Saturday 28 April 2012

General Problems

Inevitably, I suppose, I’ve run in to a major problem in writing rules for the classical period, by which I mean roughly speaking Greece from Marathon to the Successors. The problem is with generalship, and so I’m putting it here in the hope that someone can have a good idea to solve it. I’ve been reading a paper by Meissner (Journal of Military and Strategic Studies (2010) 13, 1, 4-27) and it has rather thrown the issue into stark relief.

Basically speaking, early Greek generals did not do very much. At Marathon in 490 BC there were ten generals and a Polemarch. So there was no central command, and the Athenian forces were commanded on a ‘tribal’ basis. This system itself only dated from 510 BC, and the generals, or ‘army commanders’ came along a bit later to supplement the polemarch (‘war leader’), which was an aristocratic post. Anyway, as you probably know, the strategoi were divided and Militades waited until it was his day in command to attack the Persians.

The invasion of Greece in 480 BC forced further changes. Cities entered into alliance, set up a council to debate decisions and had an overall commander, a Spartan on land and an Athenian by sea. The Greeks actually did not agree on strategy very closely – the Spartans preferring to defend the Isthmus of Corinth and the Athenians wanting to be further forward, largely because Athens would be undefended. After Salamis, in the winter of 480 – 479, this nearly broke the alliance.

Incidentally, Herodotus’ description of Thermopylae indicates that the Greeks varied their hoplite tactics, by introducing feigned retreats to draw the Persians on. Try finding that in any set of wargame rules on the period…

Anyway, control of the army was exerted directly by the Spartan general Leonidas. He, despite his own doubts and preferences, carried out the orders he had been given. There was no local council to decide on the battle – the strategic plan, of defending the pass, had been drawn up and Leonidas was to execute it. There was a supreme command and a council of war making the decisions, and no actual argument on the battlefield, a significant change from Marathon.

As the century evolved, so did command structures. In Athens the democratic government kept control over military operations by popular assembly and the personal accountability of post holding commanders. Accountability for the vast cost of military adventures was, in particular, kept by the assembly. The commanders of expeditions were given their orders and resources, although obviously, the local commanders in theatre were able to make some sorts of decisions. They could not, however, simply go somewhere else.

Obviously, this sort of control could occasionally go awry, particularly after the Peloponnesian war when mercenaries were attracted into Persian service. Xenophon gives an account of his adventures after the battle of Cunaxa, which sees to Ten Thousand reorganise itself (on a democratic basis) after the main commanders had been killed, and march to the Black Sea.

With Xenophon we arrive at a period when there started to be theorizing about warfare, strategy and tactics. Xenophon himself was at the fore of these developments, although his writings about them are mainly part of his attacks on the Sophists. Xenophon argues that the Sophists teach tactics, but not logistics or judgement, and certainly not strategy. Nevertheless, there was clearly a debate to be had, and we have now come a long way from Marathon in just one hundred years.

Alexander, of course, bought his own particular (and possibly peculiar) personality to the whole question of strategy and tactics. Due to his victories the world, and the requirements of the military situation, changed. While the campaigning was fairly straightforward and the aim of the battle was the destruction of the enemy army, the control of the general, Alexander himself, could be clear, direct and personal. It was part of the mystique of Alexander that he intervened at the critical point, when the battle was in the balance. In short, he did the hero thing.

When, however, Alexander was having problems in, for example, the Hundu Kush against, essentially, an insurgency, his charisma and control started to slip. The Macedonians had to operate in smaller groups under lesser commanders. Overall direction had to be decentralised; there were, simply insufficient Alexanders to go around. This lead to dissention in the high command and, ultimately, violence as Alexander felt he had to act against some of his commanders. Perhaps they really were plotting against him. Or maybe he was just getting paranoid.

Nevertheless, we now have a more or less complete change in the mode of command of an army. From a diverse group of ‘democratically’ elected representative, through a supreme command with a council, to a charismatic leader with a group of companions to a diffused anti-insurgency campaign, we see a major alteration in the method of command.

The problem is that all of these, if we are going to have a single set of rules to cover the period, have to be assimilated in one command system. It will not be good enough, I think, just to assume that Militades is simply an earlier version of Alexander. On the other hand, the other generals at Marathon may have had more influence than we think; who else ordered the Greek wings to turn in, after all, but the tribal generals?

I’m not sure I can really think of another period of history where the structure of command changed so radically in one ‘era’. Granted things did change during, for example, the English Civil War, but not quite so significantly, and Roman command structures remained pretty well as they were during the early Empire.

So, short of having something really complex, how can this be accommodated in a rule set? Or am I worrying to no sense?


