Non random wargames

[Jerboa]

Aurelian uses random card drawing: you draw a number of cards after suffling from a discard pile.

Source: Video N3

http://dnir.net/JerboaNet/Jindex.htm

[Jerboa]

 

Man on Man (1996) R. Bracey

These are ancient miniature rules where combat is decided by deterministic charts.
I’ve tried to contact the author that signs ksi45, from uk, without success.
Interesting rules, still do not know if they can be different from other deterministic rules listed.

http://dnir.net/JerboaNet/Jindex.htm

[Phil Dutré]

Isn’t this whole debate why Game Theory was invented – i.e. to deal with uncertainty when making a decision, and trying to derive the best odds for what decision to make?

In a wargame, a randomizer (dice, cards, whatever, …) is not necessarily there to add randomness in the system, but rather to add uncertainty, that is otherwise difficult to model in a 1 player vs 1 player setup with no hidden information.

Whether you like your games having more or less or no randomizers is a matter of taste, and has nothing to do with the wargame or its portrayal of a military battle as such.

 

[Phil Dutré]

Jerboa wrote:

Plus, there are scientific studies that prove that human anticipation for the results of a random draw are above the random average: those studies were done with card drawing trying to guess what color, number ,etc.

Sure, those studies exist. But those studies are also flawed.

There is no such thing as being able to predict a pure random event without having some sort of information advantage.

[Phil Dutré]

Another point:

A game might contain random elements – as seen from a player’s perspective – without randomizers being present.

E.g. if you would use a paper-scissors-stone mechanic (many games have variations of this mechanic), I have to pick one of three options. However, I have no clue what option my opponent might pick. I could guess, or try to out-think him, or use my intuition. But in essence, from my point of view, to determine my optimal move, I have to assume my opponent is making a random choice. He might pick deterministically, using a procedure unknown to me; but to me, that’s as good as random. Game theory was developed to make decisions based on such uncertainty in knowledge. At it’s most basic level, it boils down to a decision table. See also the seminal “The Compleat Strategyst” (http://www.rand.org/content/dam/rand/pubs/commercial_books/2007/RAND_CB113-1.pdf).

If each player would secretly assign different strength to various units (e.g. Stratego, or as I understand it, AWE 10), that introduces a random element in the game from the point of view of the other player.  The secret information might reveal itself during the game, hence reducing the amount of hidden information (or perceived randomness), but that doesn’t mean initially there is no random element present. Actually, an optimal tactic might be to assign the strengths of your units randomly yourself, to prevent the other player from outsmarting you. That’s a common strategy in game theory, preventing a bias in your decisions that could be exploited by the opposition.

Anyway, the point I’m trying to make is that the division between random and deterministic elements is a grey zone. Hidden information is essentially a random element in the game if I have no or limited means to acquire that information, and it is at least random for as long that hidden information has not been exposed.

That doesn’t mean that all game mechanics are perceived as equal, or are as much fun. Some mechanics (randomizers, hidden info, …) might work better than others in certain game engines. And often, randomness is perceived differently when you change the mechanic. E.g. the simple exercise of letting a player draw any random card (“I am in control!”) vs drawing the top card of a stack (“This card is forced on me!”), can make a psychological difference. But that’s a different discussion …

[Jerboa]

Phil Dutré wrote:

Another point: A game might contain random elements … without randomizers being present. 

Good point.
I’ve been using randomness as a synonym for non-randomized. To be more accurate I should have wrote games without randomizers instead of non random wargames.
In order to understand a multi-variable complex event, there are objective methods of analysis besides pure statistics: simplified models and limit situations. J. Williams in the The Complete Strategist seems to address those, from the introduction.
Chess is a good example for a deterministic game. Backgammon is a good example of a partially stochastic game. The big issue here is how we do classify Stratego, again a simple example of what AWE 10 would partially be.
In fact when I consider that luck may be involved in a chess game, as in any real life event, that does not automatically allow me to classify chess as partially random. The fact that a chess player cannot predict the full option tree from a given position, implies that his choice will be influenced by intuition, experience, plus even knowledge of his opponent play style (intuition and decision).

There is another key example that must be introduced into this discussion: pocker card game. This game has a random variable, that is after-shuffling card drawing, but then there is the bluff and betting sequence. This dynamic, not deterministic sequence, is not the same a random variable, namely because of the nature of the decisions taken. I call this intuition/decision.
Therefore I think that dismissing intuition from game dynamics is not reasonably possible.
But to get closer to understand your stance I will consider that both a randomizer and intuition/decision are stochastic variables of about the same value.

To keep it on target, it would be interesting to get a clear – no grey areas – classification of Stratego.
1. Characteristics: no random variables, in the sense that there are no randomizers.
2. Setup decision of hidden value elements: the decision for a completely random distribution of strength is a valid strategy as any other.
3. Intuition/decision drives the game.
4. Progressive revelation of forces involved, often culminating in a deterministic end.

Even we if consider that 1-3 are all just random, point 4 indicates a game progressing towards non-stochastic decisions. In the end, the stochastic degree of Stratego cannot even remotely be compared to backgammon.

The first tentative classification was Step Decision Deterministic.
Or Progressive Near-deterministic, or, if you want, Step Decision Stochastic (no problem).
My point is that this ‘SDC’ system has little to do with partially stochastic with randomizers, the backgammon example, that is so close to traditional wargames.
But more than that, there must be a definite classification, something I cannot get at the moment. Even if Stratego is finally considered as partially stochastic, that will be useless in terms of useful game description, i.e. a sensible game classification.
 

http://dnir.net/JerboaNet/Jindex.htm

[Jerboa]

Not sure if everyone knows how Swords & Spells worked, but FWIW, It was by applying the average of D&D hit dice of attack weapons and beings represented. A stand of 10 men wearing mail and shield and carrying longswords has hit dice or defensive value of: 3.5 (average of a d6) x 10 = 35 and attack value of 45 for long swords which have an attack dice of a d8. Penalties to attack values are added as per the D&D rules, so if they were attacked by a group/stand of 10 orcs with spears (d6), for example, I believe the penalty is -2 for spear versus mail and shield, the orcs would inflict 35-20=15 points against the humans. You kept track of the damage per stand and there were tables to use to figure out what the attack value of 4 men left on a stand might be. if anyone has questions I can dig out the rules and look it up.

Excellent summary.
But my doubt is about a truly random element being present, on artillery and morale rules. The first may be negligible, but if the second does influence the game outcome that’s crucial.

 

http://dnir.net/JerboaNet/Jindex.htm

[Author]

Posts