Command Ops: Artillery under AI control and you

[wodin]

I don't go as far as you only above battalion. However maybe I should do. I'd have thought mortar management etc being abit too much micromanagement which is why I always left Battalion support assets attached..with the long range arty units you only have to move them once or twice during the scenario (some times they don't need to be moved at all)

ORIGINAL: RockinHarry

ORIGINAL: wodin

I always at the start of a scenario separate the arty units and give them then move them to a designated firebase with a defend order..or just defend in situ order..moving them up when they become out of range. It's one of the first thing I do.

Excatly! That´s what I do too. The combat Bn´s and their HQ have a way better time organizing stuff, without all that subordinated support elements. I do the same for AI side and it´s elightening how much better the AI plays with mortars and Arty pulled from the combat Bn´s and have them concentrated under some dedicated Rgt or Div HQ.

[Phoenix100]

If I followed at all what Bletch-G said to start off this thread, shouldn't we be placing the arty wherever we like (in range of targets) and then attaching it (via Cntrl-click, say) to an HQ higher than the assaulting HQ it is to support (one higher would be ideal), then giving that HQ a defend in situ order or some such? In order to ensure that the AI calls on the arty to support the assault?

I thought the idea was - following some recent posts - that we couldn't even trust the Bn mortar asset to put in a prep bombard to assist an assault, that it would be better to attach even this to the higher HQ.

[RockinHarry]

ORIGINAL: Arjuna

Harry,

Sorry for the delay in replying but I did miss this post. I have installed IE 10 and since then I notice that some times it automatically marks things as read even though I haven't actually read them. Either that or I have opened this, got distracted and its marked it as read. Anyway my apologies.

ORIGINAL: RockinHarry

So here´s some questions:

What actually makes a particular unit call for "Arty support"? Does it actually need to be attacked by nearby enemy ground troops, or is it rather the TLOS rating of enemies that might yet not be immediately "threatening"?

Each minute a unit will check for active enemy - ie those that are assaulting, reorging, firing or bombarding - and inactive enemy - ie those that are not assaulting, reorging, firing or bombaring. It first checks its visible threats and then it checks its sides known enemy concentrations. If its static it searches out 2000m. If its moving it checks ahead along its route.

How much influence on that has the "Default Enemy APer FP" and "Default Enemy AArm FP" as set in the ESTAB for Axis and Allied sides? Do these influence just intel, or also "threat" perception and thus AI engagement behavior?

In deciding to respond the AI determines if the enemy is active - ie assaulting, reorging, firing or bombarding - or inactive. The probability of being added to the eligible threats list is based on their active status, with active enemy having a greater prob, and the total targets within the enemy - ie pers, vehicles and guns.

How much influence has the
"% of enemy to kill" and "Total kill points" settings in Victory Conditions, on AI engagement and Arty usage?

None at the moment but it's a point to consider. Thanks for raising it.

With regard to german and allied doctrines, there surely can be made some more in depth debates. I see a number of errant assumptions above...

Debate away.[:)]

Ah thanks! That explains most my observations with regard to on call support behaviors. So there´s a bias towards "activity" and less towards actual "threat". Could probably explain why there´s little prep bombardments to be used on KNOWN enemy units, which are set to "ambush", since they stay longer in sort of "inactive" mode. Also counts for defenders which for unknown reasons do not engage closing attacking enemies and thus stay longer "inactive", although these are the greatest "threat", sitting directly at the path of enemy attack.

So far I thought that Arty is mainly called on enemies that can be hit and hurt easily (moving in the open, yet not an actual threat to any friendlies), so that enemies which are near friendlies (and thus provide a more real threat) receive less Arty attention, when already deployed in good defensive terrain or dug in. That oftenly leads to situations, that enemies that actually need prefered Arty attention (at a Vic Loc, or any assault Obj), will be first rushed with ground units and once this assault does either succeed (enemy pushed from defense position) or not, THEN Arty would be called. That´s a costly affair for both sides. If the ground attack does not succeed, then it´s actually little late for calling Arty support. If it succeeds, then the attackers do it at a higher cost than necessary. If it succeeds too easily (defender retreats without really much pressure), then the defender receives a terrible Arty punishemnt, while retreating out of the dug in positions.

My idea about "% of enemy to kill" and "Total kill points" settings was that it could likely influence Arty usage to achieve more overall kill points (more firing at targets of opportunity). That´s obviously not the case. OTOH setting these to Nil would not necessarily get Arty to focus more on achieving victory locations (more prep bombardments, focus on actual assault locations, less firing at targets of opportunity).

