Saturday 15 June 2019

Models and Targets

Here, as they say, he goes again. I have been saying for quite some time that wargames rules, and associated figures and terrain, are models or, strictly speaking, sets of models. To that end (among others) I have been reading this weighty academic tome:

Weisberg, M., Simulation and Similarity (Oxford: OUP, 2015).

Now, Weisberg’s book is a discussion of the role of models in science, mostly, and, as such, turns on a number of distinctions which might not apply to a wargame. However, I think the overall description of models that Weisberg comes up with does apply, and I also think that there is some assistance in the book with regard to some of the things about models and wargames which have puzzled me recently.

Overall, Weisberg describes a model as being one end of a system, the other end being the target, the real world object being modelled. Thus, in the Bohr model of the atom, the model is the mathematics and concepts associated with it. The target is the real world hydrogen atom and its spectrum. Similarly, in the predator-prey system, the model is the linked differential equations and the real world target is the number of foxes and rabbits (or whatever). Finally, of course, a model of the battle of Waterloo has the historical event as its target.

Weisberg proposes that there is a similarity function between a model and its target. That is, the modeller tries to maximise the overlap between the phenomena the target possesses and those which are part of the model. The modeller also tries to reduce the number of structures in the model which have no real-world equivalent. They may also, in due course, make the model more complex to represent more features of the target, of course.

I have noted before how this tends to happen in physics, specifically atomic structure calculations. The model is created and results calculated. Broad agreement, say with the energy levels of carbon, are found, but then fine structure is incorporated by considering electron spin-orbit coupling, and then hyperfine structure and so on. A broad brush model is refined to something that really quite closely resembles the target through a process of refinement.

In wargaming, of course, things are not quite so simple or predictable. There is only one historical incident which we can call the Battle of Waterloo. We can, and do develop many sets of models for this target, but each model will (or may) emphasise different elements of the known elements of the target. Thus a model developed by someone with an interest in artillery may well differ from someone who believes that cavalry were the decisive arm. If we add to this the variety of background assumptions that have to come with any model (for example, most sets of wargame rules rely on arithmetic, geometry and the fact that cannon-balls move in a straight line) then we can land up with a complex set of complex models, all of which purportedly target the same historical event.

The practical upshot of this seems to me to be that there is never going to be a perfect set of wargame rules. This, to most wargamers, is hardly likely to be a startling revelation, but hopefully it might give some of the evangelists for one rule set or another pause for thought. The point here is that different models (even in the sciences) pick out different phenomena as the important ones to simulate. Some others, which other people might think are important, fall by the wayside. Thus the grounds for many a dispute over rules and interpretations are laid.

It is possible to have a model system which has no specific real-world target. Thus, for example, Richard Feynman famously modelled a perpetual motion machine. Naturally, everyone knows that such machines are impossible; there is no target real-world system in this case. What is interesting for this is why the machine is impossible. The model, however, still has no target.

I wonder if this might be the answer to my question of a few posts ago about ‘historical’ wargames without historical armies or events. For that matter, it could also be the answer to what fantasy and science-fiction wargames are about. These are certainly models or sets of models. They satisfy the criteria for being a wargame – figures, terrain, rules and so on. However, there is no real-world target for them. In this sense, then, they are exactly target less models.

A target less model is not necessarily a bad thing, as the Feynman example tells us. It provides some insight into how the world works, even if the upshot is that the world does not work like that. A wargame of a generalised Napoleonic battle may still generate insights into how Napoleonic warfare proceeded, and why some things happened as they did. Of course, it is a little more difficult to see this in fantasy and science fiction wargames, although some things (like basic physics) remain the same. I was once helping out creating a set of science fiction rules and got into an argument about why laser weapons were not subject to the inverse square law of diminishing effectiveness. Sadly for those why tried telling me I was wrong, I do actually know what I am talking about here.

Still, most fantasy and science fiction wargames are based on some aspects of real life. The best fiction in the genres is actually related to the world as it is (or was when the work was written). That, of course, raises the rather ghastly concept of the Warhammer universe being a comment on the ways politics have been developing in the later twentieth and early twenty-first centuries. This may be hard to imagine, perhaps, but also alarmingly possible.

Anyway, I think that the answer to my question of a few posts ago is this: my wargame was a model, but had no real-world target. That does not invalidate the model, of course, but in my case, as a historical wargamer, I probably need to use real-world targets to check that validity.

No comments:

Post a Comment