This turn: every roll
State after the turn
Combat results table notional
Escalation and politics
Oil and exports
Monte Carlo:
Notional model, not a prediction. Most numbers are assumptions you can change; sourced values are marked.
This turn: every roll
State after the turn
Combat results table notional
Escalation and politics
Oil and exports
Monte Carlo:
Set up. Pick the U.S. objective: seize the island and hold it as leverage over Iran's oil exports, raid it and withdraw, or blockade it from offshore without landing. Spend the 100-point budget on forces, then set Iran's posture with a preset or force by force.
Play. Each turn is 12 hours. Carrier aircraft strike Iran's coastal launchers, drone teams and boats. Iran fires missile salvos and drone waves at the ships, and destroyers try to stop them; swarms of fast attack craft (small armed boats) close in, and helicopters try to break them up. Landing groups cross mined approaches, air lifts face the island's air defense, and the two ground forces fight on the combat results table: the ratio of U.S. to Iranian strength picks a column, one die picks a row, and the cell sets each side's losses. Iran ferries reinforcements from the mainland and strikes troops on the island with drones. Each turn Iran also rolls to widen the war. The log lists every chance and every die: a roll (a random number from 0 to 1) below the chance means the event happens.
Monte Carlo. One game is one draw of the dice, so it can mislead. The Monte Carlo card replays your setup 1,000 times with the same seeded dice and reports how often the objective is met, the escalation that came with it, and the notional oil price. Key drivers changes one lever at a time, replays the same 1,000 games, and shows how much each lever moves both the objective and the chance of a major escalation.
The result box at the top of the setup panel gives the share of those 1,000 games that met your objective, the shares that ended mixed or lost, and how often the war widened into a major escalation.
Outcomes. Seize: the U.S. holds Kharg when it has secured the whole island and still holds it at the end; the landing is defeated when the force is destroyed or pinned and outnumbered more than two to one after the defender's terrain bonus; anything else is contested. Raid: the force must take the airstrip area and pull out; a bad withdrawal counts as costly. Blockade: success means Kharg loaded a quarter or less of its normal volume on average across the game, failure 60% or more. A major escalation is a strike on Gulf energy infrastructure, or an escalation index of 50 or more.
This is a teaching model in the style of a board wargame. It uses no real unit positions, beach surveys or order of battle, and it does not draw or name any part of the oil terminal. Iran's forces sit in a notional zone and the U.S. force in a notional sea area on purpose. The island inset shows two broad approach halves and the airstrip at its published airport coordinates, and nothing else on the island.
Only a few numbers come from open sources: Iran's pre-war crude exports of about 1.6 million b/d and Kharg's share of about 94% (Kpler; CRS and most outlets say about 90%), the Brent spot baseline of $71 on the last trading day before the 2026 war (EIA), and about 2,200 Marines in a Marine expeditionary unit, or MEU (U.S. Marine Corps). The menu items, Iran's force types and the rough 1979 garrison figure rest on open descriptions (DIA, CSIS, FRUS, The War Zone, USNI News). Every probability, cost, rate, force count, oil-price sensitivity, escalation weight and the combat results table is notional. The table below lists them all; edit any of them in the panel and the whole page updates.
Kharg is not hypothetical terrain. In 2026 the United States struck military targets on the island, blockaded Iran's ports and publicly weighed seizing it (see the timeline). This game is not a model of those events or of any plan. It borrows their general shape, and the history is there so readers can compare the model's assumptions with what happened.
The dice use a seeded generator, so the same link replays the same game and the same 1,000 Monte Carlo games. Strength points are notional: one point is about 1,000 troops. Distances and the island's size are measured on OpenStreetMap coastlines, simplified to about 150 m so no harbor structures show.
Left out: Iran's air force and submarines, cyber and space, U.S. bombers and land-based missiles, civilians and oil workers on the island (about 8,000 residents in the 2016 census), weather and sea state, the time needed to assemble the U.S. force, legal authorities, and any negotiation during the fight. Use the model to reason about how the pieces trade off, not to estimate any real operation.