Back: Articles and information on Clem Sunter

Nuclear game theory: time for a review

By Clem Sunter

On 30 October 1961, Russia detonated the most powerful nuclear weapon of all time, before or since. The site was a remote archipelago in the Arctic Ocean, north of the Russian coast. The blast was equivalent to between 50 and 58 megatons of TNT or 1 400 times the combined power of the two atom bombs dropped on Japan in August 1945. Think of taking one step a metre long versus walking 1.4 km. That is how huge the bomb's power was by comparison.

Named "Tsar Bomba" and nicknamed "We'll show you", the device weighed 27 tons and measured 8 metres by 2 metres. The subsequent mushroom cloud was 64km high but, more importantly, it destroyed all buildings in a village 55km from ground zero. Dropped centrally on any major city today, it would virtually flatten it in a single blast and instantly kill between a quarter and a third of its inhabitants.

The MAD logic

The reason I am giving these statistics is that classic game theory examines every possible strategy a player must consider when taking into account all the possible responses of his or her adversary. It obviously applies to nuclear conflicts as well. Right now you have the two nations most heavily armed with nuclear weapons locking horns over Ukraine - Russia and America. They did so in 1962 over the Cuban missile crisis and narrowly missed a full-scale nuclear exchange as a result of mutually misunderstanding each other's position. Once again, they are eyeball to eyeball like two chess players poring over the next move.

John von Neumann, a US Hungarian mathematician, was the founder of game theory in 1928. He later coined the acronym MAD which stands for "mutually assured destruction" in the nuclear game. The argument was that neither of two players would resort to nuclear conflict if a first strike by either one triggered a response by the other which caused an overwhelming loss to the initial aggressor. Thus, a second strike capability on both sides guaranteed peace between them according to the principle of minimising your maximum potential loss. The latter is a core element of optimising your strategy under game theory (known as the minimax theorem).”

The current state of the nuclear game

In 1985 there were estimated to be 68 000 active nuclear warheads around the world. As a consequence of disarmament treaties since then, the figure is now put at 4 100 with the bulk being fairly equally distributed between Russia and America. The remainder are in the hands of the UK, France, China, India, Pakistan, North Korea and Israel. To keep the equilibrium demanded by MAD, treaties between the two principal nuclear players have also limited the numbers of strategic launchers, heavy bombers and anti-ballistic missile systems owned by each. However, a destabilising factor is that while 17 300 warheads are reckoned to have been decommissioned, they have yet to be destroyed. Equally, a single inter-continental ballistic missile can deliver as many as 10 separate warheads at a time. It can also be accompanied by 40 decoys meaning that 50 rockets are required to neutralise an incoming missile. Secrecy means outsiders have no idea of the latest technical advances in attack or defence systems.

As an aside, Ukraine had around 5 000 nuclear warheads when it became independent of the Soviet Union in 1991. By 1996, all these weapons had been voluntarily sent to Russia for disassembly. I wonder how Ukraine feels now about that handover.

If John von Neumann were alive today, I would want to ask him the following questions. Maybe someone from the Institute for Advanced Study in Princeton, New Jersey, where von Neumann was based, is sufficiently well versed in game theory to act as a substitute:

1. Is MAD still a valid concept which makes a nuclear conflict highly improbable? Does the Doomsday scenario still predominate? Or do military strategists now play nuclear war games where they examine less apocalyptic aftermaths? If so, do they look at a conventional war preceding a nuclear exchange, or the other way around?

2. What would be the estimated damage and loss of life today caused by a nuclear conflict between the US and Europe on the one hand and Russia on the other? What harm would be done to the environment in the short and long term?

3. What would be the impact on the rest of the world and would the whole world recover from such a global war as quickly as it did after 1945?

4. What are the probabilities of the current game in Ukraine remaining localised or deteriorating into a direct military conflict between the major players with potential nuclear consequences?

5. Does the chance of terrorist groups acquiring nuclear weapons nullify MAD since they belong to no fixed abode which can be destroyed if they attack first?

6. How does emotion feature in the mathematical models of game theory since patriotism as much as logic can be responsible for the onset of nuclear war?


I have not come across a single word in the printed/social media, on the internet, on television or radio or anywhere in the public domain addressing this issue. As a scenario planner, you always look at the end game to decide what your next move should be. That is why Herman Kahn wrote his famous book On Thermonuclear War in 1960, followed by Thinking about the Unthinkable in 1962.

Neither Putin nor Obama talk about a quantum leap in confrontation; but they wouldn't, would they? Every step will be calibrated so that it does not cause too much anger between the Russian bear and the American buffalo. In the end, a compromise may be reached which is classified as neither appeasement nor victory, but which for the time being is a solution.

Alternatively, 2014 may go down in the history books as the beginning of an unanticipated slide into a third world war. Hence, it is now time to review nuclear game theory before the rivalry between Russia and America becomes unstoppable in its progress towards a nuclear showdown.

Back: Articles and information on Clem Sunter