Some thought has been niggling away at the edge of my consciousness for a while, now, and I have finally managed to bring it out into the open and take a look. It was originally caused by a paper of generalship I was reading, and then tucked up as an interesting thought, but it has been re-awakened by a thread on one of the mailing lists I keep an eye on (but I forget which).
The issue is this: I have covered how generals can plan, react to events in the middle of a battle and so on. But what is much more difficult to ‘model’, in I think any sense, is the fact that general can, and did make mistakes.
I dare say that we can all think of mistakes that were made by commanders. Off the top of my head, Prince Rupert is often blamed for losing control of his cavalry at both Edgehill and Naseby rather than exploiting the fact that the Parliamentary infantry had a flank exposed. At Raphia, Antiochus attempted to exploit the victory of his right flank, while Ptolemy launched a counter attack with his centre. Antiochus, on incomplete evidence, assumed that he had won, and turned out to have lost the battle.
I dare say that these sorts of accounts could be multiplied throughout history, and these are simply the tactical or grad tactical errors. Strategic mistakes are also manifold. As many of you probably know, Monty’s third law of warfare is ‘Don’t start a land war in Asia’. Both Napoleon and Hitler made this mistake, and paid for it in spades. As a matter of fact, it could be argued that Alexander of Macedon made a similar sort of error when he invaded India.
As a wargaming issue, I am sure that you can see the problem. Wargamers, standing over the table watching all the action going on, are unlikely to make these sorts of errors. The information presented to them is much more complete than any real commander and his staff could possibly have hoped for.
As mentioned in a previous post, there are some things that you can do to limit the damage. Scenarios can be disguised, so that Edgehill looks like an American Civil War battle, or Gettysburg is presented as a Franco-Prussian war clash. All this is well and good, and at least might throw the players initially as they try to match the set up with the real battle. But once the action starts, the knowledge that the players have of the status of the units and the state of the battle is far more complete than any real commander.
It is, I think, very hard to find any rule sets that tackle this as a problem. I suppose DBM, with its irregular commanders who could lose control of themselves and their troops made a stab, at least, in the direction of making sure that troops did things that generals did not expect or even, particularly, like or want.
In a similar way, older rules had some sort of ability for certain troops to run out of control. For example, the old WRG Renaissance rules by George Gush classed (some) Royalist cavalry as ‘A’ class morale, which had to check morale when, for example, they saw enemy troops for the first time. All well and good, perhaps, except that it does not really account for the times when the Royal horse did behave as their commanders wished, or, for that matter, when the Royal horse did not just charge off on a whim (which in fact was most of the time).
The normal response to this sort of problem is for wargamers to start muttering the word ‘umpire’. However, in this case, I am not at all sure that an umpire, except under special conditions where the players are given highly restricted views of the table, really solve the problem. Most wargamers are sufficiently experienced in the way of wargame to know, roughly, how things are proceeding and how they might turn out.
Only if the commanders are in a separate room, fed incomplete information from forward commanders and subject to intermittent cannon fire can we really start to ‘simulate’ the sorts of conditions where commanders make real time mistakes.
Of course, we could argue that as normal humans, we make mistakes all the time, and that would be true. Even Grand Masters make errors in chess, but I suspect that, in the cosy world of a friendly wargame, general mistakes are less common than they would be in the real world.
One way of simulating mistakes might be via a campaign game, where it is relatively easy (using a matchbox system, or an electronic equivalent thereof) to create some uncertainty in the minds of commanders, and thus create the conditions where errors due to lack of information might occur.
This is fine so far as strategic moves may go, but does not really create the battlefield conditions where tactical mistakes might happen.
I suppose the most sophisticated attempt to deal with this sort of issue is the Piquet series of games (which I have seen but not played). Here, you can see what combats are occurring but not how it is going, and thus, at least, some of the information is concealed from the player. A similar situation can arise in a role playing game where the players can see that they are inflicting damage on the enemy, but not how much damage they can take.
The problem with both of these, as compared with a conventional wargame, is that both have a fairly high threshold for preparation. In a role playing game, the game master has to set up and control the non-playing characters, and asses the state of their morale (for want of a better word). In Piquet the card deck has to be created, and a whole load of equipment is needed to run the game (at least in the demonstration games I have seen at shows).
Maybe, then, this is a problem that does not have a particularly good or elegant solution, but until it is tackled, we cannot really gain much insight into the problems of command.
