"Stealth" or "Guerilla Warfare"

[jamoecw]

 

This is incorrect.  Creatures have their own individual detection radii.

Yeah, to make sure I have this right in my mind:

 

If you have a Rogue with a Stealth of over 9,000, maybe he has a 10ft ehh... detection radius, we'll call it. But, enemies have a perception radius (just as your character has in regard to things trying to hide from him). So, it's still easier for him to sneak past a drowsy Human with a perception radius of 5 ft than it is to sneak past an alert Skuldr with a perception radius of 50 ft.

 

Whereas, even though the drowsy guard's detection radius is only 5 ft, it might be really tough for your unstealthy, plate-clad Fighter to sneak past him if said Fighter has a perception radius of 30 ft, since he's got to stay THAT much farther away from even the guard's tiny perception bubble.

 

Also, since your detection extends out in a radius around you, I'd wager that sound's incorporated into that. Of course, I do hope that things that affect sight and sound can modify the radii, be it lighting, armor material, floor material (grass quieter than cobblestone), etc. I only say that because, technically, all that was mentioned was "based on the character's actions." Though I realize that that was just an example of the fact that it could be affected, and not necessarily a list of all the things that can affect your base detection radius as calculated straight from your Stealth skill.

 

i'm fairly certain that it is just 2 different circles, if they overlap, then you have been detected.  though another way to do it would be if someone enters into either circle, so a sneak walking into a wide perception circle, or a guard walking into a sneak's stealth circle.  there are pros and cons to each, but the former gives more importance to the stealth skill when trying to sneak past guards.

[JFSOCC]

I wonder if silent kills are going to be possible on any but the weakest of opponents, since it's unlikely that you're going to insta-kill any enemy, and equally unlikely that an alterted, engaged enemy won't call for help.

 

Maybe if a rogue deals the first blow, you have 15-30 seconds to kill the guard before he can call for help. It'd be a rogue special ability.

[Lephys]

i'm fairly certain that it is just 2 different circles, if they overlap, then you have been detected.  though another way to do it would be if someone enters into either circle, so a sneak walking into a wide perception circle, or a guard walking into a sneak's stealth circle.  there are pros and cons to each, but the former gives more importance to the stealth skill when trying to sneak past guards.

That's exactly what I was suggesting, though admittedly in perhaps more detail than was necessary. And, to be clear, when they overlap, you're detected, but not quite fully;evidence of your presence is noticed in some way (the detector goes into investigation mode, but not alerted mode). Then, when they overlap enough (and/or if they overlap again with someone who's already investigating), you're fully detected.

 

And yes, I need to work on my wordiness. 8P

[jamoecw]

 

i'm fairly certain that it is just 2 different circles, if they overlap, then you have been detected.  though another way to do it would be if someone enters into either circle, so a sneak walking into a wide perception circle, or a guard walking into a sneak's stealth circle.  there are pros and cons to each, but the former gives more importance to the stealth skill when trying to sneak past guards.

That's exactly what I was suggesting, though admittedly in perhaps more detail than was necessary. And, to be clear, when they overlap, you're detected, but not quite fully;evidence of your presence is noticed in some way (the detector goes into investigation mode, but not alerted mode). Then, when they overlap enough (and/or if they overlap again with someone who's already investigating), you're fully detected.

 

And yes, I need to work on my wordiness. 8P

 

i imagine it would be a time thing, if there is overlap long enough then they will figure out what they're seeing.  that way rogues can still backstab, but they can't sit 5 feet away and lay traps for 5 hours.

[Lephys]

i imagine it would be a time thing, if there is overlap long enough then they will figure out what they're seeing.  that way rogues can still backstab, but they can't sit 5 feet away and lay traps for 5 hours.

Time and possibly also distance. I'd imagine an either/or; your radii overlap for, say, more than 5 seconds, OR, your radius actually overlaps with the target, itself. Whichever comes first.

 

It would be interesting, however, if visual and audible perception were somehow separated. It'd be nice if, for example, you could sneak within 15 feet of a sentry if his back was turned, but couldn't do the same if he was looking straight at you (you're quieter than you are invisible), in some circumstances.