[RockinHarry]

ORIGINAL: wodin

I don't go as far as you only above battalion. However maybe I should do. I'd have thought mortar management etc being abit too much micromanagement which is why I always left Battalion support assets attached..with the long range arty units you only have to move them once or twice during the scenario (some times they don't need to be moved at all)

ORIGINAL: RockinHarry

ORIGINAL: wodin

I always at the start of a scenario separate the arty units and give them then move them to a designated firebase with a defend order..or just defend in situ order..moving them up when they become out of range. It's one of the first thing I do.

Excatly! That´s what I do too. The combat Bn´s and their HQ have a way better time organizing stuff, without all that subordinated support elements. I do the same for AI side and it´s elightening how much better the AI plays with mortars and Arty pulled from the combat Bn´s and have them concentrated under some dedicated Rgt or Div HQ.

The main point is, that many these Bn support assets in the ESTAB reflect some particular unit compostions at a given time during BoB and MG operations. These aren´t alwas standard and in RL, things are usually kept concentrated for the point of main effort. Even mortar units were concentrated under unified command to enable really effective barrages. 3-4 mortar units conentrated for few minutes at highest ROF do way more, than entrusting the Bn AI to use these at the right place and right time, doing 10 Min of rather harassing fires.

Beside that, the AI handles movements, formations and assaults way better if not freighted with support or line support units from my experiences.

[RockinHarry]

Here´s an interesting report of the commander of german Arty Bn/Abteilung, employed at the eastern front:

Google translated to english

Original german

[wodin]

Harry I'm aware of the issues of keeping support units attached at Battalion level as I've mentioned it a few times. However I suppose one I'm to lazy to micromanage them, two I try and put faith in the game and when it seems odd blame the commander for being stupid infact thats on the whole I play the game..more in a roleplay type of way than get the perfect tactical implementation. I do detach all Arty above battalion level as I said, however I give very few bombard orders, unless I see a juicy target or someone is under threat. Again I try and roleplay as much as possible and am willing to except silly things happening (as long as it isn't game breaking)at times. Arty management has never really been something I've manually managed, 90% or more of the time I just leave it upto the AI when and where to fire.

Obviously there are things that have bothered me like routing behaviour (which is now sorted, hurrah)and the other major issues that have been brought up here. On the other hand there are things I miss or am unaware of or willing to as I said blame bad commander leadership. The quick death of the mortar units and support assets does bother me though and has done for years i suppose. I'd rather there be a modifier to make them abit harder to kill than other units like an Inf Coy, again to make up for the fact there is no micro terrain in game so the types of places these units would be set up aren't represented on the map which leads to getting spotted and killed to easy. As you know from our chats over the other game..the more abstract the terrain the more niggles you'll get if you have a hi fi combat\spotting mechanics

[BletchleyGeek]

ORIGINAL: dazkaz15

Looks like adding each battery size unit adds between 100-500m to the Frontage or depth of the attack formation.

I now have just noticed a number 4 problem though.

4, When adding artillery with motorised movement tables (that will be most of them) to your assault, you can now only place the assault marker on terrain that motorised units can reach.
Not very handy for taking out isolated outposts in the middle of a forest, but ok for the majority of objectives I guess.

Thank you for your tests, Daz. I am not sure the frontage of a force should be increasing as one adds non-line units to it, but Dave might have a good reason for doing so in the formation code.

Regarding 4), that shouldn't happen unless there isn't a good firebase location, reachable by motorized movement and within range of the target location. I see some simplifying assumptions here which are getting in the way of things.

[jimcarravall]

ORIGINAL: RockinHarry

Here´s an interesting report of the commander of german Arty Bn/Abteilung, employed at the eastern front:

. . .

The report appears to discuss nearly a month of division level operations, and if I interpret the word "shot" correctly in the artillery tabulation, it resulted in expenditure of just over 2100 artillery rounds by 34 batteries using both direct and indirect fire on defined targets while the division was in a defensive role.

Don't know how to interpret the report of 13,278 "shot" in what appears to be a footnote to the artillery data.

If that is total artillery ammunition expenditure, my guess is the difference is between fire targeted on known objectives for specific tactical needs (as defined in the tabulation -- a circumstance which exists in BftB), and harassing fire sent into an area without specific targeting or anticipated results -- a circumstance that doesn't exist in BftB (or you'd have an easier time accounting for the duration of your Operation Veritable efforts [;)]).

If one exists, it would be interesting to see a similar report (ideally for the same division by the same author) for a period of operations in an offensive role.