  1. Hi,

    First, thank you for your thoughts, I like to follow them very much. Especially because from time to time I am writing wargaming rules myself.

    I think the problem in this one may be in fact that periodization is always faulty. Strict cuts of historical periods have no place in real life. We can observe "Constantinople was sacked today", but we cannot observe "The Middle Ages ended today". Even if we were aware of the fact that somebody calls "our" present time "the Middle Ages".
    So, maybe the "classical period" in Greece military history is simply too broad to grasp as one? I think something similar happens with most or all time periods when we consider military actions. For example, ways the war was faught in 1939 were much different than in 1945, and this is not a long timespan. Same may be said about revolutionary France/napoleonic wars.
    Military art evolves all the time and strict periodization (in the meaning that all things remain largerly unchanged during given period) is probably close to impossible. It may even be argued that changes in military art evolve faster than in other areas of human life, at least during a conflict. The reason is that all failures are paid dearly in lives and resources (and personal careers), so motivation for development is really great.

    After saying that, it comes to mind that writing wargames rules may be even more difficult than making up something coherent, playable and giving nice impression of "inherent military probability". As if it was not that easy anyway...

    Best regards,

    Adam Lewandowski
    Warsaw, Poland

  2. Hi,

    Thanks for your comments, and I agree.

    The problem is that the end of this discussion leads us to design rules for every battle, because each one is contingent and needs special rules to get it "right", whatever 'right' means. And that way madness lies, I think.

    Phil Sabin's Lost Battles takes the historical outcome as being 'average', in some sense (or I suppose we could say 'normative'), and then explored variations from it. The problem is that battles are contingent and unique events, so statistics do not really help us.

    On the other hand, we cannot just give up.

  3. Hi,

    You are right, but rules for every battle may be viable. In a way, most boardgames depicting a single battle do this.
    Not that I am a strong supporter of such direction of rules writing, but it is possible in a not very complicated manner: prepare main core of rules as standard ("average") version and for battle X change points 5 and 11 for solution A and B. For battle Y change point 3 for C, 7 for D etc. etc. This need not to be too complex as there are usually only few specific features which should be addressed (at least few we know about). So we come back to the basic problem of how to write good core rules.

    Coming to think of this, comercial rules vendors should be pleased, as they could sell many supplements (one for every battle). Also this could be more interesting for players than the usual army list supplements - battle description, maps and similar could be included in addition to specific army list. I mostly use home rules and therefore I am not a huge buyer of commercial rules, but having a nicely studied and prepared battle scenario might be a strong incentive to change my habit. Even if I would be using home rules anyway.

    Best regards,

  4. Yes, I think boardgames (or some section of boardgaming) does go for the specific battle - I've always thought that very limiting myself.

    I guess the problem with your suggested approach could be that people would then start arguing over which interpretation of the battle should be used.

    Some rules do have options, mind you, like PM: ECW which has additional rules for cuirassiers, highland supermen, Scottish archers and so on. But they were because people expect them to be there, not because we thought they were important or made a difference.

    My experience of publishing, for what it is worth, is that no-one writing or publishing wargame rules is going to make a fortune, no matter how many supplements are published (I guess GW could be an exception). But something on the web could be managed, I guess.


  5. Hi,

    I am not into business of publishing, so you are probably right, I just came with a quick thought. It still looks nice to me, even if nobody will go that way.

    Back to more important thing - interpretation. I think you showed the problem in your blog - if participants of actual battle sometimes do not agree on major facts, so how could we as players? But this is not an inherent problem with "battle rules" approach.

    However, I do not think there would be much dispute about battle scenarios. If we have the data, most specifics are more or less established and the casual player will not have much problem with it. Even if one thinks there were less than 100 thousands Gauls in Alesia relief army and another thinks there were more than 300 thousands we just end up with variants of the same battle. Writer of the rules may just prepare his version and say "If you like to play it hard, take three times as many Gauls."

    If we do not have sufficient data (which happens quite often), well, than we have to come up with something with what scraps are available. Than we can only put up argument of "My version is way better than yours because I say so." However, for the casual player it may be tempting to receive ready to use scenario on some less known fight, assuming that the author made some research. Making this research by oneself may be hard or impossible, depending on skills and access to sources.

    I think I will stop my typing now, because it turned off course and I think it starts to belong to some forum better than here. And if you allow me I would probably fed you up with comments. :)

    Best regards,

  6. Hi,

    I'm not fed up at all with your comments.

    It would be good if we could have some rules with battle specific modifications, but I do not suppose they would sell well. Most rules do not, after all; I don't think PM: SPQR has broken even yet.

    I think we may need some basis for determining what is acceptable in terms of interpretations, but I guess any reasonable popularisation of a battle would generally be OK (some of the recent Ospreys are decent; older ones a bit less so).