At our club we have invented something I will translate as "situation assessment test". We use it anytime we come across some situation problem which is not obvious to solve. The test is used to tell what a particular commander, in a particular moment is going to do. It is very universal, so we use it in different war eras.
This may be done in different ways, but usually we throw for: "visibility" (how well the commander is aware of what is going on), "optimism" (or lack of it), "skill" (how well the commander organized his actions). We can use the "know all" problem to advantage this way. As gamers we most often know what is most profitable to do in a particular situation. Also we see the more or less exact positions of troops, their strength etc.
For example, there is a chance of attacking enemy in superior situation - outnumbered, in bad formation or something similar. But when testing we may end up with the commander having low "visibility" so he will not take advantage as he is not sure what is really going on. Or he could see the opportunity but get a "low skill" assesment, so he could attack but instead of moving around the flank he will hit straigth ahead. We get quite a range of options.
When we test both sides or more units, it gets even better, to the point when we create a self-going game (if we test every commander through the chain of command in every situation). It is very nice to use this in creating information submitted between different commanders - when one has "high optimism" he will try to explain his high casaulties with more enemy casaulties or great advantages gained. When "low optimism" one will call for reinforcements despite already having numerical superiority. And so on.
This method also works great in campaign games which, I agree, make it much easier to receive feel of what the "fog of war" is.
I remember one ACW scenario, which was prepared by myself and I had part of units under my command. Obviously I knew the enemy's possibilities, which included two variants of outflanking march. I used the above method for simulating decisions by the commander I was playing, which ended in a great success for the other side, but the game was very good indeed and looked quite realistic, with "my comander" first deploying some forces on the flank, than sending part of them to attack, and when the danger was imminent, franticly sending and calling reinforcements from other positions.
The problem with this method is that you have to game with fairly "gentleman" players, because when someone wants to bend the test to his "needs" there is nothing to stop him. But I suppose solo players should be glad to use it.
Thanks for your comments; I think you have a good system there, although some might consider it to be a bit dice heavy. You can't please everyone, though.
We do seem to come back to this idea of some sort of reaction test for commanders a lot of the time, and there do seem to be some systems for implementing them.
I wonder how much we do use them though.
I think using such methods (at least our version) requires some imagination (what factors may be in place) and goodwill (so you are willing to accept someone else point of view and not go into pointless argument). Experience also helps a lot - it creates necessary trust between players and makes easier putting up ad hoc alternatives.
Concerning heavy dicing - it depends on one's liking, as always, but I suppose many people fall in the trap of thinking that more dice throwing makes a game more random. It is not so, because it depends on probability. Also, more throws usually means that extreme results usually cancel out (I get one here and you get one over there; if one side gets all, well, it is obvious they were blessed this time :) ). So more dice throwing may likely mean a game is less random (which I like).
I have seen very nice (playable and seemingly realistic) rulesets with low amount of dice throwing. But I felt really down when one or two bad dice rolls decided the outcome of a whole battle. For some reason it is easier for me to accept defeat when my opponent had to score good rolls 30 times to 20 of mine than just 3:2. I think it is because I feel that "my troops tried hard" (they scored many local successes, however small).
Yes, agreed that many die rolls tend to blur or smear out the luck factor. I've just finished Philip Sabin's Simulating War and he says the same thing.
I do wonder, however, if there is a critical point in many battles (both wargame and real) where there is a critical point. Something has to swing one way of the other.
Or is it an accumulation of small points which suddenly go critical?
I would say that in most cases there is no single critical point. Some may argue that for example death of an overall commander is single and critical, but I suppose it plays together with other factors as "a commander too many" or in special cases where the confidence of an army lays on some single person ("the one and only Mahdi" or similar).
So, IMVHO I think there is accumulation of small points and critical points are not "one second" instances, but rather the summit of a wave, which are colorfully depicted in literature, because it makes it better to read.
Concerning ancient battles, many descriptions seem to point out no more than several critical points. Yet, the battles took many hours, seemingly after these critical points happened as well. So, at the minimum, the critical moment was not so obvious to many of those taking part in the action.
I think I tend to agree.
The accumulation of small results (individual combats in real battles, die rolls in wargames) tends to constrict the next possible moves to a more limited manifold than could have been the case if the result had been different.
So the narrative of the battle evolves at each minor point, rather than suddenly changes at one crucial interaction.
Anyone got any good counter examples?