 

Of course, if it's abstracted to just collective circles, that's not the end of the world. I'm sure it'll still be handled well, and make for some interesting effects of stealth on various scenarios. Regarding the waiting 5 feet away thing, though... It makes a lot of sense that your radius might shrink when you're not moving, especially if you're hiding in a corner shadow or something, or within some cover (like a shrubbery). So, in a way, I can see you hiding 5 feet from someone and not being detected. But, I would think most forms of movement/action, at that range, would at LEAST warrant investigation. Unless, of course, they simultaneously result in the death of the potential investigator.

 

I think the overlap of your detection radius with the target, itself, resulting in your detection would be a pretty intuitive use of the radii, though. Josh said that if your circles overlap a bit, you spark investigation but aren't actually detected yet. I wonder how large that buffer is and how that is handled, exactly (like with time/further overlap?).

Edited August 23, 2013 by Lephys

[jamoecw]

the problem with getting within a certain distance for full discovery means no stealth backstabs without magical assistance, i could see that as the radii overlap they increase the detection rate, so that a stealthy rogue can still run up and stab someone without them going on full alert.  though ya, visual cones would help that out considerably and allow for a auto detect at a certain distance.

[Lephys]

^ I was thinking very generally and did not reach the problem with backstabs yet. Point noted. Of course, there's gotta be more to the current system, or it would already fail to allow backstabs, what with radii touching/overlapping equaling investigatory status. Unless something reduced your radius to 0 temporarily to allow for backstabbing.

[JFSOCC]

[Lephys]

 

 

While I very much like the idea, might I ask how this system differentiates between investigation and discovery? It seems like they're either one in the same (a guard will NEVER be prompted to change his behavior and look/move in your direction based on hearing/seeing SOMEthing that he isn't sure yet is you, until you run out of ticks and he just-plain discovers you), or investigation negates the tick-pool system (you get to a certain number of ticks, or you get within a certain range, and the enemy turns to investigate a sound/image, even if you weren't moving or doing anything out of the ordinary... even if you were behind him).

 

Does that make sense? Simple example: If you sneak up behind a guard, then just stand still, you shouldn't actually incur ANY ticks, until he turns around. If you're being perfectly still, you're not generating any sound.

 

Or, is the tick system intended to come into play only once you're being investigated? If that's the case, then the only problem I see is the whole "why am I ticking ever-closer to detection if I'm hidden from view and completely still/silent, but simply happen to be within a certain distance of the target?" when you're behind him or whatnot.

[JFSOCC]

oh that is a flaw in the system I hadn't considered.

 

Perhaps investigation could be based on a counter-skill. If a stealth character enters your sight range, no matter what, the NPC alert state changes one level to "suspicious" or for every 5 ticks spent in NPC observation range, there is a x chance of a change in alert state (base 10%) to suspicious.

the perception/observation/guard skill would add 1% per level.

So base 10% after 5 ticks, 20% after 10 ticks and so on. Whereas a character with 7 levels in perception would have 17% after 5 ticks, 34% after 10 ticks, 51% after 15 ticks and so on.

Obviously if the number of ticks runs out before suspicious state is triggered it would still trigger auto-discovery. No hiding 10 minutes right behind someone.

I suppose "dead zone" could be a special rogue ability where "up close" does become a safe spot as long as the rogue doesn't move, as per his specialisation in stealth.

 

--

Another alternative approach is to have the suspicious state activated when you run out of ticks, instead of auto-discovery. A suspicious NPC will usually (unless he or she has specific orders to hold his or her post.) look about, making discovery likely.

In that case the distant vision cone would also yield instant discovery.

---

 

I sincerely hope that lighting will have influence on npc vision, aiding or foiling stealth attempts.

[Lephys]

Well, it seems to me that the system of depleting ticks serves as a direct (though abstract) representation of the suspicion/investigation process, as fewer ticks brings you closer to absolute detection, and more ticks takes you further from detection. So, it just seems like the only thing that's really missing is something that actually toggles the ticks. In other words, under one set of circumstances, ticks shouldn't even start ticking. Then, under all other circumstances, suspicion is sparked, and ticks begin ticking in their respective areas.

 

Rogue or non-Rogue, if someone has no reason to even suspect that anyone's behind them, then no amount of time spent behind them should bring you any closer to detection, if that makes sense. Same with in front of them. If you're hiding behind a thick shrub, then it shouldn't matter how long they stand there looking that way, or how far into their inner vision cone you're residing, unless they suspect that they saw someone/something, or that someone's possibly hiding in the vicinity.

