Should invisibility be scrapped?

[Jetrel]

I strongly disagree with the removal of invisibility. REALLY strongly disagree. Eventually, we will need a better AI to handle it, but it is a feature people will always be asking for, and IMHO, it is a waste to throw such a thing away - especially when certain units like the shadow basically base their existence off of it.


The only gameplay problem with invisibility is that it is expensive to spread out units to detect it. Perhaps certain units should be more capable of detecting invisibility - having a field of "true sight" which spreads 2,3, or even 4 tiles out (mages and such).

Regardless, though, it, in all of its different incarnations, is a feature that almost all major RPGs and strategy games use, and the existence of it makes wesnoth much more flexible.

[Elvish_Pillager]

it is a waste to throw such a thing away

Regardless, though, it, in all of its different incarnations, is a feature that almost all major RPGs and strategy games use, and the existence of it makes wesnoth much more flexible.

You still haven't stated ANY reason why it's any use at all.

I wouldn't mind a little inflexibility when it prevents thing which are wholly bad.

- especially when certain units like the shadow basically base their existence off of it.

The shadow plays so much second fiddle to the Wraith that it needn't exist anyway.

The only gameplay problem with invisibility is that it is expensive to spread out units to detect it. Perhaps certain units should be more capable of detecting invisibility - having a field of "true sight" which spreads 2,3, or even 4 tiles out (mages and such).

And let the arms race begin...
Really, having abilities that have no function but to counter specific abilities is something that has been stated time and again that shouldn't be done.

[cobretti]

(Blah blah) In fact, one of the things that is most distinctive about Wesnoth is that you know everything that's going on, and you can make strategic decisions based on real knowledge and not make said decisions based on guesses. (Of course, you make 'guesses' all the time on what random numbers are going to be, but there you are actually informed of the probabilities and don't have to leave that to guessing too. Besides, the combat luck 'evens out' but it doesn't make sense for strategic luck to even out.) And lastly it kills undoing too! I like the undo feature because it lets me get away with not spending lots of extra time just making sure I don't misclick! Invisibility is even worse in this case because the click-to-attack feature doesn't let you undo from it! That ruins the original reason said feature was proposed!

My eyes have been cleared! Remove all non-undoable events from the game now!

Also, ranting about why everybody show leave Wesnoth and play some other games because you don't like a feature Wesnoth has, and you want to be removed, is a bit harsh. There's no need to do so, try to calm a bit, please.

Many people don't like being ambushed in the fog. True. But many people don't like to have their units killed, even when they put them in really bad positions, and call the other people cheaters. If anybody can learn to use terrain effectively, he can also learn to use fog effectively, too. Same for ambush.

As for ambush itself, I like it, and think it should stay, but I'm only stating my opinion, as the final decision doesn't belong to me.

[cobretti]

Same for ambush.

As for ambush itself, I like it, and think it should stay, but I'm only stating my opinion, as the final decision doesn't belong to me.

str_replace("ambush", "invisibility", $my quote);

[Elvish_Pillager]

My eyes have been cleared! Remove all non-undoable events from the game now!

Well, the ambushes in Crossroads are almost as bad as ambushes using Ambush.

Also, ranting about why everybody show leave Wesnoth and play some other games because you don't like a feature Wesnoth has, and you want to be removed, is a bit harsh. There's no need to do so, try to calm a bit, please.

Right...

if you want to play an unbalanced game... well, it isn't Wesnoth.

If anybody can learn to use terrain effectively, he can also learn to use fog effectively, too. Same for ambush.

Nope, this is incorrect. Terrain and fog/ambush are entirely different.

Terrain: Place your unit here and you will be less likely to be hit!
Simple. Easy to learn, maybe a bit (lot) harder to master but it's simple.

Fog: Place your unit here and you may be able to prevent your opponent from using tactics which may or may not enable him or her to attack your position is such a way that you may or may not get more units killed/injured than you would have if what you saw was all that was there!
It takes a LOT longer even to LEARN fog, and to 'master' it would be... impossible, for me. I can never abide that much guesswork.

[Disto]

I hate fog, but i like ambush as many a time have i got badly injured in battle and been able to slink off and get away without trouble, as my favourite units are Avengers.

