How to Encourage the Cooperation of Artillery and Cavalry?

[Mr. Doran]

It does not seem like it for regular guns. Guns that are deployed near squares or happen to be overlapping by coincidence seem to have a higher survival rate than they otherwise would if they had not been overlapping; but as I said that just seems down to chance. The only guns that are integrated fully into a square are regimental batteries from what I am aware.

It would be nice to have this feature although I do not know on how many accounts it is historical. It must have happened at some point on at least some occasions I imagine. Wellington is cited as saying that he expected his gunners to follow this behavior but instead they often took flight instead.

[EXELMANS]

This attitude was common among gunners,
Like placing regimental guns in the corners of infantry squares.

---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Cette attitude était courante chez les artilleurs,
Comme de placer les canons régimentaires dans les angles des carrés d'infanterie.

[Mr. Doran]

I think we need some additional doctrinal options regarding AI cavalry behavior and protection of artillery. Even with direct support liaised to guns cavalry still can run off and leave their positional guns to die. I feel the AI in general does not hold the same value a player may to it's positional artillery. This can be mitigated by the constant detachment of artillery of all types but I feel it leads to less artillery being deployed due to it's vulnerability and the lack of AI willingness at times to defend it. The most frustrating case of this being when a cavalry corps has significant amounts of horse artillery at it's disposal and simultaneously fails to utilize it and lets it taken.

[sandman]

BTW, the AI in SoW Waterloo does a pretty good job protecting the batteries. Also self-protection by redeployment when threatened is a common practice of the game AI. 

[Mr. Doran]

I've played ~500 hours of SOW; given mainly on KS but never the less. Even with what HW:N is lacking I could not be paid to touch SOW again. You can just as easily TC and cheese cavalry into enemy guns with ease even at regimental scale cavalry; with squadron scale cavalry it makes it trivially easy. And it's not like its a problem only relegated to the regular RTS camera. It is more than easy enough to do in HITS. It is so easy to do that Kevin from KS eventually changed it so you could not give direct orders to cavalry units that are TC'd to prevent such abusive behavior. He has programmed some fairly realistic forced AI behavior regarding cavalry but even that is not enough to stop a determined player from TC taking your guns.

[JMM]

Hi,

Several functions in the AI are in charge to protect the ART units.
Probably these functions must be tested a lot and must be imprved for a right behaviour.

This behaviour is in the list for checking during the beta test campaign.

JMM

[sandman]

I've played ~500 hours of SOW; given mainly on KS but never the less. Even with what HW:N is lacking I could not be paid to touch SOW again. You can just as easily TC and cheese cavalry into enemy guns with ease even at regimental scale cavalry; with squadron scale cavalry it makes it trivially easy. And it's not like its a problem only relegated to the regular RTS camera. It is more than easy enough to do in HITS. It is so easy to do that Kevin from KS eventually changed it so you could not give direct orders to cavalry units that are TC'd to prevent such abusive behavior. He has programmed some fairly realistic forced AI behavior regarding cavalry but even that is not enough to stop a determined player from TC taking your guns.

First of all, from my side there isnt any effort to make you play SoW, even if I got paid for, since you stated you made a certain decision for yourself. The fact that you didnt consider to try the mod and the setup we´re playing, doesnt change anything about it.

Secondly I did SoW for 634 h, much of the time on MP. But thats not the point. Some people gain more insight in 100 h than others in 1000 h.
After all, I disagree with what you alleged. Of course its basically possible that you loose all of your cannons. But thats simply not the case in nearly all of the MP games I had, no matter if larger numbers of Cav squadrons came around or whatever.
The common command for MP is the division. The division mostly has 1-2 batteries. Assuming the human player keeps up overview, does anticipate enemy movement and is skilled enough to use the AI stances properly, he mostly will be able to keep at least a good amount of the cannons.
Furthermore, as I stated in my previous post, if managing the AI-stances correctly, the AI in SoW does a much better job in terms of self protection than the one of the most recent version of HW could. Anyway, thats not relevant, if the human player cares for deployment, redeployment and emergency withdrawing according the principles I mentioned above. At divisional level the subordinated AI is more like a second level reassurance anyway.
Even at corps level, if some divisional AI acts independently, the batteries have a reasonable higher life expectancy than in this version of HW, where complete batteries often get extinguished within one single Cav raid. Most of us know, that it is absolutely normal, that cannons and gunners get lost, destroyed or killed in the chaos of napoleonic battles. And in SoW, apart from AI self protective redeployment, even if Cav squadrons infiltrate the lines, assuming the tactical basics are practiced decently, normally at least parts of the batteries can be remained.

[Mr. Doran]

Survival rates of artillery under AI hands I think are at least partially responsible ironically in part by the vanilla SOW AI algorithms for cavalry are not as quite as savagely blood thirsty for artillery and not always the most observant or brightest. Where in HW:N they seem hell-bent most of the time at taking artillery whenever there is an opportunity to. The biggest advantage in terms of compassion to survivability that SOW has is that the battery re-limbering operation takes far less time than it does in HW. My complaint stems from not the battery's lack of self-preservation instinct but the fact you cannot grantee mutual support between cavalry. When it does happen it works pretty well. When it doesn't happen you ask yourself why you are being punished. But by the same token, trying to get the AI cavalry algorithms to defend anything in SOW is a total night-mare. KS had some success over the vanilla algorithms but nothing I ever trusted the AI in pulling off on a consistent basis.