 

I'll tell you what else might work: two sub-systems of ticks (kind of like what you said about the ticks just leading up to discovery, except without the "vision begets instant discovery" bit). One for suspicion, and another for detection. If you're hiding behind a bush, in the target's sight range (like 10 feet away), then, the longer he's looking around in that direction (being an alert sentry and all), the greater the chance he's going to eventually notice some little inch of you visible through that bush. Obviously the cover would matter here. A wall would completely conceal you, no matter how long you stood there. But, movement within sound-relevant ranges would use up ticks (basically action-based, instead of time-based) towards suspicion.

 

I like the idea of a depletable, finite resource like the ticks much better than a % chance of being detected. A master of stealth, even if seen, is going to warrant suspicion where a nooblet Rogue would've warranted full detection. Shy of a complete lack of stealth-supporting environmental factors, or a blatant lack of active attempt at stealth, I don't see even a 1% chance that a master of stealth is going to just suddenly be completely detected as an intruder, because he's that much quieter and better at concealing himself, etc., under exactly the same circumstances as one of lesser skill.

 

With something like ticks, under bad circumstances, you have a definite smaller window of opportunity (distance to cover, time sneaking, etc.) to stealth around targets, and under good circumstances, you have a larger one. Plus skill: more skill causes the ticks to deplete more slowly, translating into greater sneakability, and less skill depletes them more quickly. A master might be able to move across 20 feet of shadow without even sparking suspicion, while maybe a novice could get about 15-feet before triggering suspicion. So, he could even make it all the way across (because shadow is on his side), but he has to take extra measures to avoid the target who's now making his way toward that area to make sure he didn't see something there, while the master has no such worries. That sort of thing.

 

The simplest system would be something like "when target is suspicious, double the rate at which ticks deplete." But, there's a lot of leeway there. Your "ticks just lead up to suspicion, then after that, you're either seen and detected or not-seen and not-detected" system is a definite option. I just wanted to clarify that I wasn't trying to shoot that down. I'm just sort of getting carried away brainstorming possibilities, hehe. Factors in this are like playdough! I just like seeing what we can make with them.

Edited August 27, 2013 by Lephys

[JFSOCC]

yeah, but when people are suspicious you want their behaviour to reflect it, patrolling, searching, or calling it out to their allies.

[Lephys]

yeah, but when people are suspicious you want their behaviour to reflect it, patrolling, searching, or calling it out to their allies.

Undoubtedly. I had no intention of suggesting otherwise. I only think that, even when someone is already suspicious, there's a range of increased suspicion between "suspicion" and "detection." It's not as if suspicion automatically gives sentries the ability to have their senses override all other factors of noise and concealment, to the point of "Yep, I suspect that I may have heard yet another strange noise, and therefore I now know that there is definitely an intruder and exactly where that intruder currently is!"

 

And I know that wasn't the intention of your idea, so I just wanted to make sure we addressed it, is all. I think having a set of depletable points between suspicion and detection maintains that range of visibility/audibility, while a hard "if you're in this area/radius/cone, you're automatically detected simply because someone is more alert than usual" doesn't really do so. A chance works better, in that respect, but it still provides that really weird "even though this area was pitch black and you're quiet as a mouse, this guy not only glimpsed you, but also glimpsed you so well that he knows you're a person and exactly where you are and will now sound the alarm and attack you" scenario, because chance. Not to mention the times when you have almost no chance of going undetected, and you somehow successfully do. If you're that visible, it makes more sense that you'd simply be easier to detect. Not that a coin toss could still decide that you just don't get seen.

 

I like chance maybe affecting the potency of a sentry's suspicious searching. That might be cool. But, in that, it's not a dice roll between "do factors count for anything" or "don't they and you're simply detected?," but rather the exact effect, within a range, of his searching efforts. Say he rolls a d20 when he's suspicious, for example. Maybe if he rolls below 10, your Detectability ticks, in your given spot/circumstances, decrease at a rate of 3/second instead of 2/second. But, if he rolls above 10, they decrease at 4/second instead of 2/second. That sort of thing. There's no doubt that they're going to decrease, or that they're going to decrease at an increased rate (as compared to his non-suspicion).

