-Lease "Persian Corridor" to the Soviet Union, sat on the only flat approach into the country.[22]

That flat approach has been the geographic constant of every campaign against Iran. The Zagros mountains run sixteen hundred kilometers along the country's western face, climbing from the sea-level Khuzestan plain to four-thousand-meter ridgelines in roughly a hundred kilometers; the roads and railways through them thread a handful of narrow passes that defenders have used since antiquity.[24] In the winter of 330 BC, the satrap Ariobarzanes held one of them, the Persian Gate near modern Yasuj, against Alexander the Great, walling the defile at a bend and raining boulders and arrows on the trapped column; by most accounts the position held for weeks, and fell only when prisoners showed Alexander a goat path around it.[23] Twenty-three centuries later, Iraqi armor crossed the Khuzestan flats in days in 1980 and then spent eight years failing at the foothills.[24] Iran's own most audacious operation of that war, the night crossing of the Shatt al-Arab at al-Faw in February 1986, when thirty thousand troops in small boats took the peninsula in roughly a day behind a rainstorm and a feint at Basra, remains the region's standing demonstration of what a planned amphibious surprise can do.[20]

No party to the 2026 war has attempted the ground page, in either direction. The entire conflict has been fought in the air, at sea, and along the coasts, which is to say inside the 1856 template and outside the 1941 one.

The unused inventory

Read together, the rehearsal record sorts into a ledger. The tactics already cashed in 2026: minefields in the strait, swarming small craft, suicide skiffs in civilian traffic, navigation spoofing, missile saturation of land targets, and the pierside destruction of a conventional fleet. The tactics still sitting in the record, used once in a game or a war and not yet in this one: the coordinated salvo-plus-swarm strike on a capital ship in confined waters; the courier-and-silence command system that blinds signals intelligence; the warship casualty event that Internal Look treated as the war's true widening point; an amphibious surprise on the al-Faw model; and any ground operation at all against, or through, the Zagros.

None of this is prediction; The War Dispatch confines its forecasts to the labeled Forecasting sections of its conflict coverage. It is an observation about inventory. Wars tend to draw down the stock of rehearsed ideas before they invent new ones, and this war's combatants have so far drawn down the cheap shelf. What remains in the record is the expensive shelf, the moves that games and history priced in carriers, divisions and months. The June fighting is a reminder that both sides are still browsing.

Open in the interactive site →