Properly Presenting SOW Waterloo

[mitra76]

For the square there're two distances of checking:

- a long distance at 250 yards where is checked if the nearest enemy is a really dangerous cavalry (sufficient strong to charge me)
- a short distance check at weapon range which checks the cavalry danger in all the quadrants.

The checking distance is reduced if the cavalry came from behind or flanks.

A checking of all the enemy around you can be done, but repeat this every loop for every unit can be heavy, because it as to be done both in the target and no target flow.

Of course because of LOS on the enemy is influenced from the ground that distance is always relative. The cavalry can come against me unseen until late.

[Saddletank]

What I have noticed Mitra is infantry staying in a square a long time with only minimal cavalry threat. We played a Prussians-v-French game a couple of weeks ago where just a single Chasseur a Cheval squadron kept about 4 or 5 Prussian battalions in square even though I had 2 Uhlan squadrons manoeuvring towards the enemy cavalry.

Could infantry in square be made to check and come out of square if nearby friendly cavalry is seen to be equal or stronger to an enemy cavalry threat?

[mitra76]

Yes but my suggestion is to work inside the fearcheck routine which is used also for determine if cavalry going to charge and in the long check of square formation order (now I don't remember if also for the short, I'm in office).

This is the most sure system to be sure a situation where infantry feels it isn't right to form the square, also block also the cavalry from charge: friend cavalry cannot countercharge in time or have the path or LOS blocked by the ground.

[HoldTheLine]

I have downloaded the SDK and confess in the past couple days I've been looking at these routines - such as infantry in danger from cavalry - and I can see that the kinds of changes being discussed are possible with a modded AI dll. However as has been alluded to there could be unintended consequences from adjusting the reaction distances. It would absolutely be possible to alter the criteria used to determine reactions to threats - or how close a unit is to a friendly etc.

All of this I must say is due to the generosity of the devs providing the actual code, project file, headers etc to do this work. Of all the games I've bought this is unprecedented and I applaud it. I have been able to successfully compile alternative AI libraries without trouble. Thank you NSD!

However this is getting off the track. Perhaps a separate thread is warranted.

[Marching Thru Georgia]

I'm fairly certain my description of how the game treats infantry and cavalry threats is accurate. In an MP game this weekend, using the KS mod, where musket ranges are shortened to 110 yd., average Russian light cavalry had their way with the French Imperial Guard.

One of our Russian divisions was heavily engaged by the Hollanders. The rest of the Guard began marching towards his flank. I moved a regiment of uhlans and hussars forward to intercept the Guard. My intention was not to engage them, but merely slow them down and allow our 2nd infantry division time to move up and engage.

The cavalry commander deployed his squadrons forward, (great cavalry AI by the way), to menace the enemy. Occasionally the Guard would form squares, but most did not. The result was a cavalry romp, riding down Old and Middle Guard units that did not form squares. Nearly all of these attacks were frontal so the Guard had plenty of opportunity to see the cavalry and react, but given the shortened musket ranges they did not.

[mitra76]

I have a doubt: isn't that the Guards were without a stance set at Brigade level when the spot the cavalry for the first time? Stance if not set manually or from the division play, is triggered at musket range, but without a stance the more complex and autonomous orders of battalions are not processed (like for the square->leave square).

[Marching Thru Georgia]

The French side was controlled by the AI. So the division and brigade AI does not set a stance?

[mitra76]

yes they are set, they are specifically important for the AI, but like for Gettysburg they are triggered or at the engaging distance (weapon max range) or when the division send the battle orders (when the first of brigade\batteries of division is engaged, all the others receive their battle orders (place + stance)).

In Gettysburg the consequences of a stance still not triggered were less visible because they did less and less complex actions.

[Marching Thru Georgia]

Was it likely that the shorter musket ranges caused the Guard not to form square as often as they should?

[Saddletank]

...and is it possible to disconnect that link between musket range and reaction distance?

One thing this might allow to be represented is if you can set a reaction distance longer for poorer quality troops so as to represent them forming square early out of fear or lack of good officers, etc. A cavalry threat could then pin poorer quality troops into square from further away, making them less "reliable".