Would help validate / invalidate ammunition expenditure rates for the AI in the game.

Player-controlled rates would be different, since the player operating the command can define the duration and expenditure for any planned bombardment.

[BletchleyGeek]

ORIGINAL: phoenix

If I followed at all what Bletch-G said to start off this thread, shouldn't we be placing the arty wherever we like (in range of targets) and then attaching it (via Cntrl-click, say) to an HQ higher than the assaulting HQ it is to support (one higher would be ideal), then giving that HQ a defend in situ order or some such? In order to ensure that the AI calls on the arty to support the assault?

I thought the idea was - following some recent posts - that we couldn't even trust the Bn mortar asset to put in a prep bombard to assist an assault, that it would be better to attach even this to the higher HQ.

There's a finer point in this discussion, regarding what it means to "assist an assault". Basically, it might be two things. One, to execute a preparatory bombardment of the target area. Two, to provide on-call support as the attack (or more generally, any task) execution progresses. One isn't currently supported by the AI, you need to do that by hand. Two works fine, yet it can be further refined.

This split between preparatory barrages and on-call support, has sprung up a conversation between me and Harry, about how we can implement reasonably historical artillery doctrines with the current AI and user interface we have. I think we can do quite well, yet it does indeed convey a certain amount of micro-management.

I'm not too worried about micro-management against the AI, in multi-player games is more worrisome. Any observations you have on using artillery on a MP game are welcome, phoenix

[dazkaz15]

ORIGINAL: Bletchley_Geek

ORIGINAL: dazkaz15

Looks like adding each battery size unit adds between 100-500m to the Frontage or depth of the attack formation.

I now have just noticed a number 4 problem though.

4, When adding artillery with motorised movement tables (that will be most of them) to your assault, you can now only place the assault marker on terrain that motorised units can reach.
Not very handy for taking out isolated outposts in the middle of a forest, but ok for the majority of objectives I guess.

Thank you for your tests, Daz. I am not sure the frontage of a force should be increasing as one adds non-line units to it, but Dave might have a good reason for doing so in the formation code.

Regarding 4), that shouldn't happen unless there isn't a good firebase location, reachable by motorized movement and within range of the target location. I see some simplifying assumptions here which are getting in the way of things.

It should be me thanking you for your opening post. It was very informative, and I very much liked the concept as it was not something I had tried yet
.
In fact it worked great as regard to providing on call support, and helped out a lot with not having to micro manage the arty for the assault.
That's why I would love to see the points I have bought up discussed, and fixed or find ways we can work around the problems
.
The whole concept of adding the arty assets to the battalion for the assault not only adds to the immersion, but as mentioned takes some of the micro out of using arty for your important assaults.

[BletchleyGeek]

ORIGINAL: dazkaz15
It should be me thanking you for your opening post. It was very informative, and I very much liked the concept as it was not something I had tried yet
.
In fact it worked great as regard to providing on call support, and helped out a lot with not having to micro manage the arty for the assault.
That's why I would love to see the points I have bought up discussed, and fixed or find ways we can work around the problems
.
The whole concept of adding the arty assets to the battalion for the assault not only adds to the immersion, but as mentioned takes some of the micro out of using arty for your important assaults.

No worries, Daz.

I think we should try to have more threads like this, tackling concrete problems with concrete examples, and providing saved games so others can comment on what one is doing and also "get into your boots". Videos are great, since they're quite clear when it comes to describe what's going on, but they aren't interactive and that somewhat deprives from learning things efficiently. One needs to try stuff by himself, and get feedback on it [:)]

[Arjuna]

dazkaz15,

I just checked my code and it already exracts bases and any long range arty from the frontage calc. Do you have a save I can test?

IGNORE I can reproduce it.

[Arjuna]

I had put that Base and Arty culling code inside the UpdateDrawFrontageDepth() but forgot to add it to the CalcInitialFrontageAndDepth() that is called when the Auto box unchecked. My bad.

[Arjuna]

Fixed. In this screen shot I have given orders to the 35th Tk Bn with the 22nd Arty bn attached. The frontage is 700m the same as it without the arty attached. Please note that the code culls the forceGroup of any bases and long range arty ( ie > 5km range ). I have not included short range arty - mortars or IGs - as these can often be included in the assault, especially if the assault is a long one.

[wodin]

Bletchley, well I for one am concerned about micromanagement against the AI..if it can be cut back all the better.

