forum.freeciv.org

A call for discussion: when you defeat a unit, all other units in the same location die, too (unless in a city or a fortress). However, there is an server option to change this, yet, it seems that nobody is playing that way. I'm curious why. Did anyone try and was it so bad, or people simply don't want to try something different?

Hmm, it's likely to punish people for amassing all their troops in one poorly defended spot, or for example, allowing a unit to kill an entire stack of workers at once.

In my opinion, if you enable this option, then it is no longer important the way you move your troops in the board. You just need to pile them all in the best defensive terrain, and advance this way until you reach a hill or mountain next to the enemy city.
I think it would make it harder to defend against advancing enemies and it would make unit tactics even less important compared to economy. That is why I personally do not like it.

and it would make the game very longer...

Not to mention the effects on which it would have on pricey units. Usually I use my Stealth/Bombers specifically for the purpose; the destroy enormous stacks of units in enemy territory.

"Killstack" is one of the rules I'm strggling with. There is argument for and against it. What makes it bad is that it is totally contra-intuitive and lacking realism. The rule also makes it very hard to conquer cities and land troops by boat on the shore. In reality it makes a difference for the attacker whether they bring in 1 company or 4. Not to speak of the resources taken to build them. So it doesn't seem to make sense that if one company is eliminated, the orther 3 are too. - In the game you effectively cannot bring more than one unit next to the enemy city, which then can easily be defeated. Bringing along many troops doesn't seem to matter very much at this point.

I agree the game benefits from the rule, as long as unlimited number of units can mount in a single stack. However, the latter could be modified into something more intelligent.

I am against killstacks because they aren't realistic.
Its not like an entire army will just lay down and die when one of their defensive units dies.
And its also unrealistic to suggest that "if they can beat this, they can beat the others just as easily" because More Dakka really does work in real life.

(a mass of armies in one place are in essence one giant army)

There are X tons of dudes there and they're all going to fight. And more units means more people that your army has to kill, which means more chances for their (much larger force) has to inflict harm on them.

It IS strategic to put a bunch of armies in one place.
In fact spreading your armies out is a bad idea because it makes them weaker.
The only strength to having them spread out is for patrolling and border control.

If you want to make sure that an area is protected, that is what you do.

If you want that gold-bearing mountain with a mine on it to remain under your control, that's where you put 5 defensive units at, and you build a fortress around them too.

And it should take that much firepower to take them out too.
Being in a city or outside of one doesnt change that.

----

I am for the reverse of a killstack rule . . .
For all of the units in a space, a running total is tabulated of all of their stats (attack/defense/hit points) and then when an attack happens they roll with their Combined Attack, Combined Defense, and take damage homogenously.

If a unit is moved out of the space, their initial hit points are reduced by the percentage of the hit point % was sustained from the last time the "big army" was changed.

Well, firstly, one thing that isn't realistic is placing an infinite amount of army on one spot. The second thing that isn't realistic is having a huge army enter enemy territory and functioning completely normally without and need for supplies. The third unrealistic thing is that the army defends equally regardless of what direction it is attacked from: front, flank or rear, especially if it is advancing into the enemy territory.

So, to counter all those things and to prevent abuse of placing 20 defensive and 30 offensive units in one place and then going on a rampage, "killstack" is introduced. Once those three problems (and maybe some others I didn't think of immediately) are solved, we'll be able to remove "killstack". But, as it is now, it is a necessary evil. I don't like it, but removing it would open the Mouth of Hell regarding strategical thinking.

RSSwizard wrote:I am against killstacks because they aren't realistic.

Speaking as someone who likes to disable the kill stack rule (and, in 2.6, stack protection against diplomat actions): How realistic killstacks are depends on the situation you are modeling. ISIS being attacked by the USA is an example of a situation were killstacks adds realism. A battle from the first generation of modern war (before quick battlefield communication, nukes, bomber planes, artillery, machine guns, rifled muskets, etc) is an example of a situation were killstack subtracts realism.

sveinung wrote:

RSSwizard wrote:I am against killstacks because they aren't realistic.

Speaking as someone who likes to disable the kill stack rule (and, in 2.6, stack protection against diplomat actions): How realistic killstacks are depends on the situation you are modeling. ISIS being attacked by the USA is an example of a situation were killstacks adds realism. A battle from the first generation of modern war (before quick battlefield communication, nukes, bomber planes, artillery, machine guns, rifled muskets, etc) is an example of a situation were killstack subtracts realism.

I like the line of reasoning behind this, and to a point it also seems like it depends on the units involved. A diplomat would be a person, not an army of them; a cavalry unit is likely not a single horse and rider. A group of cavalry attacking a diplomat unit would be able to kill a person, and a single diplomat could bribe/persuade a group (or group's leader), but it's unlikely to be able to do so to an entire army or stack of units. The level of plurality (one, a few, dozens, hundreds) seems important to this concept and it's also not generally well defined in game (currently it's just an abstraction, which makes it easier on everyone).

It also makes sense to me that some units would be more likely to defend a stack or perish protecting everyone else (defense units that cause the tile to ignore killstack); think 300, a group working to stop an army from crossing a pass. What happens to the conquered tile vary, possibly pushing everyone else back as one option perhaps? And there are units, bombers being a great example, that attack a geographical space more than attacking a unit (where killstack, or at least damaging all units makes complete sense). Other units, like diplomats, could be thought of as targeting the leaders who then bring the rest of the troops with them and would never affect more than a single (or very few) units on a tile stack.

Getting any of this into rulesets however is a whole different challenge. Given that many features are based on existing Civ games, there's not any precedent that I'm aware of for this time of awareness of plurality of a unit or it's affects, there's also potential for conflict between units that attack everyone versus units that defend a tile. Adding in potential interpretation is also likely to cause some disagreement in precedence and who can do what, while each person/server can pick their ruleset, picking a default might be more difficult (or just leaving it off as the default, killstack remaining all or nothing by default).

You could likely manage something basic at least as a script based on the unit_lost server signal. It would certainly make for an interesting proof of concept. I'm not sure if you can define an arbitrary tag in the unit definitions for this though, but I imagine it should be doable.