random vs deterministic model!

[JMM]

Bonsoir,

The wargames generally use a random function and a matrix datas to find the losses of each opponent.

In HistWar, a mathematical model allows to compute the losses.

What do you think about this?

JMM

[Le Tondu]

Thank you for asking JMM.

Does it have to be one or the other?  I feel uncomfortable when one has to choose between two polar opposites.  I think that somewhere in the middle is probably better.

Why not some kind of a mixture of both?  

It seems to me that much existed in the realm of chance.

Why not have a mathematical formula where certain parts of it are randomly generated?  

[One could go even deeper and randomly decide which parts of the mathematical formula that will be randomly generated.]

This way the results could never be exactly the same two times in a row.  Sort of like how life is.

I hope that this helps some.

[CBR]

My first post here...

I dont know much about this game except drooling over the screenshots    but how much is deterministic? Morale, shooting, melee?

I think I would prefer some random element in the game. That is part of real life and especially war...just so many things going on that you shouldnt be able to know precisely what the result is.

I take it you mean that if 2 equal regiments are shooting at each other they, then the game engine will produce same amount of casualties of say 400 over a certain amount of time, why not +/- 20% or something like that?

I dont know how many different modifiers there are, but is it not too easy for an experienced gamer(or a number cruncher) to know exactly what will bring victory?


CBR

[Frédéric Walther]

Mainly deterministic with an additional pinch of random... because life must not be tasteless !

[Duc_dEarl]

Bonsoir,

The wargames generally use a random function and a matrix datas to find the losses of each opponent.

In HistWar, a mathematical model allows to compute the losses.

What do you think about this?

JMM

Ideally... I would prefer an algorithm that produces a mix of random-mathematical results...

Influenced by things such as: realistic ballistic trajectories, Line of Sight & environmental factors (smoke, fog, rain, trees, walls, fences, defilade, enfilade, etc.), troop expertise, ammo quality, weapon quality..

As well as influenced by the usual factors: target range, ranks firing, ammo supply, weapon type... etc., etc.

Also... I would like to see realistic firing disciplines used for musket volleys.. i.e. feu roulant (rolling fire) by peloton, by rank, by file, by division, diagonal fire, fire-at-will, skirmisher fire, reserved carre fire (against Cavalry), etc, etc..

Cordialement, Duc_dEarl

[Malaspina]

Sorry for resurrecting this very old thread.
However, I haven't been able to find an answer to my doubts. I've been recently doing some quick "tests" piching regiments of differing quality one against the other. Elite troops, even if smaller in number, always seem to have the better of lower quality regiments (all other conditions being equal).

My question is: does the game use random or deterministic calculations for engagements?

I.e. if I engage a conscript regiment with my French Imperial Guard, am I 100% sure that I will always win, or is there a remote chance that, due to an unlucky dice throw, my Guards will break?

I remember conducting similar tests with CMBB, where I had a single panzerschreck pitched against a T34. 90% of the time the schreck missed the first shot and soon routed under the tanks withering MG fire, but one or two times he did get a lucky first shot which killed the T34 and won the battle.

[quartermaster]

Is there much of a difference between

•     a spread of results with the variation delivered through a randomising function such as a dice
•     a mathematical model with statistical variation of the results

       
I would agree there is a difference between total random results and fixed and repeatable mathematical outcomes.

The 6-sided dice is a rather crude tool for fitting to some distribution of results.   

[Gunner24]

I.e. if I engage a conscript regiment with my French Imperial Guard, am I 100% sure that I will always win, or is there a remote chance that, due to an unlucky dice throw, my Guards will break?

I don't know the answer, it's a good question......if these "Regiments" were the same strenght, lets say 2,000 men each, and it was only them on a 1v1 basis with NO other outside influence, then how could the conscript Regiment hope to win ?.  If it was a straight 1v1 and with nothing else in the mix the result should be certain - in this example.

If the "test" was two concripts Regiments in the same conditions, then it's a different situation......it would need to be about 50/50 as long as there is no other influence on the battle between these two Regiments - all assume no ground advantage and both open fire at the same time and engage in melee with the same number of men etc etc.