[Elvish_Pillager]

my favourite units are Avengers.

Same here. I actually think Ambush makes Avengers overpowered.

[Disto]

I don't, we all have our own opinions.

[Dragonking]

Maybe do a poll - people will be able to vote what they think about invisiblity.

[MadMax]

I say keep invisibility.

1. It does add fun to the game. Ever sneaked up on the enemy leader with around 3 or 4 shadows

2. It does add strategy, because the other player must have either patrols or defenses to avoid an invisibility attack. Also, unless you are playing Age of Heroes, you get to see the enemy level their avenger or ghost, meaning you have advance warning of a possible attack. And woses are too slow to use invisibility effectively anyway.

[Dave]

Eventually, we will need a better AI to handle it

Where is this better AI going to come from? Writing an AI with the assumption of complete information is hard enough; writing one that infers things like invisibility is much harder still.

No-one but me has written anything but the most trivial AI changes for Wesnoth, and I think that writing an AI that handles invisibility well is almost infeasible. In fact, I have some ideas for a new, improved AI, that'll handle most abilities, including healing and leadership; but invisibility is an ability it won't handle, because it's just too difficult.

If people who want invisibility to remain do so based on the assumption that we'll one day get an AI that can handle it well, I think they need to re-think things. Such an AI is probably not going to happen.

---

I'm afraid I also don't understand why so many people seem to think fog of war is 'must have'. The realism argument is meaningless to me.

I like fog of war in some contexts: particular if I play a single scenario with human allies vs some AIs, I like it. But some people seem to think it's always good, and it 'adds strategy'. I really don't see how so much -- I think it adds luck more than anything.

A few popular commercial games in the mid-90s used fog, and since then it seems to be stuck in people's head that every such game must have fog, and that having fog is simply better. I think this is as wrong as the notion that every strategy game must have a building component and a fighting component -- a notion taken mostly from the same set of games.

There are different games with different rulesets for a reason. Some people will like a few of the games, some people will like many of them. Some games have luck, some don't. Some have limited knowledge and guesswork, some have complete knowledge. Some are based on building, some on fighting, and some on both.

Wesnoth is a game that is based fundamentally on fighting with no building (except of units), on involving chance to resolve combat, and on mostly complete knowledge. We do provide an option to reduce the complete knowledge, but people should realize that the fundamental design is based on complete knowledge.

David

[Jetrel]

FoW adds one thing to strategy - scouting. You have to scout to determine where your enemy is, and what he is up to. If you see everything, all the time, your scouts are no longer there to gather information, but are instead there purely to secure resources.

Personally, I don't really care either way - wesnoth is a bit slow-paced to use it well, though it can be a lot of fun because it completely changes the play of a level, especially if there is more than one castle per player.


(Correct me if I am wrong on any of the following, as you know better than I)
The primary problem with our AI is that it is tactical, not strategic, at least, not outside of recruiting and village grabbing. Given the basic prerogative of moving toward and engaging enemy units, with a special target of the enemy's leader, establishes a simple-but-effective "strategy" that allows the AI to function.

But you will never see the AI doing stuff like transplanting himself to a better located castle, or using larger-scale flanking tactics, or even planning an assault to land on a specific time-of-day. Or looking at a map and deciding to move the majority of his troops into a certain area with favorable terrain. (In all of this, I mean this on a large scale - the AI in small situations will move onto a better terrain for an attack that turn, and might run from an enemy when weakened, but this is only on a per-turn basis.


The challenge lies in having an AI that establishes "plans" for each unit or for specific enemy units - the AI that would, controlling elves, make a specific point of hunting down that one goblin pillager in order to ensure the safety of his army of woses. That it started hiring, and was planning to hire in order to hold a specific section of forest, allowing it increased income so it could then mount an assault past them into the enemy castle.

-------------

So we could use a better AI, in many ways. And we have a nice "plug-in" system for putting better ones in. They'll come. You don't have to write it - someone else probably will, IF all the other parts are in place.

Greasing the wheels like this is what makes our development flow - I, for example, am perfectly capable of coding my own videogame. But I am here, because here, I don't have to - I can work almost purely on the art, and can learn what systems of animating and constructing sprites work best. (Really, I'm quite happy with what I've learned - I might not be the best artist on the project, but I know I can outdo anyone here at animation - something I had never done at all beforehand).

