Marginal Differences...

The Polemarch has written a typically excellent blogpost, riffing on some things about bias which I had mentioned in an earlier post on this blog. But, as these things do, it inspired some thoughts of my own on something he mentioned, namely game designs and the dependence of victory and defeat on marginal differences: he is experiencing something of this in his current Italian Wars campaign based on the boardgame, Machiavelli. I am experience something similar in test games of Europe in Agony, ready for my Thirty Years' War campaign.

To rehearse the argument, a game can be very fairly balanced in that all sides are equally likely to win, but this is in turn heavily dependent on some marginal differences in conflict outcomes early in the game. So in The Polemarch's example, Venice and Turkey are both highly motivated to control the Adriatic early on the game, there is no alternative strategy for either side which has the same pay-off, and the result of their conflict is highly unpredictable and basically dependent on small swings of fortune in the outcomes of dice rolls. Something similar pertains at the beginning of Europe in Agony, when the Protestants are highly motivated to attack Vienna, it is a 50:50 thing, but winning or losing the Early War scenario at any rate is highly dependent on the outcome.

And this is inherent in real warfare: it is often, not always, but often, a zero-sum game in which an action that benefits one side must harm the other.*  In real warfare this is one of the true importances of the 'fog of war': it stops decision-makers on either side from effectively calculating the utility of actions, and thus both sides can take actions which might ultimately benefit the other side.  But in a boardgame, this is relatively uncommon, and the odds can be calculated more exactly.  This then takes away the importance of the decisions, since the optimal path is clear, and then why would someone playing properly not take it?  This I guess also ties in to the thinking behind blogger Paul Leniston's recent decision to write a new set of Napoleonic rules after hundreds of games with a previous set: they had just become too well known between him and his opponent that (he felt) the games were becoming predictable, since the logic of the rules was too well known to both parties.

However, wargaming has another problem: it (very broadly speaking) aspires to some degree of realism. But there is nothing in the laws of nature which dictates that these problematic situations should be rare in military conflicts. So, in probabilistic terms, the initial set-ups of the wars mentioned above might be 'realistic' in the sense of generating the most plausible outcomes of the conflict; and this might be true at more tactical levels too. 

It is slightly difficult to know what to do here.  One can of course sacrifice everything to the game, but that is not always entirely satisfactory, since re-creating to some degree the issues and situations of history is a major attraction of historical wargaming. But, allowing that still grants us as rules-writers, scenario-designers and players a latitude, and we should always probably err on the side of the game where we can.  Specifically, I think we should be careful about designing rules and scenarios where the early turns are the most crucial rather than the least, as far as we can.

(* I am by no means a game theorist, but I assume that there are lots of outcomes in real wars which look like interractions between the two combatant sides but are really interactions between elements of the same side e.g. a combatant nation and a supportive country supplying weapons and money but not participating directly; this pressure helps to keep war from descending into total barbarity in all instances).