Vision Rework

[b2198]

Currently what i'm seeing is that invisibility is not really useful as a mechanic, because detecting invisible units is almost as easy as just getting vision of the area they are coming from, so what I suggest is separating the mechanics of vision and detection of invisible units in two different stats, I'll call them "vision range" and "detection range" in this post. It would work like this:

-visible units are seen if they are inside the vision range of the enemy

-invisible units are seen if they are inside the vision range of the enemy AND inside the detection range of the enemy

-current units that don't detect invisible units could be converted to the new system by giving them a detection range of 1 (only detect when they are adjacent)

-current units that detect invisible units could be converted to the new system by giving them a detection range equal to their vision range, before changing them on a per-unit basis

-city patrol (or however it's called in english, the research that gives +1 vision range to buildings currently) would give both +1 vision range and +1 detection range to buildings

-units on a mountain would get +1 vision range, but no bonus to detection range


And some of my ideas for unit changes after this:

-watchtower would still have 8 vision range, but would have 5 detection range, instead of the current 8

-flying scout would get its vision decreased from 8 to 7, and its detection range from 8 to 4

-ranger would get its detection range decreased from 7 to 5, becoming comparatively better than flying scouts at scouting on some maps

-light cavalry would get its detection range increased from 1 to 3

-assassin would get its detection range decreased from 5 to 4, but would be able to move through mountains (maybe also a small rework on them, reducing their speed from 3 to 2 but giving them an ability that works similar to elephant run, so that they can move faster if they aren't attacking or using another ability)

-scout would get its vision range decreased from 7 to 6 and detection range from 7 to 5

-spy would maintain its detection range at 6, becoming the best unit for detecting invisible units, but way slower than scout in most cases

-guerrilla would get its detection range increased from 1 to 3

-towers in general would have their vision range reduced by 1


Now, I know, this would require some changes to the way the engine handles visibility, and I know that there are higher priorities on the list, but this is a suggestion for later (probably after Unity version is completed), and, at least by how I imagine it works currently, wouldn't be a change that big, code-wise, and would greatly improve the vision game, instead of the current "who can sacrifice more units to get vision of the frontlines?", that doesn't leave much use for invisibility.

[Endru1241]

Few remarks from me:
1. detection range could theoretically be different from sight even now
All it would take is removing all "can see stealth" specs and give active abilities, that decrease sight and grant spec.
There is also system special reveal, but I am not sure if it also shows invisible units. Plus not sure how convenient in usage it is.
2. I have no idea why only very specific units among those not capable of seeing stealth are listed for detection range increase.
E.g. if guerilla, why not skirmisher?
But above it - wouldn't that hugely decrease usage of scouts?
Why even bother with it if light cav can see the same range and only have 2 less in detection, while still capable of attack and tc capture?
If I'd get 2 turn guerilla with 3 detection range - I'd rather use those and as a bonus eliminate any enemy scouts or spies in 1 attack.
3. There is very important information either missing here or being not realized.
Units do not see anything as they move.
NOTHING.
There is only sight check on start of the turn, end of the move and I think, but not sure - on sight refresh, which includes movement end of other units.
So e.g. flying scout with detection of barely 4 tiles could easily miss unit (1,2 - flying scout initial and end positions, E - stealth enemy):
_____=_____
____===____
___=====___
__=======__
_====2====_
__=======__
___=====___
____===E___
___=====___
__=======__
_====1====_

4. If detection range would be a parameter, then there would be a possibility to decrease it to 0.
How would we deal with it when encountering non-steppable stealth unit?
Unit would see what it encountered and what forced it to stop?
And if yes, then with detection range of 0 - it would loose sight of such unit on refresh.

5. If we were to be making stealth changes - I'd opt for making it not as inborn parameter, but rather as an effect - reveals stealth units of categories [] on tile.
When used in aura - in range of x.
When used in some effect with area - in area of x of target tile.
Not very useful for AoS, but it would give interesting options for more modern variants with submarines, stealth planes, camo units, regular soldiers, guerilla.

6. Why would units not detect things outside of their sight, but in detection range?
I'd rather give mentioned above effect a spec or parameter that would make it work on stealth units or not and any kind of detection would be resulted in revealing unit.
7. Is really stealth that much important to this game to bother with such changes?

[b2198]

2. I have no idea why only very specific units among those not capable of seeing stealth are listed for detection range increase.
E.g. if guerilla, why not skirmisher?

Yeah, skirmisher would make sense too, I just mentioned the ones that came to mind at the moment.

But above it - wouldn't that hugely decrease usage of scouts?
Why even bother with it if light cav can see the same range and only have 2 less in detection, while still capable of attack and tc capture?
If I'd get 2 turn guerilla with 3 detection range - I'd rather use those and as a bonus eliminate any enemy scouts or spies in 1 attack.