Programmers are the single thing which open-source has in greatest supply. I'm willing to bet that there's another guy like you, who's ticked at the AI of a game like civilization, and thinks he knows of ways to improve it. Normally, he'd never dream of doing such a thing, but - given the chance to make just an AI for a game like wesnoth, and not to have to fuss with anything else - he might bite the hook.

You know what? We really should advertise that we need such a thing. I'm not sure where - but ... somewhere out there, there has GOT to be a programmer who would want to do that - in fact I'll bet there are people who dream of having a job at a place like blizzard making such things - what better resumé than making an ai for us?

So, it might be incomplete, but let's not burn the near-finished bridge. It would be nice when the day comes that someone finishes it.

--------------

The other chief concern is, that humans have no problems working with invisibility, or fog of war. Some people only play wesnoth for the multiplayer.

We have gotten by just fine with the ai knowing next to nothing about almost every ability. It might make the ai easier to play against, but you can easily handicap yourself, and the game is hard enough in spite of it. We would be better for a smarter AI, but the AI we have is as good as many other, enjoyable games - Blizzard's games used to have some hideous disuse of AI abilities - in fact, even in Warcraft III, some abilities are not used at all by the AI!

ex: "control magic"

And yet, I would consider their removal to be a great detriment to the game - balance comes by making these good for people to use, but leaving the units useful in other ways, and making the abilities such that they cannot be exploited. ex: "Control magic" would allow a player to, at significant cost in mana, turn a summoned unit to their side. What was nice was that the unit capable of it could still live a full life with his other two skills - spell turning, and mana burning. Control magic was also difficult to use, and no summoned unit was powerful enough to utterly ruin ones opponent (except perhaps the infernal, which is immune to the spell anyways).

------

And the last thing is - it isn't really a problem if the ai doesn't know how to use it, due to its rarity. I think that we should not add any new units with invisibility, to preserve this, though.

Most of all, I just really care that we don't blow the opportunity to get an AI which can handle it, by throwing away everything we have for it right now. We have *everything* but the AI. All - ALL of the other pieces are in place. The basic support for it, even if we don't use it on any units, should be kept in the code.

I do also care that one, specific, unit keep it, unless you can come up with something that's a way better fit that it has now: the shadow/nightguant.

I wouldn't mind seeing ambush removed as an ability - giving woses better defense in forest, and giving rangers a skirmish that only works in forest. That would provide (mostly) the best of both worlds. Shadows have a time based invisibility, so you can't park them and keep them invisible forever, removing the most grievous exploit.


I am also curious - very curious, as to what bugs were caused by the invisibility.

[Jetrel]

(I really wouldn't mind seeing FoW go, as I have played with it a whopping... 1 time.) It was interesting, certainly, but the game is a bit slow paced for it.

If you could somehow come up with an ability which lets the shadow "hide at night and blend into the shadows" and yet is not invisibility, feel free to remove invisibility as an ability used on basic units - keep in mind though, it could be great for special instances in campaigns.

I can think of one great campaign scenario idea, where delfador casts a spell on konrad that makes him invisible until he is detected, allowing him to somehow pull a big surprise on an enemy force.



As far as idea-mining for the shadow's replacement ability, I always liked the ability the Myrdraal had in The Wheel of Time - the ability to "teleport" from shadow to shadow. Hideously powerful, sure, but cracking nice.

Perhaps at night, shadows could move 50% faster? Perhaps at night, shadows could have much better defense?

Just please don't pull basic support for invisibility, because I can think of many uses for the ability where the AI's inability to understand it is not an issue - in fact, in the above campaign idea, it is an advantage.

[Dave]

(I really wouldn't mind seeing FoW go, as I have played with it a whopping... 1 time.) It was interesting, certainly, but the game is a bit slow paced for it.

I don't think anyone is suggesting FoW should go. It can really give alot more flavor to the game, and the code has been relatively easy to maintain.

David

[turin]

I don't think anyone is suggesting FoW should go.

EP definitely is. (look at his sig). And i'm considering supporting him... i don't think it adds much that would not be covered by shroud, and there are some problems with it that shroud does not have, namely, you LOSE some information when you move a unit.