So my friend and I were playing a 1v1 game on a tiny map. He randomed Broken Lords, I randomed the Cultists. By the time we met, I had converted 12 minor faction villages and I had all those free units roaming all over the map. He, on the other hand, had a really strong army and he sent it straight towards my capital. What option did I have? Converge all my units on his army, that's what! As he neared the edge of my territory he put down a city to act as a forward operating base. I surrounded it and laid siege to the base. Every turn his army tried to break the siege by attacking one of the 6 units surrounding him. I unchecked all the reinforcements and had that one unit retreat, causing it to take a little bit of damage. Every turn he took siege damage to his whole army and I merely swapped out the unit that got damaged by the attack.



TLDR; I was able to overcome a superior army with weak units by exploiting the retreat and siege mechanics.



Screenshot:

Propbuddha wrote:
So Settlers should be able to run around unescorted and players don't have a responsibility to scout out the area before sending a Settler in?

I just explained that escorting a settler is insufficient to protect it in the exact same post you're quoting from.

Antistone wrote:
You need to overpower your intended target by 2:1. But unless I'm mistaken, reinforcing on defense doesn't cost an action point, so you don't need to overpower potential counter-attackers 2:1. You aren't leaving yourself vulnerable to other threats.

Oh, no, you're not mistaken in that.

I notice that no one seems to have commented on any of the ideas I brought up in my first post.

Oh, I was talking about your suggestions when I said that nothing proposed would address the problem-- without the ability to move through retreated enemy units. Which was something you had originally mentioned.



Your suggestions would help, but it would still be slow. Even if the besieged army manages to kill a unit every turn, it's still under siege and cannot escape. So long as the sieger can spare a unit-- any unit at all-- the besieged army remains trapped. If they can attack until units stop retreating, the sieger just doesn't retreat.



Now, that's going to be a problem with most solutions. I would prefer if the speed with which the besieged unit could escape had at least a tiny relationship to the difference in power between it and the units it was attacking, which is why I made the particular suggestion I did-- given sufficient power, a 6 unit army could actually break a 6-unit siege in a turn, even though it was encircled. (In reality, doing so would be risky, and it would take more than a turn, but probably not six.)

There's a public document that shows how close various parts of the game are to being finished?

Here. I guess it's actually "code" that's at 15%, but I can't help but feel that the lion's share of that is AI.

And they released the game while a major category still said 15% (or less)?

Ehh, it's just a matter of high how they set their expectations. In it's current state, Endless Legend is miles ahead of a great many games. All else being equal, I'd rather they aimed higher than lower.

Propbuddha wrote:
Please cite them. I'd especially like to hear the ones that can't be solved by not placing troops in a dangerous situation in the first place

The game could be redesigned for this, but if you just take out retreat, large portions of the game change.



Mostly, you can't scout because you don't have scouts. It's better to have a bigger army than the same army and a scout, because all that happens to the scout is he gets killed all on his lonesome (or drags your army in via slow reserves to a disadvantageous skirmish).



Currently, default vision range is less than default infantry move. So if you wanted infantry to move without risk of getting counterattacked by enemy infantry, you'd be stuck moving at 2 tiles (which gives you enough leftover move to run away if you spot an army.) Against cavalry, it'd be impossible. So use cavalry scouts? But for your separate scout stacks (which are being paid for at full army prices), you're still limited to 4 tiles of movement-- to be safe against infantry. -1 tile of safe movement if they're a tier ahead of you on tech. -2 tiles if they have cavalry more numerous than your scout.



What about demons? They move 8 tiles at default. (And they always seem to home in on my troops unerringly. Not sure that it's magic though.) No amount of scouting is going to protect your settler from an army of roaming demons.



Never mind the fact that unpacified villages-- including those you've parleyed with for a quest, the first criterion for which might be "settle this province"-- spawn units that move and attack on the same turn that they spawn.



You can do sharp senses, you can assimilate Jotus-- and they can do talismans and assimilate Erycis.



I expect that I could come up with further examples if I thought about it more.