[jamoecw]

 

 

 

While I very much like the idea, might I ask how this system differentiates between investigation and discovery? It seems like they're either one in the same (a guard will NEVER be prompted to change his behavior and look/move in your direction based on hearing/seeing SOMEthing that he isn't sure yet is you, until you run out of ticks and he just-plain discovers you), or investigation negates the tick-pool system (you get to a certain number of ticks, or you get within a certain range, and the enemy turns to investigate a sound/image, even if you weren't moving or doing anything out of the ordinary... even if you were behind him).

 

Does that make sense? Simple example: If you sneak up behind a guard, then just stand still, you shouldn't actually incur ANY ticks, until he turns around. If you're being perfectly still, you're not generating any sound.

 

Or, is the tick system intended to come into play only once you're being investigated? If that's the case, then the only problem I see is the whole "why am I ticking ever-closer to detection if I'm hidden from view and completely still/silent, but simply happen to be within a certain distance of the target?" when you're behind him or whatnot.

 

well people give off body heat and increase the moisture in the area near them (they are made of water).  so if you get close to someone just by living you can be giving away your position.  then add in little things like yawning, turning your head, stretching, etc. and you end up causing a change in states which may change your dead zone enough to start detecting someone glued to the back of your head.  one of the roman emperors was paranoid and had polished marble put into most of his construction because it is shiny and people who are sneaking up on him alter the ambient light reflecting off the marble enough to warn him.

 

i could see 1 point for movement each radii, +1 for each of the inner due to ambient factors (so staying still cuts discovery in less than half).

 

as for instant discovery, perhaps instead +24 ticks/sec instead, thus someone not really alert might get jumped from in front if the rogue was super stealthy and quick.

 

also perhaps each radii should be a set distance, and as the perception increases you gain extra radii, so you detect even faster close in at the same scaling (at say whenever you radius increases by the original amount, so if the starting radii is say 30ft for the outer circle and they all expand as you put skill points into perception, when the radii gets to 60ft. then you'd get another inner radii the same size as the original version, that detects at 7 ticks/sec in addition to the others at 1, 3, and 5.).  that way highly perceptive people can't get jumped easily from behind around corners, without making them able to detect things through walls (which would get weird at high levels).

Edited August 28, 2013 by jamoecw

[Lephys]

well people give off body heat and increase the moisture in the area near them (they are made of water).  so if you get close to someone just by living you can be giving away your position.  then add in little things like yawning, turning your head, stretching, etc. and you end up causing a change in states which may change your dead zone enough to start detecting someone glued to the back of your head.  one of the roman emperors was paranoid and had polished marble put into most of his construction because it is shiny and people who are sneaking up on him alter the ambient light reflecting off the marble enough to warn him.

Granted. But, that's a bit closer to someone than I intended. I'm just saying that, for example, your sheer presence shouldn't really alert anyone to anything. I mean, you'd probably agree that if there was a 10-foot-thick solid stone wall between you and an enemy, then he probably would have much difficulty either seeing or hearing ANY indication of your presence, correct? And yet, you'd still be within the radius within which he's capable of detecting things, under normal circumstances. Varying circumstances are represented by changes in the system's factors (movement on loud floor = increased noise radius, etc.), not represented by the base system, itself.

 

It's kind of like line of sight with a ranged weapon. Sure, you can hit something within 30 feet, maybe, but if it's on the other side of a 10-foot-wide oak tree, you have to move before you can hit it. Your radius is simply the range at which you can make an attack, not the total area in which you are always capable of making an attack.

 

So, I just think we need to make sure the system we're sort of brainstorm-assembling here actually logistically allows for base ranges/radii/mechanics, etc, which are then modified by relevant circumstances, whatever they may be.

 

I'd rather it just say "well, he's looking the other way, so he's not going to SEE you right now" than sit there and have to convolutedly calculate how many stealth ticks we need to give a Rogue because he should, feasibly, be able to sneak past this guard behind his back, or exactly how to build levels and position sentries so that the overlapping radii always allow for an appropriate amount of tick-depletion to provide the means to sneak past them (i.e. no guards blocking 5-foot-wide alleyways, because you'd ALWAYS have to sneak so close to them that your potential for detection would spike crazily.