ORIGINAL: Bletchley_Geek

ORIGINAL: phoenix

If I followed at all what Bletch-G said to start off this thread, shouldn't we be placing the arty wherever we like (in range of targets) and then attaching it (via Cntrl-click, say) to an HQ higher than the assaulting HQ it is to support (one higher would be ideal), then giving that HQ a defend in situ order or some such? In order to ensure that the AI calls on the arty to support the assault?

I thought the idea was - following some recent posts - that we couldn't even trust the Bn mortar asset to put in a prep bombard to assist an assault, that it would be better to attach even this to the higher HQ.

There's a finer point in this discussion, regarding what it means to "assist an assault". Basically, it might be two things. One, to execute a preparatory bombardment of the target area. Two, to provide on-call support as the attack (or more generally, any task) execution progresses. One isn't currently supported by the AI, you need to do that by hand. Two works fine, yet it can be further refined.

This split between preparatory barrages and on-call support, has sprung up a conversation between me and Harry, about how we can implement reasonably historical artillery doctrines with the current AI and user interface we have. I think we can do quite well, yet it does indeed convey a certain amount of micro-management.

I'm not too worried about micro-management against the AI, in multi-player games is more worrisome. Any observations you have on using artillery on a MP game are welcome, phoenix

[BletchleyGeek]

ORIGINAL: wodin
Bletchley, well I for one am concerned about micromanagement against the AI..if it can be cut back all the better.

Don't take my statement as an absolute, wodin [:)] In MP micro-management in a real-time (even if pausable) game can break the game. And any reduction done on micro, which becomes apparent during MP, will entail a reduction in micro playing against the AI as well.

[RockinHarry]

ORIGINAL: Arjuna

Fixed. In this screen shot I have given orders to the 35th Tk Bn with the 22nd Arty bn attached. The frontage is 700m the same as it without the arty attached. Please note that the code culls the forceGroup of any bases and long range arty ( ie > 5km range ). I have not included short range arty - mortars or IGs - as these can often be included in the assault, especially if the assault is a long one.

Maybe it would help to keep the longer range mortars (4.2" and 120mm), as well as IG bit more behind the HQ and reserve location. It´s sufficient to have these support more from near their maximum ranges to support any attack or defenses. Something like max minus 1km would do the purpose (attack), so these could also reach the area behind (or before) the enemy position, to deal with counterattack forces or some in assembly positions. Defense setup could then more be like max range minus 2km something. Shorter employment ranges are only usefull if one intends to support friendlies in the neighborhood, but that would be the case then where I rather subordinate these guns to a Rgt HQ instead and do micro managing.

[Arjuna]

Harry,

You have to factor in cases where the user specifies an FUP that maybe 3 or 4 kms from the objective. In such, you have to push the mortars and IGs forward of the FUP so they can reach the objective. The way we handle that is to have them join the assault and they reassess as they assault forward. Because they are support units they are placed behind the line units in the assault. When they get to a point where they can bring eff fire onto the objective plus a margin of range to allow them to hit enemy reinforcing the objective, they then deploy and go onCall.

Now the code that sets the default frontage when you uncheck the Auto box, this is called before any planning in undertaken. It's just an estimate at this stage. So it has to cater for the worst possible case. If we don't factor their persQty into the equation then the force may be too bunched up and therefore more vulnerable to enemy fire during the assault.

[dazkaz15]

Awesome

Thanks Dave [:)]

[wodin]

This is why though when you check out unit casualties at the end of a game..esp if you don't micro manage these support units 90% or more are Mortar units\support units for Btn. Due to the small amount of men in the unit compared to a inf coy they soon get depleted and also in real life Mortars would on the whole be set up out of LOS as much as possible..but this sort of terrain isn't in the game (think courtyards etc) so they end up in LOS way to easy and then wiped out.

This is the last remaining bug bear for me in the game. I can live with it as I'm use dot it.

ORIGINAL: Arjuna

Harry,

You have to factor in cases where the user specifies an FUP that maybe 3 or 4 kms from the objective. In such, you have to push the mortars and IGs forward of the FUP so they can reach the objective. The way we handle that is to have them join the assault and they reassess as they assault forward. Because they are support units they are placed behind the line units in the assault. When they get to a point where they can bring eff fire onto the objective plus a margin of range to allow them to hit enemy reinforcing the objective, they then deploy and go onCall.

Now the code that sets the default frontage when you uncheck the Auto box, this is called before any planning in undertaken. It's just an estimate at this stage. So it has to cater for the worst possible case. If we don't factor their persQty into the equation then the force may be too bunched up and therefore more vulnerable to enemy fire during the assault.