AI Artillery Fire

[Der Alte Fritz]

I am writing a scenario which features some Soviet Spotter with LOS to some German trenchs at start of scenario. All went well and they used to bombard the trenches quite happily.

But having made some changes to the scenario, now all the guns fire on one flag about 2000m behind the front line which blows up some trees and not much else. Have tried parking lorries beside the trenches and other tricks but I cannot seem to get them to fire on the trenches I want.

Any ideas?

[RockinHarry]

hm....not having any idea which way you setup the scenario and which changes you made since, I assume the random detection routines make the german trenches not visible, when you want it to to the russian AI, or the strategic AI thinks it makes more sense to bombard the german held (?) flags. The flag ownership routine is pretty random too, as you know. All random factors involved I think.

[Tero]

Have you assigned preplanned bombardment ?

[Der Alte Fritz]

No - these are the Ais spotters. I was just wondering if there were any tricks to make the ai fire where you want him to.

[Tero]

AFAIK only preplanned bombing missions are the only solution.

[Joachim]

IIRC the AI targets not just flags but spotted units for a prep barrage.

Are the lorries in the open, in good view of AI units? This should do. If not, try valuable units like tanks.

[Bannon DC]

AFAIK only preplanned bombing missions are the only solution.

Tero -- he is talking about controlling the AI's spotters. Designers can not give the AI orders. You have to have a bit of luck to get the finicky AI to work the way you want it to.

[Bannon DC]

Fritz --

Been there... you have one part of it working correctly and then some minor thing changes and it all goes to hell.

I was going to suggest a HT... a more valuable target.

Basically the AI has to see something to shoot.

Try bringing the spotters in on turn one. Place some TRPs where you want the AI to fire. (If this is an player vs the AI only scenario, this will be no problem).

You can sprinkle TRPs as you necessary. The time-to-fire for the AI will be minimized to 1 minute for most artillery. This greatly increases the chance the AI will actually use its heavier artillery.

Bringing in the AI spotters on turn one prevents them from throwing up useless smoke screens as preplanned missions as well and stops the random flag targeting.

Good luck

[RockinHarry]

Using TRP´s looks like a workable solution at first sight, but most my experiences were and are still disappointing, mostly due to the random factors involved and the AI moving around the spotters when it should not. When testing the AI during gameplay, I "surrender" the game at opportunities to check the map and whereabouts of the enemy units particularly the AI spotters. Most of the time, these are on the move and start targeting only when as part of the first infantry wave are already very close to human players positions. Needless to say, many spotters get lost, before the AI gives any opportunities to start reasonable actions.

CMSF "baked" scenario format would give great opportunities to CM1 to solve some major problems. Although it does not work in CM1 for the AI player, one can partly use some "tricks" for H2H/PBEM games.

Setup the player which you intend to have the artillery spotters target particular areas at given times and place these very close (few meters) to the friendly map edge. Start the game and in the first orders phase let the spotters target desired areas in the enemy main battle zone (or anywhere else). Use the timed targeting method (Q key) when appropiate. Then give the spotters movement orders off the map. Since they are placed very close to the map edge, they´ll move off during the following action phase. NOW it would be great if one could save the game after resolution of action phase #1 AND back in the editor save the scenario with "preplanned" artillery strikes to be used by an AI player. Well..it´s CM1 and not CMSF, but some sort of preplanned arty strikes can be forced to one human player in H2H games this way. Off course the scenario needs to be delivered in "saved game" format and there´s the problem with whether the players are willing to play double blind and whether they like to play a game with units they can not move in any setup phase, since it´s gone this way. Maybe BFC comes back and includes a "switch" that enables to have a human played game to be played by the AI midways. Can´t be THAT hard to code (I think...)

Other creative ways to use any "preplanned" (by scenario designer) artillery strikes in a H2H game would be to have some devastation created at the start of the game, causing random suppression and damage at particular units, having craters at desired areas, smoke and dust (CMAK), as well as random fires at bombarded places. One can even simulate friendly arty strikes, which are out of cpntrol of the human player (since the spotters moved off in turn 1, the player owing the strikes wouldn´t even know).

Off course one can place fires, craters, suppressed units ect. already in the scenario editor, but the more awkward way mentioned, deliveres some "surprises", as well as more realistic visual results.

