Direction of Retreat

[Arimus]

A question for TOAD, what determines the direction of retreat and why does it seem like units tend to retreat to the flanks rather than to the rear?
Also it seems like since the patch, units that would have routed now divide and retreat to the flanks.

[larryfulkerson]

ORIGINAL: Arimus
A question for TOAD, what determines the direction of retreat....

I always thought that units retreated toward their HQs or the nearest HQ if they didn't have one. I could be wrong however.

[Industrial]

ORIGINAL: larryfulkerson
I always thought that units retreated toward their HQs or the nearest HQ if they didn't have one. I could be wrong however.

That is correct

[Arimus]

My game play experience doesn't seem to support that. It seems to me like they retreat to the nearest friendly unit and if divided, a part will retreat in a different direction, even if that direction is toward the enemy.

[golden delicious]

ORIGINAL: Arimus

My game play experience doesn't seem to support that. It seems to me like they retreat to the nearest friendly unit and if divided, a part will retreat in a different direction, even if that direction is toward the enemy.

Sometimes but not always. I think there's a degree of randomisation.

[JAMiAM]

They will retreat toward the nearest supply source, or HQ unit, and generally into a "safe" location (i.e., a friendly unit) to minimize disengagement attacks. Unfortunately, when supply sources are very near by in lateral, or forward, positions, rather than "to the rear" the units will drag themselves along the front line making things worse, or even forward, where they often get stuck in an exposed bulge.

Making unit retreat more intelligently is on our list of things to work on.

[Industrial]

ORIGINAL: JAMiAM

Making unit retreat more intelligently is on our list of things to work on.

Allow players to place 'rear area' pins on the map, and units will always try to disengage towards the neares pin, that way players would have the most control over retreats.

[golden delicious]

ORIGINAL: Industrial

Allow players to place 'rear area' pins on the map, and units will always try to disengage towards the neares pin, that way players would have the most control over retreats.

Units don't generally necessarily flee where their commanders want them to.

[Arimus]

More player control isn't necessary, just slightly better retreat logic.

[Industrial]

ORIGINAL: golden delicious

ORIGINAL: Industrial

Allow players to place 'rear area' pins on the map, and units will always try to disengage towards the neares pin, that way players would have the most control over retreats.

Units don't generally necessarily flee where their commanders want them to.

*cough* *cough*....
units will always try to disengage towards the neares pin....

Of course they wont always manage to do it, but as long as the tactical situation permits it and they weren't outright routed they should be able to retreat towards that pin, which might for example be the last remaining bridge over a super river.
A retreat is a somewhat organized withdrawal, while a route is a unit beeing overrun, panicking and not really following orders. So retreating units should in most cases be able to follow orders and move towards the pin.

[Menschenfresser]

The best is when one of the attacking units evaps, leaving a hole in the line and the defender retreats forward, as if attacking.
 
Seen this one quite a few times...and it only gets funnier. [:)]

[golden delicious]

ORIGINAL: Industrial

*cough* *cough*....
units will always try to disengage towards the neares pin....

Of course they wont always manage to do it, but as long as the tactical situation permits it and they weren't outright routed they should be able to retreat towards that pin, which might for example be the last remaining bridge over a super river.
A retreat is a somewhat organized withdrawal, while a route is a unit beeing overrun, panicking and not really following orders. So retreating units should in most cases be able to follow orders and move towards the pin.

When Manstein's Army Group South was retreating out of the Ukraine, he had a disagreement with one of his commanders over which direction in which to retreat. Manstein wanted the unit to retreat one direction to continue to present a continuous front, whilst the subordinate wanted to retreat in another direction to ensure the survival of the greatest possible part of his command.

I think in the end Manstein had his way- but you can see how the man on the ground might make a decision which doesn't correspond to the desires of his superior. If the unit is retreating from combat (which is the situation we are discussing) then it is the authority of the man on the ground which will matter. The sort of centrally orchestrated retreat you're describing would be done on the player's own turn- not in response to an attack.

The current system does produce some wierd results. But I think your solution gives the player too much control.

[golden delicious]

ORIGINAL: Menschenfresser

The best is when one of the attacking units evaps, leaving a hole in the line and the defender retreats forward, as if attacking.

Seen this one quite a few times...and it only gets funnier. [:)]

I believe this is actually a deliberate feature. Surrounded units do sometimes break out, you know.

[Industrial]

ORIGINAL: golden delicious


The current system does produce some wierd results. But I think your solution gives the player too much control.

Easy fix, base the likelyhood of retreat towards the retreat-pin on unit and force proficiency, done.

[golden delicious]

ORIGINAL: Industrial

Easy fix, base the likelyhood of retreat towards the retreat-pin on unit and force proficiency, done.

Yeah. That would be just the ticket.

[Arimus]

One thing I hate when a unit retreats to the flank is it sometimes ends up in a hex salient which is attacked again leaving the unit out of movement and supply the following turn. With zero movement, there is a good chance it will get left behind when the force falls back.
Yes, I realize this may have happened in a real war, but if it happened this often then you would think that some military history books or some of my military manuals would have said something to the effect of "retreating to the flank is a death sentence".

[a white rabbit]

ORIGINAL: JAMiAM

They will retreat toward the nearest supply source, or HQ unit, and generally into a "safe" location (i.e., a friendly unit) to minimize disengagement attacks. Unfortunately, when supply sources are very near by in lateral, or forward, positions, rather than "to the rear" the units will drag themselves along the front line making things worse, or even forward, where they often get stuck in an exposed bulge.

Making unit retreat more intelligently is on our list of things to work on.

..wot he said..

[tblersch]

ORIGINAL: Industrial
Allow players to place 'rear area' pins on the map, and units will always try to disengage towards the neares pin, that way players would have the most control over retreats.

Couldn't you do the same thing now just by carefully positioning a unit's HQ?

[Industrial]

maybe, but than you'd lose the supply boost bonus and also any artillery support from the HQ, as designers often put divisional arty into the HQs.

Another problem would be that with so many HQs in most scenarios, it won't always be clear which which HQ really is the closest, not to mention our lacking insight into the retreat algorhitm, meaning as to what radius the game will check for HQs to determine the retreat direction, what it does if the path to the closest HQ is blocked by enemy units etc.

[Arimus]

Supply points seem to have the largest influence on direction of retreat if they are close to the unit. Adjacent friendly and enemy units have a big effect also, even if you play with disengagement off. Direction of enemy attack seems to have little effect.