So it's not like the game breaks, but the rate of travel halves, and exploration past turn 15-- maybe before that-- becomes a suicide mission. So the balance changes pretty drastically.

If the retreat option must exist it shouldn't be a sure way to escape destruction and a retreating army shouldn't be able to finish out your turn like nothing happened.

Well, it's not, and it doesn't (with the possible exception of Broken Lords settlers).



Against armies with equal move, all retreating does is buy you a turn or two, which is hopefully enough to get to reinforcements, or into your own territory in a cold war.



Against armies that outpower you 2:1 (and that are played by something smarter than endless legend AI), retreat doesn't even buy you a turn. It just costs them an extra movement point.

Antistone wrote:
And the problem can't be solved by "not placing troops in a dangerous situation in the first place" because Settlers have to go into dangerous territory in order to accomplish their function.

So players don't have a responsibility to scout out the area before sending a Settler in? Is picking off unescorted Settlers too OP?



This is the type of sloppy play the Retreat function is designed for. Screw up? Press retreat and you escape.

Propbuddha wrote:
Please cite them. I'd especially like to hear the ones that can't be solved by not placing troops in a dangerous situation in the first place.

Well, one thing that springs to mind is protecting Settlers. If your Settler is involved in a battle--even a battle that you can easily win--it can get killed pretty easily, and depending on the composition of the enemy army, there may not be much you can do about it. I had one battle where I completely wiped out the enemy and took no casualties except for the Settler that was trying to hide in the corner and not bother anyone.



With the retreat option, the Settler can escape direct combat, and you can send your military units without the Settler to counter-attack and destroy the enemy before the next round (at least in principle). I'll happily admit that's not a very good solution to the problem, but it is a solution of sorts.



And the problem can't be solved by "not placing troops in a dangerous situation in the first place" because Settlers have to go into dangerous territory in order to accomplish their function.

natev wrote:
I think there are reasons for retreat to exist, and I don't think the mechanic is so broken as to require scrapping it completely!

Please cite them. I'd especially like to hear the ones that can't be solved by not placing troops in a dangerous situation in the first place.



If the retreat option must exist it shouldn't be a sure way to escape destruction and a retreating army shouldn't be able to finish out your turn like nothing happened.

Antistone wrote:
But setting that aside, do you really feel that you should never be able to retreat from a siege? That seems a bit arbitrary to me.

Yes, If your siege unit is threatened, stop the siege and run away, no need for an easy button. Of course, I'm coming at this from the position that Retreat is a stupid feature to begin with

Antistone wrote:
I notice that no one seems to have commented on any of the ideas I brought up in my first post.

I read them but you raised a bunch of situations (just as you have for my suggestions) where the solution is problematic. Ultimately, we're creating a bunch of mechanics simply so an army can have a convenient escape plan. In my suggestion (which admittedly doesn't solve the encirclement issue), I attempted to address some of the issues with existing mechanics.



A search on the forums will turn up other Retreat shenanigan threads:



/#/endless-legend/forum/22-gameplay-and-ai-issues/thread/6035-critical-endless-siege-exploit

/#/endless-legend/forum/5-general/thread/1415-retreat-option-is-a-pointless-feature

natev wrote:
When you split armies in order to deny retreat, you can't use them to reinforce, because reinforcing uses their action point. So if you anticipate retreat and want to prevent it, you need to overpower your target by 2:1.

You need to overpower your intended target by 2:1. But unless I'm mistaken, reinforcing on defense doesn't cost an action point, so you don't need to overpower potential counter-attackers 2:1. You aren't leaving yourself vulnerable to other threats.

natev wrote:
I think that G2G roadmap still says 15% on AI....

There's a public document that shows how close various parts of the game are to being finished?



And they released the game while a major category still said 15% (or less)?



Sigh. You know, I specifically avoided buying Endless Legend in early access because 4X games take me forever to play and I didn't want to invest that time until the game was done, but 1.0 clearly doesn't mean to Amplitude what it means to me. I like Endless Legend, but I'm kind of wishing I had waited another 6-12 months before buying it.