Auteur Sujet: Artillery tests  (Lu 21301 fois)

Hors ligne HarryInk

  • Grognard de la Vieille Garde
  • Major
  • ***
  • Messages: 529
Re : Artillery tests
« Réponse #30 le: 16 décembre 2009, 17:02:08 pm »
Well I'm just crap.  i marched two russian brigades against the horse 4 pdrs (which initially said there were only 4x4pdrs, then deployed at a 6+ gun battery, then (later) registered as 4x4pdr + 2 Hows).  The final regiment took out the guns but only after the other three regiments had variously panicked and wot not.  Even the victors had the hebee geebees and ran away for a bit before rallying.  Perhaps it was a post Friedland brigade with bad memories of the French artillery.

How is it that a gun takes 2 casualties per round.  what?  Does the third rank sh&$kicker dodge?  Goddam his eyes I'll have him flayed!!
"But I vil not divulge any furzer informazion!"

Hors ligne CBR

  • Capitaine
  • **
  • Messages: 107
Re : Re : Re : Artillery tests
« Réponse #31 le: 16 décembre 2009, 17:33:26 pm »
If you dropped long range artillery fire by 80%, the result would be that it would be pretty much totally ineffective.  Halting for artillery preparation would be the height of folly.  Regimental guns would be detrimental, as they wouldn't have enough effect to bother with and they'd slow down the regiment.
Or do what they did historically: move closer and/or use more guns. This is supposed to be a simulation and distances of 1200+ meters are to be considered extreme range. Regimental guns are not meant for long range fire anyway so has little to do with the effect I saw in the tests.

Citer
You can render a cannon hors-de-combat just as easily by killing the crews as destroying the physical cannon, although the only effect we see is destroyed cannons, which I would consider an acceptable abstraction.
Abstracted or not is besides the point. The effect is that guns were finished in way too short a time and with way too little ammo used. It bears little resemblance to what I have read of the battles of 1809 or 1812-13. If LG had been about the Franco-Prussian war of 1870 it would be fine I guess.

Hors ligne Hook

  • Chevalier d'HistWar
  • Modérateurs
  • Général de Brigade
  • ****
  • Messages: 1752
Re : Artillery tests
« Réponse #32 le: 16 décembre 2009, 17:39:27 pm »
Citer
Regimental guns are not meant for long range fire

Long range for a 4 pounder regimental gun isn't very long.  Look where they deploy in the game.

Citer
The effect is that guns were finished in way too short a time and with way too little ammo used.

That's why I mentioned adjusting accuracy.

Hook

Hors ligne CBR

  • Capitaine
  • **
  • Messages: 107
Re : Re : Artillery tests
« Réponse #33 le: 16 décembre 2009, 17:51:18 pm »
Long range for a 4 pounder regimental gun isn't very long.  Look where they deploy in the game.
I have not noticed use of regimental guns (not played much anyway) I take it they deploy at max distance?

Citer
That's why I mentioned adjusting accuracy

Actually I'm not sure what you mean by lowering the accuracy but not their strength? If you lower the accuracy fewer shots will hit which means fewer casualties on average per shot fired.

Hors ligne CBR

  • Capitaine
  • **
  • Messages: 107
Re : Artillery tests
« Réponse #34 le: 16 décembre 2009, 18:05:29 pm »
Erm I just realise some numbers got messed up. All the individual regiments attacking the single battery routed when the battery had fired off 7 rounds of ammo (assuming the ammo counter is correct) so that would be a total of 56 shots fired from the battery. Losses were generally 150+ so thats actually 2.5 to 3.0 casualties per round fired.

The 2 casualties per round fired came from the test where I send in 5 regiments and at one point checked the score and then ammo count. I'm quite sure I remembered it as about 2 per round fired. Maybe more rounds needed when switching targets? or I'm just not remembering it correctly. Overall the battery managed to produce 1000+ casualties and only used 1 canister ammo but I did not get to see the remaining ammo before the battery was destroyed  :|

Hors ligne Ras

  • Chef de Bataillon
  • **
  • Messages: 275
  • MizuRas
Re : Re : Re : Artillery tests
« Réponse #35 le: 16 décembre 2009, 18:16:59 pm »
I have not noticed use of regimental guns (not played much anyway) I take it they deploy at max distance?

There're no regimental guns in the demo.
4lb foot or horse need to deploy closer to the enemy than 8 or 12lbs. I suppose at their max firing distance.

Hors ligne CBR

  • Capitaine
  • **
  • Messages: 107
Re : Artillery tests
« Réponse #36 le: 16 décembre 2009, 18:20:51 pm »
Ah ok thx. Yeah they got shorter range (manual says 1250 meters)

Hors ligne JMM

  • Administrateur
  • Maréchal d'Empire
  • *******
  • Messages: 8375
    • http://www.histwar.com
Re : Artillery tests
« Réponse #37 le: 16 décembre 2009, 18:24:37 pm »
Could you send me your xxx.sav (histwar@histwar.com) (folder save/battle)
I have to understand this figures...

THX
JMM

Hors ligne Hook

  • Chevalier d'HistWar
  • Modérateurs
  • Général de Brigade
  • ****
  • Messages: 1752
Re : Re : Re : Re : Artillery tests
« Réponse #38 le: 16 décembre 2009, 18:27:18 pm »
There're no regimental guns in the demo.