[LNDavout]

My 5 cent:

a mathematical model for the minimum losses, like in the table of Kriegsspiele, and a random factor for minimum to the max of losses.

The math keeps track of all the known influences, the random factor for the unknown.

[FranzVonG]

Deterministic in macro-factors, random in micro.
Imho, a battle should never change completely due to a bad dice roll: I've played too many of those games.
At the same time, the little unexpected event (ie: the conscript regiment that holds against an elite once every 100 fight) gives to the player that "living" feeling in a game.

As far as I know, HLG is completely deterministic, and I think that's too late to change it now. In a game with this kind of fog of war, there will still be plenty of unexpected events.

[Montecuccoli]

This is an important discussion for me.

I worked many years on a napoleonic tabletop wargame for figurines and i my main thoughts were focused on NOT using dice to resolve combat but just let the players know the mathematics rules behind the game and letting them focus on strategy.

There was a huge amount of work and brainstorming to finally catch the feelings i had in mind and the final result (after 5 years) is a good one (well at least for the friends who played my rules  ).

I do not like random functions on this kind of game for (maybe my bad luck) seeing a good strategy being overwhelmed by bad luck is sooo bad for me.

Yes we all know the Napoleon statement (is it a good general? Is it lucky?) but i believe that a good strategy has nothing to do with luck itself, for every aspect of Friederich Der Grosse dictates and Napoleon ones were about minimize the effects of luck and random on the strategy.

So maths i prefer 

[GrosPaul]

Montecuccoli, you affirm "i believe that a good strategy has nothing to do with luck itself, for every aspect of Friederich Der Grosse dictates and Napoleon ones were about minimize the effects of luck and random on the strategy."

I could admit that in a perfect world and perhaps wargames with figurines could approach that.

But as well, you could also admit in real life, no one can act without beeing influenced by at least a hint of fortuity. For example the random weather conditions, the false gathering of informations, not totally reliable maps at this era, fog of war, mental state of the chiefs corps at the moment, ...

I agree with you maths should be the most important factor but spiced with very little randomized, in order to approach the reality.

You admit yourself "...  about minimize the effects of luck and random on the strategy.". Consequently, I understand you agree implicitly my answer.

Cheers,
GP.

[Montecuccoli]

Yes in that point of view i have to agree with you  

anyway i believe (and we are running on the edge of phylosophy eheheehe  ) real life is a matter of different facts bound together and "luck" has nothing to do with that facts... but i am going to far talking about life 

[GrosPaul]

anyway i believe (and we are running on the edge of phylosophy eheheehe  ) real life is a matter of different facts bound together and "luck" has nothing to do with that facts... but i am going to far talking about life 

I agree completely with you "luck" is a "weird" word. "Random" is more exact; luck is an effect of it . Random rules our all life. See a very elementary example: Before going off your home with your car you hesitate to drink or not a cup of coffee. This so simple decision may have enormous consequences, because arrived at a corner of two street depends you collide or not another vehicule. Maths are here completely out of the subject.

All right, we are talking about life but also in close relation with games. OK, I stop. Else I could tempted to develop up to the mechanisms of the univers ... and then some moderators ....  

Best regards,
GP.

[Cpl Steiner]

As someone once said, there are lies, damn lies, and then there are statistics.

The game CANNOT, in my opinion be completely deterministic because any figures from historical sources for the expected number of casualties caused in a given set of circumstances are merely averages.

For instance, if a historical source says a volley of musket fire at 100 yards range by 1000 men would on average cause 23 casualties (a made up example) this does not mean every single volley of this type will cause 23 casualties.

In statistics, we use "standard deviation" to describe the variation in expected results from the extremes at each end of the "bell curve". I think it is vital that the game take the same approach - i.e. in the example above the average may be 23 casualties but on rare occasions it might be only 16 or 17, or as many as 29 or 30, and all the results between with varying degrees of probability.

Similarly, if some historical source says a battalion will break after taking 7% casualties, there should be a bell curve of results possible such that sometimes a battalion might break after only 5 or 6% casualties whilst sometimes a battalion might take 8 or 9% casualties before breaking.

The spread of the bell curve in each situation (its standard deviation) would have to be determined through historical research, lots of testing, and also a bit of common sense.