AGEOD Forums

How does the engine compute which regions to retreat or withdrawl into?
I'm seeing alot of odd choices in play, units often choosing to retreat in a direction that effectively puts them behind enemy lines, or has the effect of seemingly retreating through the attack route.

For example, In the Austerlitz scenario, I had Spranger guarding Ausburg. During the turn, Davout moved down from Ansbach to Ausburg. Spranger avoided engaging and withdrew....to Ansbach! (crossing a river in the process!) looking at the map, almost ANY other region would've been a better, more logical choice-west to friendly forces, south to friendly forces, east to avoid entrapment..North is not only an odd choice for strategic merit, but it seems stranger in that the retreat effectively goes through the enemy front.

I've seen this in other scenarios as well..Hohenlohe-Ingelfingen tends to retreat south while moving north and being attacked from the south in Jena, and so on..

So, just wondering, what pulls a withdrawing unit into a given region?
Is it just random?

There is a big incentive to move back to the region you previously moved. Also cities are favored as regions with a high military control %.
I would be interested to get your game (the current turn and the previous one please, in a zip)

Arrr...I accidently deleted that save, I don't have it anymore.
I'll run it again a few times-I'm sure I'll have the same result before too long.


On that one, I'm guessing it must've been attracted to the 100% prussian military control?
In 1.02 bayern is slowly turning french (loyalty?)
it's still wierd because it didn't retreat back south from where it moved from, and it retreated through the attackers movement route..

Yes this is weird. Can you please check and post the content of NCP\settings\Control&Retreat.opt

Sure, here it is:

ctlContested= 5 // Minimum control gained upon entering a region (if not passive)
ctlAllowRetreat = 5 // Minimum control to have in a region to allow a retreat into it
ctlRetreatAdjCity= 5 // Interest in retreating toward a region with a city (per level)
ctlRetreatAdjFort= 30 // Interest in retreating toward a region with a fort (per level)
ctlRetreatAdjDepot = 25 // Interest in retreating toward a region with a depot (per level)
ctlRetreatLandLink = 10 // Interest in retreating toward a region, value per land link
ctlRetreatPrevSubSpaceCoeffH = 250 // Coefficient applied to the interest if the region is the one where we are coming from
ctlNoBeachHead = 10 // Minimum control to have so that a region is not a beach/riverhead if you have to cross a body of water before attacking

It should works and apparently it does not... If you get a new, weird, situation, send me the save please (and the turn before so I can replay the turn)

I am playing the 1806 campaign in the demo and also have noticed strange retreats by the Prussians. Typically, the Prussians do not fall back on their Line of Communications (LOC) as they should, but retreat away from them toward the French LOC, i.e. S or SW. They are isolating themselves in the process, but also creating chaos by forcing the French to go back to fight them.

It would be nice if the program required that no retreat could increase the length of a force's LOC, since placing the enemy between them and their supply source would be suicidal for a defeated army.

I was hoping to get some further discussion of the retreat logic in this game as I find it very questionable, even as set forth in post #5 of this thread.

These preferences may well be suitable for the quasi guerrilla warfare of the French & Indian War and the American War of Independance, but not for the major European land wars of Napoleon's era. When your armies top off at a few thousand men, supplies likely can be found in any friendly region with a city, fort and the like. When armies are upward of 50,000, established supply networks are necessary and these do not adapt easily to an enemy army between the friendly army and its supply hub.

I realize that the game doesn't currently provide "rules" for this supply line, or "line of communication", but without them the retreats I've observed make no sense at all. Armies of the Napoleonic Wars didn't just retreat into any adjacent area that was "friendliest", or had a "friendly city or fort". They fell back on their LOC toward their active supply source. This line was intensively "protected" during the campaign and during battle, and cutting it would be a decisive move by the enemy.

Is it possible somehow to adjust the coding to take these principles into account? Virtually all other Napoleonic games I'm aware of require retreats on existing LOC's where possible. This seems important to me.

The logic has been reviewed and there is nothing spotted right now, so ...

It should works and apparently it does not... If you get a new, weird, situation, send me the save please (and the turn before so I can replay the turn)

Pocus wrote:The logic has been reviewed and there is nothing spotted right now, so ...

Oh, I understand that. I was hoping to get discussion of the validity of the game's actual logic as intended. I.e., this:

ctlContested= 5 // Minimum control gained upon entering a region (if not passive)
ctlAllowRetreat = 5 // Minimum control to have in a region to allow a retreat into it
ctlRetreatAdjCity= 5 // Interest in retreating toward a region with a city (per level)
ctlRetreatAdjFort= 30 // Interest in retreating toward a region with a fort (per level)
ctlRetreatAdjDepot = 25 // Interest in retreating toward a region with a depot (per level)
ctlRetreatLandLink = 10 // Interest in retreating toward a region, value per land link
ctlRetreatPrevSubSpaceCoeffH = 250 // Coefficient applied to the interest if the region is the one where we are coming from
ctlNoBeachHead = 10 // Minimum control to have so that a region is not a beach/riverhead if you have to cross a body of water before attacking

I'm saying that there should be other "priorities" to retreat than just those listed above.

ctlRetreatPrevSubSpaceCoeffH = 250 clearly state that you try to retreat from where you arrive.

The various bonuses also give an incentive to retreat to a region where you posess a big city or a depot (so this take LOC somehow).

Now, to fully take LOC into account, perhaps is missing a variable based on military control.

I'd think that adding a military control variable would be a big improvement.

I would think, also, that a variable for the nation's capital would also be appropriate. In 1805 the Austrians fell back toward Vienna. In 1806 the Prussians fell back on Berlin. 1809, Vienna again. 1812 Moscow. 1814 Paris.

IIRC, the game "Empires in Arms" (using a region system similar to NC) has the national capital as one retreat priority. Of course, this is perhaps also because a LOC of the army ultimately goes back to the national capital (or perhaps for the English army their base port city).

As I noted earlier, in the demo I am seeing Prussian armies retreating toward the enemy frontier, not back toward Berlin on their LOC. Maybe someone with more experience in other campaigns can comment on retreat behavior there.