    I think the 'Lost Battles' approach, the idea of which is to refight the battle multiple times varying one parameter at a time might be the way forward.

    For example, Sabin observes that the angle of the battle lines at Marathon is less important than the number of Persian infantry present and, as far as I recall, that the presence or absence of Persian cavalry is less relevant too.

    I hope this is based on multiple refights - I guess it is - so it should give us some insight into the original, but, of course, it is only useful if the underlying model is a 'good' one, whatever that may be (back to interpretations again, I suppose).

    Maybe the outcomes could verify our rules, but that does seem to be a bit of a circular argument. On the other hand, if everything followed history, we would no longer have a game...

  7. Hi,

    So encouraged, I will write a bit more.

    The method of playing the same battle many times with different variants of rules is a good one. I always wanted to do this, but it is too time consuming for me. When I got time to play, I prefer to play something new. Concerning rules development it has the fault of different situation every time, so refining details is much more difficult. Yet it is still possible, although it takes years to achieve. :)

    On smaller scale, I do refights with different variants of rules as stand alone fragments of bigger fight. Than I can check some separate features in say, 20 minutes per solution in a kind of closed situation.
    The most testing I did was probably when I checked how probability on some string of events goes. I would have to look for details, but it went something like this: check for fire of one side, check for fire of the other side, check morale, check close combat results, check for casaulties. My math skills on counting probabilities failed me in such a long string (the tree of results was rather big), so I made around one hundred rolls (I was not able to go for required 300000;) ), noting the empirical results. They showed me some trend and I adjusted, tested again and so on until I was satisfied with the effect. As far as I remember this was about attacking tank unit with infantry anti-tank weapons, so I had some statistical data on actual results to compare with.

    In your last sentence I think you hit a very important feature of wargame rules writing. If we have written the rules which provide perfect historical results, we would end up with the same results every time. In fact we do not need any rules for this, as we can put our toy soldiers on the table and move them around according to our best liked book on particular battle.

    Yet, we want to game, and in order to we have to start changing history. Maybe just a little bit - for example Gustavus II Adolfus is not hit at Lutzen or does not die. Than we start to ponder over what happens next. And we land down in alternative history. If so, any rules are inherently flawed with the fact that they deviate from history from the start and no matter how we try, we cannot get to the bullseye, just strive to remain close by and get alternative history results which satisfy us. This may be the reason why so many rule sets seem so far away from historical results. They simply have to, and deciding how far away is enough remains in the domain of individual liking. De gustibus non est disputandum, so the somewhat confusing conclusion is that the good rules are the ones oneself likes best.

    P.S. Eventually, the answer to frequently returning accusation "But this is not historical..." may be: "But I like this version of alternative history which allows Roman centurions to chop through 10 ranks of stubborn enemies and remain unscathed. It is so heroic..." We have nothing to counter the statement.

    Best regards,

  8. I think that you are probably worrying to no purpose over this. Part of the culture of 'the wargame' is that every force is commanded by a player-general (or generals) who have a reasonable opportunity to command their force to victory. All that is easily achievable is a nod in the direction of making Greeks more difficult to control than roman legionaries or Napoleonic French - in the context of Polemos, using the tempo point system to reward simplicity and punish complexity in the Greeks' operations.

    Take the DBR example - all those interesting Native American C16 armies are commanded in exactly the same way as the New Model Army, but it still makes an okay game

  9. I suppose the question is 'how historical is historical?', and that is a question each individual has to answer for themselves, from their reading of history.

    DBR (might) give a reasonable game for Aztecs against Mixtecs, but it is possible that it doesn't work for the NMA.

    I suppose that the problem with probability is the complexity of the calculations; we cannot know enough to really have a good idea of what might happen to a cannon shot except in some sort of blurry, abstract way. Similarly, the could, in principle, calculate the trajectory of the pistol shot that laid Gustavus low, but we have no chance of actually making a calculation.

    So we have to abstract, average and make sensible guesses. All we can really do is work with the outcomes. It isn't quite a disaster, but it is a lot more complex than it appears.

  10. Well I was thinking more strictly of the 'command' side. A d6 PIP roll hardly gives the flavour of commanding Aztecs or the NMA, but given that the parameters of wargames are generally 2-sides of 1-or-more players without a 3rd party to referee, this severely limits the complexity you can put into a playable wargame in terms of command mechanisms. Shift those parameters and you can change a lot - committee games, 1 player commands 1 group, actual replication of historical command structures and so on.



  11. I see what you mean. I guess this goes back to game mechanics and levels of abstraction in the command and control rules.

    After all, Phil Barker argues that C&C rules in wargames are to stop generals doing thing, rather than enable them.