The game theory behind violating ceasefires

ThE CEASEFIRE between America and Iran got off to a violent start. After it was announced on April 8th, Israel intensified its bombardment of Hizbullah fighters in Lebanon (which Israel says was not included in the truce). In the early hours of the ceasefire, Iran launched drones and missiles at other countries in the Gulf. Bahrain says it was attacked by Iran’s proxy militias earlier this week. Iran is also yet to resume the normal flow of traffic through the Strait of Hormuz, and President Donald Trump has instructed the US Navy to impose its own blockade.

To many, this is not how a ceasefire is meant to unfold. But although violations can test a fragile, perhaps ephemeral truce, they need not scuttle negotiations. Ceasefires have a history of being frequently and cannily flouted without regressing fully to all-out war. The agreement to end fighting in Gaza, for example, remains in effect (officially, at least) after six months, despite Israeli air strikes as recently as this week.

The Economist collected statistics from the Armed Conflict Location and Event Data (ACLED) base on 12,333 violent events in the 60-day period surrounding nine ceasefires reached in the Middle East and nearby countries after 2020. Though armistices reduced fighting, none resulted in a total cessation of violence (see chart). On average fatal events fell by 81% in the 30 days after a ceasefire began, compared with the 30 days before.

The Armenia-Azerbaijan armistice in 2020, followed by only sporadic violence, was the most successful of the nine (despite a brief return to fighting in 2023). By contrast, temporary truces between Israel and Hamas in late 2023 and from January to March 2025 saw fighting quickly return to pre-ceasefire levels once they ended. Most ceasefires fell somewhere between those two extremes. In these cases, violence declined in the immediate aftermath of a ceasefire taking effect, but never to zero. In some instances that lower level was sustained; in others, violence gradually escalated to near pre-ceasefire levels.

The fact that ceasefires rarely prevent all violations does not render them meaningless, argues Valerie Sticher of the Centre for Security Studies, a research institute at ETH Zurich. Ceasefires, she says, are often “a form of strategic bargaining” used to recast the terms of a conflict rather than simple humanitarian acts. To adapt freely from Carl von Clausewitz, a 19th-century Prussian military strategist, it may be more useful to think of ceasefires as the continuation of war by other means.

Many ceasefire violations are accidental. Opposing factions inadvertently bump into each other on the battlefield, or aggrieved spoilers on one side attempt to derail peace talks by stoking up the violence. That does not appear to be the case in the third Gulf war. Although, in theory, the decentralised command structure of the Islamic Revolutionary Guard Corps allows for the possibility of a lone commander jeopardising talks, recent attacks confirmed by Iran’s state media appear to have the full backing of the regime’s leadership.

What to make, then, of violations in the first week of the Iran ceasefire? One explanation is that all sides are deliberately violating the truce to gain a strategic advantage, and are subsequently retaliating to deter each side’s breaches. Literature on the game theory of ceasefires suggests that agreements to pause overall hostilities introduce a threshold below which attacks can continue. The exact level of this threshold is often unknowable, but it is usually managed by both sides through a “tit-for-tat” strategy whereby each provocation meets an equivalent response.

Strikes like Israel’s in Lebanon, or those of Iran’s proxies, can put pressure on opponents at the negotiating table. Attacks above the threshold can see truces fall apart altogether, and return to the sharply escalatory dynamics observed in the Gulf before the agreement.

The violations may also reflect an understanding, refined over decades by strategists in America, Israel, Iran and Hizbullah, Iran’s Lebanese proxy, of the accepted rules of lower-intensity conflicts. Though Iran had not fought a conventional war from 1988 until last year, it has plenty of experience in low-level fighting, often despite formal truces.

Iran and Hizbullah learned how a weaker actor could create what Iranian authorities call “deterrence equations”: tacit red lines enforced by tit-for-tat attacks. Hizbullah would respond to Israeli attacks (and vice versa) in kind, with the same number of rockets and casualties—often even at the same time of day.

The long-running conflict in Lebanon and northern Israel was “Iran’s strategic laboratory”, says Daniel Sobelman of the Hebrew University of Jerusalem. Reciprocal ceasefire violations enabled Hizbullah to restrain Israel, a much stronger adversary, without triggering full-scale war.

Once the third Gulf war was under way, Iranian state media said the country wanted to reset its deterrence equation. Iran repeatedly promised to match attacks on its civilian infrastructure with retaliation on energy infrastructure and desalination plants in the Gulf. That has prompted Mr Trump to back down from his most extreme threats. Those threats, along with the closure of the Strait of Hormuz, blunted America’s ability to escalate and introduced “symmetry” into Iran’s fight with America, says Mr Sobelman.

When escalatory dynamics end, the talking can begin, says Danny Citrinowicz, a former Israeli intelligence officer. The ceasefire gave all sides an opportunity to recalibrate, if not to end provocations.

Truces often yield more peaceful long-term conditions. The Ceasefire Project, a research group, catalogued 2,203 declared truces between 1989 and 2020. About half of them concluded without a return to war or were still in force when the study ended. A third of them failed. Ceasefires in the Middle East were the weakest of the lot. More than half of the region’s 360 truces with a known endpoint failed.

Agreements are weakest early on, the researchers find. And despite Mr Trump’s confident prediction that a pause in fighting he brokered between Israel and the Islamic Republic in June 2025 after the 12-day war would last “for ever”, America’s decision to resume bombing Iran in February shows even short-term success is no guarantee of lasting peace.

Monitoring by third parties improves the chances of a ceasefire lasting. A number of countries, including America, have had observers on the ground in Israel since the beginning of the ceasefire in Gaza. But the war in Iran is different because America is a direct combatant, and is therefore unlikely to allow a third party to police its presence in the region.

That is worrying because working out a peace deal can be an arduous process. The Good Friday agreement to end the troubles in Northern Ireland in 1998 took 700 days to negotiate. Pushing talks into the future can help extend a truce, but also delay resolution of the conflict. The ceasefire in Gaza, for example, is still stuck in “Phase 1”, with the two sides unable to agree on a more lasting settlement. The Korean war has been paused since 1953 without a formal peace.

America is contemplating extending the current ceasefire to give more time for talks. Perversely, each side’s willingness to return to war, including by violating the ceasefire to gain leverage in the talks, is also a point of leverage at the negotiating table. Israel is content to see the war continue. Iran appears to prefer war over a bad deal. “If Trump decides to turn back to war, Iran will turn back, too,” says Mr Citrinowicz. “Iran is like a poker player that is not only calling but also raising.” ■