Nordfriese
Joined: 2017-01-17, 18:07
Posts: 2080
OS: Debian Testing
Version: Latest master
One Elder of Players
Location: 0x55555d3a34c0
|
Posted at: 2020-03-04, 15:32
king_of_nowhere wrote:
This would be counterproductive IMHO because it means that once you succeeded in conquering one building inside the enemy's territory, counter-attacks are made much more effectively. Also it encourages players who conquered one site to try and conquer all other nearby sites as well so they can keep what they got. This would result in attacks being necessarily followed up by more fighting from both sides, which is the opposite of what we want to achieve.
are you telling me that sometimes you conquered a building in enemy territory and then the enemy didn't try to get it back and you didn't try to conquer anything else? seems real strange.
Yep I only ever play against the AI, where this is part of my favourite strategy, see below
JanO wrote:
I think easier recapturing of captured buildings is totally in favour of making defense stronger
again, in multiplayer (which is where those balance changes really matter)
Although balancing is of course most important for MP matches, let's not forget that Widelands should still be fun to play in singleplayer
Nordfriese wrote:
But increasing the attacker's need to launch follow-up attacks is certainly not defensive
actually, it is. it forces the attacker to only attack if he has an advantage.
My preferred strategy against the AI: On first contact I usually have a few well-trained soldiers (a bit more than the AI), which I send to attack and conquer a couple of sentries. The point is to kill off some scores of enemy rookies and evade-only soldiers. Then I just leave these conquered sentries sitting there and do nothing there at all. The AI will frequently throw some low-level soldiers at them which are killed one by one and my soldiers always recuperate afterwards. Wonderful mechanism to keep the enemy in check by doing nothing. (That's why I'm strongly against forbidding such buildings to heal their soldiers). Meanwhile I balance out my economy so it starts making fully promoted soldiers one at a time which I slowly gather near the frontier. When I am much stronger than the enemy, I launch massive attacks and eradicate the AI.
right now, the problem is that the attacker can launch a couple of soldiers (fully healed, the attacker can choose) and force all defender soldiers to move out to intercept. this stops the defender from healing his own soldiers, so he will lose a war of attrition. if the attacker has more power and swarms the enemy, that's not something we want to stop
Actually I consider it a bug that all soldiers come out to defend if there are only a few attackers. Keeping the number of defenders proportional to the number of attackers would already be an improvement.
Top
Quote
|
|
|
king_of_nowhere
Topic Opener
Joined: 2014-09-15, 18:35
Posts: 1668
One Elder of Players
|
Posted at: 2020-03-04, 17:34
Nordfriese wrote:
My preferred strategy against the AI: On first contact I usually have a few well-trained soldiers (a bit more than the AI), which I send to attack and conquer a couple of sentries. The point is to kill off some scores of enemy rookies and evade-only soldiers. Then I just leave these conquered sentries sitting there and do nothing there at all. The AI will frequently throw some low-level soldiers at them which are killed one by one and my soldiers always recuperate afterwards. Wonderful mechanism to keep the enemy in check by doing nothing. (That's why I'm strongly against forbidding such buildings to heal their soldiers). Meanwhile I balance out my economy so it starts making fully promoted soldiers one at a time which I slowly gather near the frontier. When I am much stronger than the enemy, I launch massive attacks and eradicate the AI.
well, there is absolutely nothing in your proposed changes that would hamper that strategy. first of all, AI soldiers would not heal, because they are attacking, and they are in your land. second, it does not matter that they would heal if wounded, because that would require a wounded soldier, i.e. a soldier who won a combat. since those hapless ai soldiers are all killed in thei first fight, they never have a chance to heal.
right now, the problem is that the attacker can launch a couple of soldiers (fully healed, the attacker can choose) and force all defender soldiers to move out to intercept. this stops the defender from healing his own soldiers, so he will lose a war of attrition. if the attacker has more power and swarms the enemy, that's not something we want to stop
Actually I consider it a bug that all soldiers come out to defend if there are only a few attackers. Keeping the number of defenders proportional to the number of attackers would already be an improvement.
yes, i agree, and i have made several proposals for it: namely, that you could fix a minimum amount of soldiers that will stay inside the building, a minimum healt below which soldiers will not move out, and then you could manually pick if you wanted to send out some more soldiers.
Top
Quote
|
|
|
stonerl
Joined: 2018-07-30, 00:03
Posts: 327
Tribe Member
|
Posted at: 2020-03-04, 18:04
I liked the idea of having simultaneous healing of soldiers, inside a building. I don't see a reason why multiple soldiers can't heal simultaneously, this somehow seems to be an arbitrary limitation.
Also, somehow I still dislike the idea that the healing rate outside is tied to a building.
I assume the formulas I posted are not worth a consideration then (I added the tables later, BTW)?