 

*shrug*. It just seems a lot easier, for balancing and tweaking and such, to just account for factors that wouldn't actually contribute to your detection at all.

 

For what it's worth, I'm all for little things like the emperor example; if you're in a polished marble room, maybe the radius in which you begin slipping towards detection/investigation is greatly increased. I would expect lighting and fog and such to affect this, so why not room-surface reflectivity? At least, when it's an extreme difference like that (super shiny marble versus regular stone). And, same with sound; sneaking on cobblestones should be much noisier than sneaking on dirt. So, it makes sense that there should be times when sneaking on dirt -- between the quietness of the dirt and your at-least-decent ability to sneak -- that your sound radius is so small that you could sneak within 5-10 feet of someone who couldn't see you at the time without audibly causing any alert whatsoever (or even any risk of it, no matter how many times you tromp back and forth on that dirt), while sneaking on a different ground/floor material would lead to at least potential suspicion. As opposed to "doesn't matter 'cause you're so close to the person, your stealth ticks are going to deplete at SOME rate, even though you're pretty much making no noise."

 

Another thing I just realized (at least, as a collective thought), is that your radius could simply represent noise, while the sentry's radius could represent sight. So, if the sentry touches your circle, he's heard you, and if you touch his circle, he sees you. Fog/lighting could affect his circle (shadows could cut off a 30ft vision cone/radius at the 15ft mark, etc.), and floor material + boot material + Sneak skill, etc, could affect the size of your noise circle (walking on grass puts out a 5 foot circle, while walking on marble floor puts out a 20ft circle, etc.).

 

Then, your Sneak skill (maybe coupled with the sentry's Perception skill? Kind of like your to-hit chance ranges being decided by attack-vs-defense) could affect two main things: the rate at which your points deplete towards detection, and the dividing line between suspicion/detection zones in a given sentry's radius. Just like in JFSOCC's example, you'd have basically (at the very least) an outer zone and inner zone. In the outer zone, only your Suspicion points("ticks") deplete. If they're gone, the sentry notices SOMEthing, but doesn't actually know what, and investigates. At this point, even the outer zone will begin depleting actual detection points. The inner zone will immediately spark suspicion and begin depleting detection points.

 

As for the effects of your Sneak (and possibly the sentry's Perception) on the zones, if a guard has awesome perception, and you suck at sneaking, then maybe his ENTIRE perception range is one big detection (inner) zone, and the suspicion (outer) zone has been snuffed out of existence by the huge different in your skills. Or, on the other extreme, you are a master Sneakist, and the sentry is a militia farm boy, then maybe the divider gets pushed the other direction, and his entire radius is one big suspicion (outer) zone. (Or, most of it... I think both zones should maybe be maintained, even if one is quite small). If your Sneak and his Perception are evenly matched, then the divider will be at the halfway mark; the outer half of the radius will threaten suspicion, and the inner half of the radius will threaten full detection.

 

Same goes for enemy Perception vs. your Sneak for your noise radius. It's almost the exact opposite, functionally. If his Perception is really high, and your Sneak is teh suck, then your noise radius will be mostly detection and barely any suspicion. The only difference here is that your points will deplete when he is in your circle, and not when you are in his (as with sight). If his Perception sucks, and your Sneak is awesome, then most of your circle will merely threaten suspicion, instead of detection. And if they're even, 50/50.

 

The way this seems to have value, to me, is that visual and audible detection will vary under different circumstances. If it's super dark, but you're on cobblestones, then your noise radius will be the major concern, since it will be something like 30 feet and the sentries' sight radius will be maybe 15 or so. So, you want to be more quiet than you do hidden. But, if it's bright outside and there's little cover, and the ground's made of soft grass, your noise radius will be MUCH smaller than the sentry's sight radius (it's a LOT easier to see someone in broad daylight, even at 50 feet, than it is to hear them at that distance, unless they're being VERY loud). But, even with the circumstantial difference in detection radii, your skill still comes into play to further effect your radius's relation to detection chance/rate.