As far as TC cheese is concerned we obviously disagree. There are more exploitable behaviors than just trying to TC take someones artillery. The AI's propensity to charge the less than ideal target can be heavily exploited. Charging speed by default is lower than running speed. Once the AI has locked onto a target it tends to stick to it without changing it's target to deal with threats. If it is a cavalry engagement, the unit of cavalry being charged can be given a regular move order at running speed away from its attacker while a second unit is ordered to charge the attacker. With the attacking unit usually responding extremely poorly to the new threat now cutting it to pieces. The same principle works on infantry vs infantry engagements but given it is harder to pull off. Completely massacring battalions of infantry that happen to be on the receiving end of the described is another story though.

[Mr. Doran]

The addition of two behaviors I think would go a long way in the general survivability :

1. Being able to force the AI to keep it's artillery in the reserve line and by the same virtue allowing it to deploy it at it's discretion. An on/off button button for corp commanders for example. A lot of artillery would not be lost as easily as it can be if AI simply did not chose to deploy it in the forward line in highly inappropriate situations.

2. A doctrinal option that forces the corp command AI to limber its artillery if it has no attached support of cavalry if their is a presence of enemy cavalry.

[sandman]

I mentioned already that we always use courier-system. There is no direct immediate control of units.
For exemple, if a direct command is given that includes increased speed or charging, then for the time being, is little chance to change anything as the human player.
So the question is, if we are actually talking about the same basics.

[Mr. Doran]

We are talking about the same basis as in HITS. If you good at positioning your avatar as close to the front where it needs to be to pull off this type of exploitation the order delay is possible to work with. I am not saying its a 100% surefire can-do-without-fail-operation but is is more than possible to do playing in HITS.  It is unfortunately possible for this type of situation to arise even without any human interference just by coincidence of positioning and what the AI Algorithms prioritize and do not prioritize.

[sandman]

Well, still the couriers cannot catchup running or hardly in case of charging units (similar regarding subcommanders).
I rarely lost all or a significant number of my cannons to micromanaging pros (coders of mod and game engine among them).
That kind of micromanaging you are talking about, in my experience is far away from 100 % working and often makes you loose attention for development within the big picture.
I dont know the details about the KS mod and didnt use it, but I have been told that it differs in significant manners from our mod.

[Gunner24]

Back to HW : From my recollections, of several hundred MP games, the main problem with artillery batteries being taken by enemy Cavalry was [in my opinion] very simple -

The battery was often deployed too far ahead of the main line, too far away from any support to help out in time.

[Soldat Louis]

I agree with Gunner, especially when deploying in defense mode, some batteries are send too far in front.

However it seemed that this situation occured in the reality, especially for horse artillery that were deployed attached with a long rope to horses to be removed and limbered quickly in case of excessive threat. 

Coming back to the game. One have to bear in mind mind that what is shown in 3D  might not correspond exactly to the reality of the simulation. I guess that JMM will confirm that it is not possible to represent all the complexity of a napoleonic battlefield. Normaly it was not a whole regiment that was charging but waves of squadrons and most of the time one regiment was not enough to take a large batterie. It seemed that the batteries were often protected by detached cavalry squadrons or infantry compagnies. According to some readings from soldiers of this time, this duty was feared since they were deployed in line next to the guns or right behind while these latter were often under counterbatteries fire. May be JMM could represent some infantry or cav figures next to the guns in 3D WHEN and only WHEN there is a CAV or INF regiment unit nearby. I any case I was always annoyed to watch these poor lonely gunners in the 3D view.

To me it seems that in the game the vulnerability of the batteries should be adjusted with respect to several factors
1) the density, quality and initiative levels of friendly units nearby
2) the moral of nearby units
3) the number of ENY charges
4) the attacking angle of ENY units (more vulnerabity and moral loss if the batterie is threaten from the side or from behind)

Again it is a all a matter adjustements: in the current version the batteries are far too vulnerables to cavalry raids, even within a corps, but it is important not to go too far on the other extreme. Not easy, but I trust JMM to find the solution.

[JMM]

To me it seems that in the game the vulnerability of the batteries should be adjusted with respect to several factors
1) the density, quality and initiative levels of friendly units nearby
2) the moral of nearby units
3) the number of ENY charges
4) the attacking angle of ENY units (more vulnerabity and moral loss if the batterie is threaten from the side or from behind)

Again it is a all a matter adjustements: in the current version the batteries are far too vulnerables to cavalry raids, even within a corps, but it is important not to go too far on the other extreme. Not easy, but I trust JMM to find the solution.

I didn't work on this part since several years... I think I have broken the engine with the last adjustments.
All points listed are in the engine... Probbaly there is a big bug :-(

Before sending the beta test, I have to take a look at this part.

 Soon the next step :-)

JMM