Top
Quote
|
|
|
hessenfarmer
Joined: 2014-12-11, 23:16
Posts: 2749
One Elder of Players
Location: Bavaria
|
Posted at: 2020-03-04, 18:14
stonerl wrote:
I liked the idea of having simultaneous healing of soldiers, inside a building. I don't see a reason why multiple soldiers can't heal simultaneously, this somehow seems to be an arbitrary limitation.
Also, somehow I still dislike the idea that the healing rate outside is tied to a building.
I assume the formulas I posted are not worth a consideration then (I added the tables later, BTW)?
No I liked that formula as my first liking a common value would tamper balance probably. It's just I am AFK and on my tablet so typing is really difficult.
Top
Quote
|
|
|
Solstice_s_Return
Joined: 2020-01-28, 13:24
Posts: 62
Likes to be here
Location: Finland
|
Posted at: 2020-03-04, 18:16
Nordfriese wrote:
My preferred strategy against the AI: On first contact I usually have a few well-trained soldiers (a bit more than the AI), which I send to attack and conquer a couple of sentries. The point is to kill off some scores of enemy rookies and evade-only soldiers. Then I just leave these conquered sentries sitting there and do nothing there at all. The AI will frequently throw some low-level soldiers at them which are killed one by one and my soldiers always recuperate afterwards. Wonderful mechanism to keep the enemy in check by doing nothing. (That's why I'm strongly against forbidding such buildings to heal their soldiers). Meanwhile I balance out my economy so it starts making fully promoted soldiers one at a time which I slowly gather near the frontier. When I am much stronger than the enemy, I launch massive attacks and eradicate the AI.
From my point of view, if there's a perfect strategy in a game, then the game needs to be fixed. Especially if that kind of strategies are easy to conduct.
Besides, that strategy wouldn't be rendered useless. It would just require some more attention from a player's behalf and if those units get too wounded, military sites have to be burned (make sure that one of your roads gets close enough so that troops can safely return..) and another sites captured to keep the strategy going.
Finally, that kind of bonus would be completely logical, because if supply line is broken, it would be a lot harder to get medical supplies there. That'd be simulated with some lines of code, no other requirements involved.
Edited: 2020-03-04, 18:19
Top
Quote
|
|
|
teppo
Joined: 2012-01-30, 09:42
Posts: 425
Tribe Member
|
Posted at: 2020-03-04, 18:40
Nordfriese wrote:
I really like the idea. But as far as I understand the code, the healing speed depends on the building a soldier is from? So, a solder from a babarian citadel heals at 110 health points per second while a soldier from a sentry only heals 40 per second?
I don't think it's a problem. This would encourage players to preferably build bigger buildings, which are harder to conquer and thus gives another small disadvantage to the attacker. So this is fine IMHO.
Many small buildings have some advantages against one big:
- They can be built in parallel, rapidly
- Soldier exchange mechanism works in parallel, too, bringing the most trained folks to frontline sooner
- If only one solder per site is healed, there is more simultaneous healing.
The big buildings should have some advantages to offset this. Just containing more soldiers does not make them that strong defenders: The soldiers swarm out anyway. In case of many small buildings, losing one is not that bit a problem: If another is nearby, it just burns down. No need to baby-sit & dismantle.
Top
Quote
|
|
|
Nordfriese
Joined: 2017-01-17, 18:07
Posts: 2080
OS: Debian Testing
Version: Latest master
One Elder of Players
Location: 0x55555d3a34c0
|
Posted at: 2020-03-04, 20:05
well, there is absolutely nothing in your proposed changes that would hamper that strategy. first of all, AI soldiers would not heal, because they are attacking, and they are in your land. second, it does not matter that they would heal if wounded, because that would require a wounded soldier, i.e. a soldier who won a combat. since those hapless ai soldiers are all killed in thei first fight, they never have a chance to heal.
Quite my point The (attacking) AI would not benefit at all from recuperating, no matter how we implement it. The (defending) soldiers in my newly …acquired… sentries however would be penalized severely if the healing rate of buildings is made dependent on their connection to a warehouse, because the strategy only works if my soldiers regain the health points they lose in the intermittent enemy assaults.
I assume the formulas I posted are not worth a consideration then (I added the tables later, BTW)?
What I dislike about your formula is that the healing rate should not depend on the soldier's maximum health. Otherwise the difference between defence and health upgrades might become negligible. (The main difference between them except for single-point edge-cases in battles is that health training decreases the relative healing speed)
Besides, that strategy wouldn't be rendered useless. It would just require some more attention from a player's behalf and if those units get too wounded, military sites have to be burned (make sure that one of your roads gets close enough so that troops can safely return..) and another sites captured to keep the strategy going.