 

For example, a novice Sneaker might still have trouble sneaking on quiet ground past a sentry 20 feet away if his noise radius is 15 feet, but most of it is Detection zone (automatic investigation, begin depletion of detection points towards full detection). Whereas, a pro sneaker, under the exact same circumstances, might easily walk on even louder ground (giving him a 30ft circle instead of a 20ft one) because so much of his noise radius is merely Suspicion (outer) zone, and his points deplete more slowly on top of that. If that makes any sense.

 

They say a picture is worth a thousand words... Maybe I should just draw a couple of pictures. 8P

[jamoecw]

http://www.youtube.com/watch?v=AukXagFIEws

 

when thinking of systems, try to make these fantastic, yet non magical, methods possible.

 

i especially like the last geisha one.  though given enough time, i'm sure she would have been noticed, in fact if all of the cutscenes were n the video you would have seen the ones where the ninja/general did notice and killed her.

 

also keep in mind that different creatures have different primary senses, so just having vision on the perceiver would take into account dogs and such.  creatures that glow would have a visual detection radii larger than their body as well.

 

i'm under the impression that detection is at least partially rogue-like (i think that is the current buzz word for dynamic FoW), i also figured that the sneaker would have a small 'noise radii' (not just sound, but general perceptual stuff) and that when it would overlap with the detection radii  you'd start collecting ticks, once full the perceiver would have enough info to become alert, that delay from full suspicion to alert would vary based on the perceiver's wisdom (or whatever) and different things that confuse the situation (like the rogue playing dead, or the rogue being some creature that the perceiver doesn't register as a creature).

 

as for building levels, keep in mind that things aren't static, you can (and should) do things to alter the detection grid the guards create with their perception radii.  on thing you could do is make a guard ai that moves towards areas of low perception in order to investigate thoroughly, and another that shies away from such places, as they are dangerous, then select the appropriate one for the guard.  so if you use some water arrows to take out a torch, the guards will move to either investigate or keep themselves from getting stealth killed, which alters the detection grid, and if it was a carefully calculated high efficiency grid then there is now an opening (hopefully as the player you made the opening where you need it).

[Lephys]

Yeah. I apologize for the length of my last post, heh. The only thing (and it's a bit minor) that bothers me about both sound AND sight being handled by the same radius is that you can be super quiet while being extremely visible, or you can be invisible and extremely loud. So, I dunno... I think Josh mentioned that some things will see better, and some things will hear better, and that will be a functional difference, so I'm curious as to how they're handling that.

 

Maybe the best way to do it is simply to have a separate visibility radius and audibility radius for each character/creature? That way, if you have a torch (or glow, as per your example), your VISION circle grows, but you could still be making very little sound, so your noise circle stays the same size. So, you could, even with a torch, sneak past something with terrible eyesight and decent hearing. Or, you could sneak past something that's deaf, but has excellent vision by sticking to the darkness, even if you're being quite noisy.

 

In other words, if each person just has one radius, and their overlap prompts investigation, then how do you let something's hearing be less effective than something's vision? If your circle represents both things, then how do you know whether it's your noise or your visibility that triggered? And how to you represent the difference?

 

Another thing I'm curious about: Will the system work the same way with you detecting [/i] enemies? It'd be really cool if you could HEAR an enemy, but couldn't SEE it, and the system somehow indicated that to you differently than seeing something without hearing it. *shrug*. Also, do you have a "this is how far out I can be seen" ring AND a "this is how far out I can see" ring? Because, that could get particularly confusing, too. If you're super stealthy, but you have awesome eyesight, wouldn't your "this is how far I can see" radius overlap an enemy's "this is how hidden I am" radius at a much greater distance, while your "this is how hidden I am" radius wouldn't be anywhere NEAR their "this is how far I can see" radius?

 

Also, as far as distractions and alterations to the stealth grid, I'm very much in favor of that stuff. But, despite that stuff, a Stealth skill of 100 should allow for sneaking through much tighter spots than a skill of 5, even without relocating guards and altering the grid. Otherwise, the only thing Stealth would boost is your ability to throw rocks and douse torches.

[JFSOCC]

Also, as far as distractions and alterations to the stealth grid, I'm very much in favor of that stuff. But, despite that stuff, a Stealth skill of 100 should allow for sneaking through much tighter spots than a skill of 5, even without relocating guards and altering the grid. Otherwise, the only thing Stealth would boost is your ability to throw rocks and douse torches.