IMO turn cost still balances that out, if you need immediate vision, or just a 1 turn unit to hold off for a little longer a TC that's being attacked, you make a scout, and the 2 less in detection would also make a bigger change than it looks, for the same reason you said in point 3., if you need fast TC capture across medium to long distances while still scouting the region, and can afford the 3 turn cost, you make a light cavalry, the 3 detection range would be just a small bonus, as it would make the opponent have to plan the route for its stealth units around it, instead of getting almost next to it and still being safe in most cases.

3. There is very important information either missing here or being not realized.
Units do not see anything as they move.
NOTHING.
There is only sight check on start of the turn, end of the move and I think, but not sure - on sight refresh, which includes movement end of other units.
So e.g. flying scout with detection of barely 4 tiles could easily miss unit (1,2 - flying scout initial and end positions, E - stealth enemy):
_____=_____
____===____
___=====___
__=======__
_====2====_
__=======__
___=====___
____===E___
___=====___
__=======__
_====1====_

Yes, I am aware of this, but that would also mean it is a bit harder to detect them, leaving more room for them to scout without being spotted, but reducing their usability for stealth detection.

4. If detection range would be a parameter, then there would be a possibility to decrease it to 0.
How would we deal with it when encountering non-steppable stealth unit?
Unit would see what it encountered and what forced it to stop?
And if yes, then with detection range of 0 - it would loose sight of such unit on refresh.

Would it be possible to guarantee a minimum number for a stat? If not, then it would probably need another factor in the check, like "if inside vision range AND (inside detection range OR unit adjacent) then reveal unit".

5. If we were to be making stealth changes - I'd opt for making it not as inborn parameter, but rather as an effect - reveals stealth units of categories [] on tile.
When used in aura - in range of x.
When used in some effect with area - in area of x of target tile.
Not very useful for AoS, but it would give interesting options for more modern variants with submarines, stealth planes, camo units, regular soldiers, guerilla.

Well, a "search for stealth units around the selected tile" ability would probably also work, if the passive detection is turned off, also I didn't think about giving an aura of detection, that would probably also work mostly like what I am suggesting.

6. Why would units not detect things outside of their sight, but in detection range?
I'd rather give mentioned above effect a spec or parameter that would make it work on stealth units or not and any kind of detection would be resulted in revealing unit.

Hm, yeah, maybe something like a spy could learn about the location of some stealth unit somewhere it can't see by some other means, so that could make sense in some cases.

7. Is really stealth that much important to this game to bother with such changes?

Well, not currently, but given how much information is important in this game, improving the mechanics for obtaining and concealing information would make it even better IMO.

[Endru1241]

Summing up your explanations - fair enough.
Provided, that stealth would be boosted by the change enough to be more important.
Although some kind of saboteur would be needed imho.

Let's see others opinion.
Especially the dev - it's his game and I think that this would be huge enough not to ever call it simple rebalance.
Rework is probably good word, just like you named it.

[makazuwr32]

Currently i am thinking about changes within current system to stealth and how to increase value of stealth units.

One of ideas was saboteur unit who will deal with buildings (disable production, deal damage, reduce armor, vision, actions...). Ofc that will come with much more costly and more limited ways to detect invisible units.

[Endru1241]

Currently i am thinking about changes within current system to stealth and how to increase value of stealth units.

One of ideas was saboteur unit who will deal with buildings (disable production, deal damage, reduce armor, vision, actions...). Ofc that will come with much more costly and more limited ways to detect invisible units.

Do you have any idea for production disabling with current effects?

[makazuwr32]

In aof we have curse effect. That one can be used for example.

[Endru1241]

In aof we have curse effect. That one can be used for example.

I am pretty sure it's hardcoded. I don't see effect json for it in files and there is no information regarding anything similar in modders lounge topic effect definition json.
Just like animate death.

[makazuwr32]

Well you could simply increase production cost instead for example. Ability similar to money courier but lasts for several turns and increases production by +1 to compensate basic production of building and eventually "stun" it.

[Endru1241]

Well you could simply increase production cost instead for example. Ability similar to money courier but lasts for several turns and increases production by +1 to compensate basic production of building and eventually "stun" it.

That is what I planned to use.
Not sure if it'll work though if I set it as related effect.

Well, I'll have to experiment, like always.

[phoenixffyrnig]

I missed this first time around. I really like this idea, it would add a lot. I appreciate it might not be high on the priorities but well worth keeping in mind I think.

In answer to the question is stealth really that important? At the moment, no, because it is too easy to counter stealth (unless they are hidden in something). But if this comes into play then stealth would become a viable strand of play, especially as we have stealth fighting units as well as agents.

[makazuwr32]

Actually maybe we must make similar change in aof...
Especially when i look through all planned invisible units.

[Stratego (dev)]

curse uses this:

public static final int SYSTEM_SPECIAL_EFFECT_DISABLE = 115; //Disables unit: considers unit "disabled" also removes all action point and movement points of a unit, on effect removal re initializes them.