In other words, a strategy that currently consists of attacking once or twice and then doing nothing but defend oneself for hours would be changed so I have to frequently attack. I though we wanted to take the focus away from attacking…
Top
Quote
|
|
|
stonerl
Joined: 2018-07-30, 00:03
Posts: 327
Tribe Member
|
Posted at: 2020-03-04, 20:56
If we only make it dependent on the health level we could use this formulas:
healing_rate_per_second = building_heal_per_second * (10 + level ) / 10
healing_rate_per_second = 35 * (10 + level ) / 10
Then we would have the same factor for every tribe (10% increase per level). The healing rate would then be like this:
Barbarians
Level |
Outside |
Sentry |
Barrier |
Tower |
Fortress |
Citadel |
0 |
35 |
80 |
130 |
150 |
170 |
220 |
1 |
38,5 |
88 |
143 |
165 |
187 |
242 |
2 |
42 |
96 |
156 |
180 |
204 |
264 |
3 |
45,5 |
104 |
169 |
195 |
221 |
286 |
Empire
Level |
Outside |
Blockhouse |
Sentry |
Outpost |
Tower |
Fortress |
Castle |
0 |
35 |
60 |
80 |
100 |
150 |
170 |
220 |
1 |
38,5 |
66 |
88 |
110 |
165 |
187 |
242 |
2 |
42 |
72 |
96 |
120 |
180 |
204 |
264 |
3 |
45,5 |
78 |
104 |
130 |
195 |
221 |
286 |
4 |
49 |
84 |
112 |
140 |
210 |
238 |
308 |
Atlanteans
Level |
Outside |
Guardhouse |
Tower Small |
Tower |
Guardhall |
Tower High |
Castle |
0 |
35 |
75 |
100 |
120 |
140 |
170 |
200 |
1 |
38,5 |
82,5 |
110 |
132 |
154 |
187 |
220 |
Frisians
Level |
Outside |
Wooden Tower |
Wooden Tower High |
Sentinel |
Outpost |
Tower |
Fortress |
0 |
35 |
40 |
70 |
100 |
160 |
170 |
220 |
1 |
38,5 |
44 |
77 |
110 |
176 |
187 |
242 |
2 |
42 |
48 |
84 |
120 |
192 |
204 |
264 |
Edited: 2020-03-04, 21:06
Top
Quote
|
|
|
GunChleoc
Joined: 2013-10-07, 15:56
Posts: 3317
One Elder of Players
Location: RenderedRect
|
Posted at: 2020-03-04, 22:03
Nordfriese wrote:
Quite my point The (attacking) AI would not benefit at all from recuperating, no matter how we implement it. The (defending) soldiers in my newly …acquired… sentries however would be penalized severely if the healing rate of buildings is made dependent on their connection to a warehouse, because the strategy only works if my soldiers regain the health points they lose in the intermittent enemy assaults.
The proposal was not to heal soldiers in the field if the building has no connection to a warehouse. Soldiers inside the building will still be healed as usual.
Busy indexing nil values
Top
Quote
|
|
|
JanO
Joined: 2015-08-02, 11:56
Posts: 177
At home in WL-forums
|
Posted at: 2020-03-04, 23:43
Then I got that wrong, too (even if I did not post anything since that proposal was made).
@ Nordfriese
I think I play exactly as you do. Nevertheless I always wonder if I even need an opponent for this. It is about just managing the economy, conquering the whole map and training my army - without being disturbed by the AI. Eradicating the AI... sometimes I don't even finish that, because it seems either unnecessary, unfair or boring. For this type of game we would need a way more complex economy system (talking about factor 10 minimum in number of wares + research, allocation tasks, experience for all workers, maintenance cost...) where economy itself is the enemy. As in all those Tycoon-games or 'Banished'.
In widelands it often makes me sad that our strategy against AI is even possible, because this AI behaviour seems not very efficient or intelligent. It is just putting an attack, even with one single rookie, above everything.
Back to topic:
I like the proposed changes in healing.
Still long term project should be like king of nowhere (and me, too) proposed full control on defending forces.
I dream of a system which looks as following:
- make the soldiers inside buildings sortable by the player (drag and drop)
- on top of the line of soldiers put a line of shield-icons which can be clicked to allow or forbid the corresponding soldier to leave the building for defense-counterattacks
- below the line of soldiers a line of sword (axe, spear...)-icons, clickable to preselect soldiers for attacking (only those show up in attack-windows on enemy buildings)
- a second line of soldiers which is normally empty but hosts all wounded soldiers. It overrides the player selection for defense and attack temporarily (makes them stay inside) but puts every soldier back in its former position once it is healed. Wounded soldiers can also be send back in position by clicking on them
- the line for wounded soldiers could get a selectable threshold (maybe)
- soldiers for counterattacks leave from left to right (or from right to left, whatever seems more intuitive)
- soldiers without conterattack-allowance stay inside until an enemy knocks on the door
- new buildings give all soldiers (except the first one) the defense-counterattack-tag (maybe? - we need some default behaviour, even if it is not like this)
- two new areas/fields inside the building's window; one that releases soldiers towards the warehouse when a soldier is drag&dropped onto it (this replaces the current solution for throwing out a soldier by clicking on it), and a second field that manually releases additional soldiers for a defense-counterattack (only when enemy soldiers are spotted)
Top
Quote
|