I like the idea that skill in stealth allows a skilled player more options, but doesn't make stealth an automatic overpowered solution.

on the flip side, even a non-stealthy character can sneak past a guard, he or she just has less time to do so, and needs to stay a greater distance away. It allows stealthy characters to shine while not making life impossible for the army boots characters.

 

It's obviously the bare bones of an idea, a concept.

you can work with environmental adjustments, creaking floorboards and stairs, loud traps, distractions, or apparently polished marble floors.

you could also have different means of detection for different senses, dogs using an abstraction for scent: the ability to see where a player/creature has tread their feet, without giving you information about that creature, a path which after enough time clears up. (x amount of time that the 'smell' lingers).

perhaps ciphers, as it seems to be part of their core concept, could see what has been touched or used by their quarry with the right soul abilities, useful for tracking, but also for seeing solutions to answers, or traps, or finding out what happened with little Timmy's corpse. (items used recently might light up for the cipher while the ability is active)

 

Josh mentioned in a previous thread that stealth was intended to be more than just a dice roll like in older IE games, but not as big as for a stealth game (which makes sense) so perhaps we're going overboard.

It's just, I don't think this would be so hard to code.\

but then I have no idea really.

[Lephys]

I like the idea that skill in stealth allows a skilled player more options, but doesn't make stealth an automatic overpowered solution.

on the flip side, even a non-stealthy character can sneak past a guard, he or she just has less time to do so, and needs to stay a greater distance away. It allows stealthy characters to shine while not making life impossible for the army boots characters.

Past "a" guard, yes. Past any given guard ever? No. That's all I'm getting at. If you take a random guard, at some random location in the game (in a scenario in which sneaking is actually beneficial), then a master of Stealth should always have an easier time of things than a Novice. Throw in the range of base difficulties for sneaking past guards, and the trickiest scenarios are going to be impossible for the Novice, and still-tricky for the Master, while the base simplest ones are going to be tricky for the Novice, and easy-as-pie for the Master. It's not that being a Master of Stealth should make sneaking flat-out easy, just that it should make things relatively easier, across the board, than a lack of Stealth skill.

 

When the player's skill at distracting guards and timing things completely overrides the Sneak skill value, you've obviously gone too far, for example.

 

It's obviously the bare bones of an idea, a concept.

 

...

 

Josh mentioned in a previous thread that stealth was intended to be more than just a dice roll like in older IE games, but not as big as for a stealth game (which makes sense) so perhaps we're going overboard.

It's just, I don't think this would be so hard to code.\

but then I have no idea really.

I'm not at all trying to criticize it as anything more than a draft. And I'm not trying to bash it. My only intention is constructive. I just enjoy analyzing systems like this, and I'm wondering how to handle all the factors, is all.

 

Mainly, at this point, I'm wondering if representing, say, your character's hearing range AND the amount of sound he's generating ,simultaneously, absolutely requires multiple rings/ranges, or if there's some cleverer way of doing it. I think simply not-representing that is possibly a very bad idea. Vision and sound are your two biggest tools in providing situational benefits and detriments (variance) to sneaking, which is the easiest way to make the Skill value not just an "I win" binary switch threshold value.

[Osvir]

EDIT: Just throwing out some ideas, thoughts and stuff I Think should be considered /EDIT

Don't forget Prone Stance and the possibilities it gives (If it is in the game, it was advertised by Tim Cain pretty early in the Kickstarter Campaign)

1. Crawl under a house to access it from below (if a house is slightly elevated and/or has a hatch door under the house somehow)
2. Crawl into a crack in a wall to get into the fort grounds (Friendly Arms Inn for instance, could it be possible to crawl into the area and pass by the guards at the gate?)
3. Crawl into a Cave tunnel.
4. Prone = Swim = Same-same animation possibilities (but would require some extra Resources to be spent on water coding and scripting and "swim paths"/water pathfinding etc. etc.)
5. River/Lake swimming to get past encounters.
6. General Secret finding (Dead-ends but might found some lost dead skeleton adventurer holding some semi-useful loot). Also secret paths.
7. Crawl in bush, hide from patrols.
8a. Prone specific take-downs?