Ah, you're right!  In the beta they deploy to the side and slightly behind the regiment.

Hook

Hors ligne Ras

  • Chef de Bataillon
  • **
  • Messages: 275
  • MizuRas
Re : Re : Re : Re : Re : Artillery tests
« Réponse #39 le: 16 décembre 2009, 18:29:47 pm »
Ah, you're right!  In the beta they deploy to the side and slightly behind the regiment.

Hook
I wonder how long a regiment fights when faced with an enemy regiment with regimental guns?

Hors ligne Hook

  • Chevalier d'HistWar
  • Modérateurs
  • Général de Brigade
  • ****
  • Messages: 1752
Re : Re : Re : Artillery tests
« Réponse #40 le: 16 décembre 2009, 18:42:03 pm »
Actually I'm not sure what you mean by lowering the accuracy but not their strength? If you lower the accuracy fewer shots will hit which means fewer casualties on average per shot fired.

Which is the effect you're looking for, isn't it?  Some shots will hit, some will miss, and the rate at which the target takes casualties (and therefore the morale effect) is reduced.  Occasionally all shots will hit within a short time and you'll get some stronger morale effects.

Intentionally hitting a point target like an individual gun with roundshot at any range beyond a few tens of meters is virtually impossible, but it should happen occasionally.  Mercer was hit by flanking fire from what he claims was 400-500 yards, but the enemy used only shell and roundshot;  no canister fire was mentioned, even for that relatively short range.  He lost no guns, but did lose most of his horses and had only enough men at the end to crew 3 of his 6 guns.  While no guns were destroyed, there was considerable effect to his battery.

It's a little late in the development process to try to model things to that level, but perhaps it could be added in a future version of the game.

Hook

Hors ligne CBR

  • Capitaine
  • **
  • Messages: 107
Re : Artillery tests
« Réponse #41 le: 16 décembre 2009, 18:59:46 pm »
Yeah that is pretty much it. Although it is AFAIK deterministic so there won't be any occasionally good salvoes, just a gradually increasing hit rate.

It was indeed impossible to ensure a hit at longer ranges. As long as one got the accurate range it was all about firing enough stuff at the enemy and some of it was bound to hit something. For counter battery the French did IIRC prefer 6 or 8 pounders because of their higher ROF. Scharnhorst did believe that smaller guns were best in flat terrain where ricochet fire was at its best. In more undulating terrain the 12 pounders were better at direct firing.


CBR

Hors ligne Hook

  • Chevalier d'HistWar
  • Modérateurs
  • Général de Brigade
  • ****
  • Messages: 1752
Re : Artillery tests
« Réponse #42 le: 16 décembre 2009, 20:26:21 pm »
Yeah that is pretty much it. Although it is AFAIK deterministic so there won't be any occasionally good salvoes, just a gradually increasing hit rate.

I know accuracy is already modeled, and you can see balls hitting in front of your troops occasionally, probably other places as well.  If the code is already sophisticated enough, adding a random factor to the "on target" rounds should be simple.  I suspect this is the case.

If accuracy were not modeled then the only thing we could adjust is the strength.  By adjusting accuracy instead, we get a larger range of target effects from a battery firing, from nothing at all to what we currently see, if what I'm seeing in the game is correct. 

The final artillery effect on a target is a sort of emergent behavior resulting from a number of interrelated factors calculated in the game.  We need to figure out which factor is causing any apparent overeffectiveness of the artillery, and make sure whatever gets adjusted doesn't throw several other things off.  Actually, it's JMM who needs to decide about any adjustments, as he's the only one who knows how the game internals work. :)

I've only seen a few cases in many battles where I thought artillery was doing too good a job.  I expected artillery to be a tough opponent.  I also expected it to be a strong ally.  Maybe JMM can record the number of casualties from artillery and other sources and we can compare these numbers against actual battles.  This doesn't have to be displayed in the game, just output to a file that JMM can show us.

Hook

Hors ligne JMM

  • Administrateur
  • Maréchal d'Empire
  • *******
  • Messages: 8375
    • http://www.histwar.com
Re : Artillery tests
« Réponse #43 le: 16 décembre 2009, 23:23:30 pm »
I am going to modify the code.. because I believe there is an issue about the distance.
In fact, the distance is used
a) to unlimber the cannon if necessary.
b) to compute the casualties.

And  there is a fuzzy code on the b) part.
So, I'll modify the parameters to only work with the efficiency distance, not the maximum one with a parameter for the percentage. More clear!

JMM

[EDIT] I am just findng a bug during the flight... Problem with the distances so the cannons use the canister whatever the distance.. and the casualties are multiplied by 3 at least...
If you want, wait for the next demo before checking some new configurations...

THX Christian for your files.. I'll try to report the casualties during a march for one of your file.. (ie the first one : 8 cannons 12£ vs 2055 infantrymen)..
« Modifié: 17 décembre 2009, 01:04:43 am par JMM »

Hors ligne HarryInk

  • Grognard de la Vieille Garde
  • Major
  • ***
  • Messages: 529
Re : Artillery tests
« Réponse #44 le: 17 décembre 2009, 00:50:39 am »
Nice one.
"But I vil not divulge any furzer informazion!"