With this method it´s also possible to have destroyed vehicles on map at the games start (orders phase 2 in H2H games). Just setup strong mobile AT units at the map edge, let them create some devastation among enemy vehicles and then let the AT units move off the map edge. Save the game and in orders phase 2 and with some luck and repeated attempts before, you have the desired results. This is all very rough stuff and most likely not usable for most scenario designers and little liked by the majority of players, but it´s interesting to try and to experiment with.

[John Kettler]

RockinHarry,

I've never done a scenario myself, but it might help if you figure out where you want your Russian FOs to be, put them there, perhaps foxholed for survivability, then padlock them in place. This should stop the awkward AI deployment problem from screwing up your planned LOS to the targeted entrenchments. In conjunction with TRPs and preplanned fires, this should, in theory, solve your problem. If you're trying to simulate prep fires, suggest you look carefully at map orientation so that sheaves come down at least parallel to the entrenchments and use "Target Wide" rather than "Target" which is for point target engagement. Sadly, the game lacks the parallel sheaf which is the commonest artillery fire pattern, being the easiest to implement.

Regards,

John Kettler

[Joachim]

John, you are wrong.

Below a link and the relevant excerpt.

Maybe the whole post is relevant to Fritz.

Gruß

Joachim

CM Artillery Synopsis [Treeburst155]

http://www.battlefront.com/community/showthread.php?p=35246&highlight=artillery+parallel+edge+board#post35246

Impact Patterns for Guns

You have two choices here, "target wide" and "target". The first gives you a circular impact pattern. I like to make sure friendlies aren't within 120 meters of the target point when using "target wide". Most rounds will fall inside a circle of this radius; but even at 120 meters distance, friendlies are at risk.

With the "target" command you will get an elliptical impact zone, the deadly area being about 160 meters by 80 meters centered on the target point. The orientation of the long dimension of the ellipse can be north/south or east/west, depending on how the friendly map edge parameters are set. When the north map edge is friendly to one side, and the south edge friendly to the other, the long dimension of the impact pattern will be north/south. If one side has the west edge friendly, the other the east, the long dimension of the arty pattern will fall east/west. In cases of ambiguity, the program will cause the long dimension of the impact pattern to run east/west. This means that in most cases arty will fall parallel to the axis of advance, rather than along the front lines; but a tricky designer can easily make it so the impact pattern is the opposite of what is expected.

[John Kettler]

Joachim and RockinHarry,

The reason you get an elliptical impact zone is that it is a rough approximation of the 50% or even 100% zone you would find for a single gun firing many rounds, or a group of guns firing fewer to multitudes of rounds. In either case, the statistically varying dispersion in range produces the long axis of the pattern and is greatly influenced by such things as the temperature history of the propellant charge used, the heat of the gun's interior at the time of firing, headwinds and tailwinds, etc. Sources of errors in also statistically varying deflection (the short axis of the pattern) include round to round manufacturing variations, crosswind, bore wear expressed as EFCs (Equivalent Full Charge) fired (creating shell wobble), etc.

Guns, by virtue of their high velocities, produce narrow elliptical burst patterns, but with howitzers, which typically fire at lower muzzle velocities than guns, the pattern broadens somewhat, after which comes the mortar, whose impact pattern, as a result of low velocity and near vertical impact angle, tends strongly toward the circular and can broadly be thought of as a cookie cutter.

Classically, the object of the game is to place the MPI (Mean Point of Impact), of whatever size shoot is being performed, on the center of the target array such that it follows the long axis of that array. Since forces in combat tend to line up opposite one another, the default deployment of artillery has long been to place it parallel to the enemy, sometimes staggered to create greater depth. Subject, then, to the aforementioned errors in range and deflection, the exploding box of steel is then simply dropped on or walked through the target array until the damage criteria have been met, ammo runs out, or the firing batteries are retasked, etc.

