Investigations of a Dog

Esther MacCallum Stewart has recently started a new blog at Glod’n'Epix which covers her work on computer games (in addition to her First World War stuff at Break of Day in the Trenches). Reading this has made me realise that I have no reason to be ashamed of liking computer games. Gaming isn’t a waste of time and can even be a valid subject for academic study. Meanwhile, Jeremy Boggs at Clioweb posted a link to an article by Niall Ferguson about a computer simulation of the Second World War. This led me to think about how games can be used by historians, and what they can and can’t tell us. It’s probably no surprise that I don’t agree with Niall Ferguson on this (or lots of other things).

The software in question is Making History: The Calm and the Storm. The original version was marketed as an educational tool rather than entertainment and was mostly sold to schools. Version 2.0 will be released to the public soon, but the demo isn’t available yet so I haven’t played it myself (and so a lot of my speculation could be wrong). Niall Ferguson believes that simulations like this can be a valuable part of learning about history, particularly by playing out counter-factual scenarios. He used it to test some of his hypotheses about the Second World War and found that his strategies didn’t work. “Play the part of Britain in September 1938 — during the crisis over Czechoslovakia — and you quickly discover (as historians have long maintained) that Britain’s pace of rearmament cannot be accelerated.” This is based on the assumption that Making History simulates the war accurately enough to make the outcome of a game a reliable guide to what would have happened in reality. Is Making History accurate enough for that? Can any simulation be accurate enough?

Over the past few years I’ve spent a lot of time playing Real Time Strategy games. This has taught me that every game needs to be approached on its own terms. If you bring in any preconceptions, such as historical knowledge, or common sense, you will lose. If you want to win you need to learn how to play the game, and that means unlearning what you have learnt from real life, and even from other games. There are some transferable skills which you can use to learn how to play strategy games, but the emphasis is always on learning how an individual game works. Structuralism is a useful way of looking at games: each game is an arbitrary system which is fixed in relation to itself but has no fixed relationship with the real world (the relationship between multiplayer roleplaying games and reality is potentially much more complex, but the structuralist analogy is good enough for strategy games). Which strategies are most viable depends on arbitrary decisions by the game designers. These decisions might be aimed at promoting gameplay or realism, but in either case they are still arbitrary.

If Making History shows that the pace of Britain’s rearmament couldn’t be increased in 1938, that is because the designers at Muzzy Lane have decided it should be so. Their decision might even have been influenced by the historians who long maintained it. On these grounds I agree with Ferguson when he dismisses all other Second World War games as unrealistic and ahistorical. However, he makes a special case for Making History because “it is based on a quite astonishing quantity of factual information about the war”. Even if we take a purely empirical approach and leave aside any theoretical questions about how much we can know about the past, this seems dubious. Any empirical reconstruction of the Second World War faces two major problems (in addition to the problems of definition that I talked about last week): not enough information, and too much information. We can’t reconstruct everything that happened because we don’t have a complete record. On the other hand, the quantity of surviving records is too vast to be dealt with easily. While I don’t have any direct knowledge of the Making History game engine or data, I suspect that its model of the war is very simplified.

Simplification might not matter. We don’t necessarily need to know everything about the war in order to simulate it at a strategic level. Common sense suggests that unknown quantities at the tactical and operational levels would tend to average out and not have a disproportionate strategic impact. But I’m usually suspicious of “common sense”. What if war is a chaotic system? In chaos theory outcomes are not random. They are determined by the system and the initial conditions. However, small variations in the initial conditions can lead to disproportionate changes in the outcome, making outcomes difficult to predict. If this applies to war, it makes simulations even more problematic. In order to simulate the Second World War accurately we would need to know far more about it than we currently know, or perhaps can ever know.

Perhaps one of the most unpredictable systems is the human mind. Different individuals can react in different ways to different circumstances. Whether Germany is played by a computer or another human player, it certainly isn’t played by Hitler himself. Then there’s the question of how far one individual, even a dictator, can really be said to be in direct control of a state, an economy, or an army.

Ferguson mentions that the branch of mathematics known as game theory grew out of attempts to simulate the Cold War. One of the limitations of this theory is that it usually assumes rational players who are aiming to optimise the outcome of the game. In practice, ideology can lead to irrational decisions (it could even be argued that the decision to go to war is always irrational because the costs of war nearly always outweigh the material benefits), while rational gameplay can lead to ahistorical outcomes.

If I was playing as Nazi Germany and trying to win the war, I wouldn’t want to commit genocide. This isn’t anything to do with bringing my own personal ethics into the game. It would just be a huge waste of resources which could be better employed fighting the allies. On the other hand, if I was playing as Britain I might want to murder all the Italian prisoners of war in order to save resources. In the real Second World War ideology didn’t allow this degree of agency. That could be simulated in a game by denying the British player the option of committing war crimes (although that in itself would be an arbitrary decision based on a particular ideological position), but what about the Nazis? Their racist ideology could be simulated by making wiping out the Jews one of their victory conditions, but I can’t imagine Muzzy Lane getting away with selling an educational game which encourages genocide. It’s interesting that Ferguson doesn’t mention this ethical dilemma despite placing racist ideology at the centre of his account of 20th century wars in his recent book War of the World.

Seeing games as tools for testing counter-factual hypotheses doesn’t overestimate them, it underestimates them. Games are games, not simulations. They need to be studied on their own terms. They are not research tools but primary source material for cultural history which are at least as important as novels, plays, films, and music. Looking at it this way, a game being an arbitrary system which is unconnected to reality isn’t a weakness. The design of the game tells us something about the cultural assumptions of the designers and the target audience. There is huge potential for the theoretical apparatus of cultural history to be applied to gaming. The idea of a player making the meaning of a game is immediately much easier to grasp than the reader making the meaning of a book. Games are very obvious examples of Baudrillard’s simulacra: systems of signs not related to an underlying reality. Representations of race, class, gender, sexuality, and the non-human in games will provide plenty of material for Marxism, feminism, queer theory, postcolonial criticism, and eco-criticism. A whole new field of game studies is opening up exciting possibilities for future research. This is yet another opportunity for military history. Representations of war in computer games can be just as interesting and important as representations of war in film or literature.