One of the problems in CM has long been that the parallel sheaf, per se, despite being the typical pattern fired by WW II artillery, isn't modeled. Instead, we see "target wide," which yields a diverging sheaf and "target" which yields a converging sheaf. Nor is this mere opinion, for I have personally seen the diagrams in the 1985 version of FM 6-20 Field Artillery employment. There, it is clearly stated that a converging sheaf is for point targets, such as destroying a pillbox, that a parallel sheaf is SOP for most engagements and is quickest to fire, since it requires no individualized deflection setting for each gun firing (all guns fire on a common deflection), and an open or diverging sheaf is for covering more frontage at lower lethality, also requiring individualized deflections for each gun in the shoot. The "as seen by the FO" diagrams confirmed the text. The parallel sheaf showed the burst from one shell overlapping to the detonation point of the next, creating a fairly dense pattern. The converging sheaf showed a very dense pattern, but of little width. The open sheaf showed the burst from one shell just meeting that of the next, producing very porous coverage, but covering a much greater width than the parallel sheaf.

Since CM players have no means to control pattern orientation relative to target, they are critically dependent on map edge assignments they have no control over, either. It is no fun at all to be stuck trying to fight a battle with the artillery landing perpendicular to the target's long axis. Not only does it waste most of the ammo, but it suppresses little and completely guts efforts to use smokescreens, too. Been there, done that, and didn't like it. Just such a situation is depicted in Hackett et al. THE THIRD WORLD WAR: The Untold Story. There, the Russian division artillery commander is ordered to support tomorrow's attack by placing BM-27 MRL fire along a certain road. When he tries to explain that to do so, he needs to move back and laterally, he immediately runs headlong into the political officer who smells defeatism and cowardice as the Russians have been battering away for days and taking heavy losses trying to get an exploitable breakthrough. Ballistics and beaten zones go down to defeat, as does the arrested artillery commander and the next day's attack. Why? The BM-27s couldn't be aligned, given where they were made to fire from, so that their impact pattern ran down the road to be targeted. Instead, it simply cut the road very narrowly, leaving most of the defenders unscathed. If we ever get CMC, then perhaps players may actually be able to site their artillery in such a way that it's usable in CMBB battles. Presently, we're entirely at the mercy of forces over which we have no control.

The siting of artillery is a real world concern and in Korea, for example, it became an absolute nightmare when limited flat ground was required by artillery, CPs, ammo and supply points all at once, especially given the issues I described about how artillery must orient on the enemy. I fervently hope BFC models the parallel sheaf and other artillery pattern control issues, such as the ability to designate artillery and mortar concentrations as preregistered boxes, rather than TRPs, in CM2 Normandy. WW II accounts further indicate that we should be able to call for shell mixes, register targets for instant refire on call later, etc. Nor do we have "On my command" fires modeled presently.

Joachim, I think what you've said about the pattern is useful and suggest you also put in guidelines for using rockets in the game, too, thus sparing RockinHarry the kind of painful lessons my troops got!

In closing, I'd be remiss if I didn't mention that I and others have repeatedly mentioned the graphic tutorials of a certain former member as being extremely valuable. I think searching under "artillery patterns" would be helpful.

Regards,

John Kettler

[RockinHarry]

Joachim and RockinHarry,

......

Regards,

John Kettler

John, I´m not actually the person to address in the shell pattern discussion ect. here! Should be DAF instead. If you reread you´ll notice I made some more in depth notes with regard to special setups for playing H2H games in a sort of "faked baked" format! Otherwise,..interesting discussion between you and Joachim.

[John Kettler]

RockinHarry,

You are correct, sir!

Der Alte Fritz,

You are the correct and inadvertently left out second interlocutor. Hope the discussion helped!

Regards,

John Kettler

[Joachim]

John,

the link provided above is a pretty good arty primer by Treeburst155. It is not my work.

I just quoted it as I was too lazy to test myself - and knew this thing was out there somewhere. Gotta bump it a bit, it deserves it.

Gruß

Joachim

[John Kettler]

Joachim,

I quite agree. I merely suggested that in your summary you include the valuable guidelines on using rockets, too. Having had some crash down on my own troops because I got too close, I think it's important. The graphic tutorials of a certain banned British member are also quite useful.

Regards,

John Kettler

[Der Alte Fritz]

Hi John and Joachim

My computer is back at the factory at present after a power supply failure so I cannot test your ideas but I will do so as soon as it comes back ..... hopefully tomorrow

cheers

[John Kettler]

Der Alte Fritz,

Good luck--with both projects!

Regards,

John Kettler