8b. Must be in Prone Stance:

- No animation abilities, but hiding behind an enemy/character/unit in Prone Stance and using a "Prone Take Down" could be an equivalent of a "Stealth Take Down" in many "Stealth Games" (Deus Ex HR, Dishonored etc. etc.).
- No animation, as said above, but could give status effects such as "Sleep", "Stun" or whatnot.

Can you move objects and possibly bodies in PE? (Divine Divinity-esque, you can move objects by clicking and dragging them). Does patrols, guards, factions notice bodies laying around? Can you pick up bodies? Does patrols, guards, factions notice dropped weapons and items? (Often times they do not, they get alarmed and hostile if they see a body, but they don't give a damn if you move the body but the weapon the knocked out guy was holding is in plain open field).

^I have more to say, but this will do for now I suppose. All things, in my opinion, to consider for a stealthy approach and how AI reacts to a stealthy-esque-ish approach.
 

Edited August 31, 2013 by Osvir

[Lephys]

I do like the potential for prone/crouching (at least in functionality), as it's silly when, in a game, there's 4 feet of cover, and the game simply says "well, that doesn't fully cover you, so you can't actually sneak here; you'll just be seen." And while you always bring a smorgasbord of ideas to the table, Osvir, and I think it's fantastic that you're able to like SUPER-brain-storm like that, I fear that quite so much detail in merely the ability to basically make yourself lower profile while moving might be a bit heavy-handed. Especially if we get all manner of various stealth-system tidbits in P:E.

 

I do, also, hope for stealth-supporting weapons, like blackjacks and such, for things like "takedowns." Maybe targets have different amounts of non-lethal "HP," and/or non-lethal damage bypasses armor and such, so a "backstab" with something like a blackjack could result in instant unconsciousness. I guess in the scheme of the P:E Health/Stamina system, such attacks could simply deal oodles of Stamina damage, but very little Health damage. And maybe you could simply toggle the attack mode.

 

Not that non-lethal takedowns are the only thing you'd do in stealth. You'd obviously make silent, efficient kills, as well. But, non-lethal stuff is typically more prevalent in sneakery than in flat-out direct confrontation. You're usually trying to get somewhere, past some sentries and such, but you don't really have anything directly against the sentries. They're just doing their job. So you try to avoid them, but if you can't, you render them a non-issue, temporarily.

[Osvir]

I do like the potential for prone/crouching (at least in functionality), as it's silly when, in a game, there's 4 feet of cover, and the game simply says "well, that doesn't fully cover you, so you can't actually sneak here; you'll just be seen." And while you always bring a smorgasbord of ideas to the table, Osvir, and I think it's fantastic that you're able to like SUPER-brain-storm like that, I fear that quite so much detail in merely the ability to basically make yourself lower profile while moving might be a bit heavy-handed. Especially if we get all manner of various stealth-system tidbits in P:E.

No crouching.

 

As for cover and such. I believe PE will have props (bushes and such), perhaps making bushes and other "cover" be "transparent" or passable whilst in Prone-Mode and give them parameters and triggers/scripts/code that they "aid" stealth (whilst in Prone-Mode).

 

IF [Character Prone] = [bush#1] can be moved through & [stealth] (or something)

 

It's not entirely "low profile" either. But it's also ideas for allowing the Player to enter 1 room from several different "entrances". The existence of having a [Prone] stance allows more "doors" (figure of speech) to be "opened" and "entered" through.

 

Potential entrances to 1 room.

 

Standing+1 Room/House:

- 1 Room/House

- 1 Door entrance (1 Door)

- Window Entrance? (1 Door)

- Climb roof+Chimney Entrance? (1 Door)

- Backdoor? (1 door)

- Total of 4 "doors"

 

+Prone

- Cellar entrance? (1 door)

- Enter house from a nearby Cave->Tunnel (1 door)

- River+Watermill Entrance (swimming, 1 door)

- Total of 7 "Doors" suddenly

 

Prone = Allows more entrances.

 

 

[Lephys]

No crouching? Why doesn't crouching get any love?

 

Also, I think prone and/or crouch (or maybe not crouch, if it's really that bad? hehe) could easily be contextual actions. In other words, you can't just click "prone" and then move around, and click "stand" and move around some more, manually. But, when in toggled Sneak mode, and/or Hide mode, and near appropriate cover (shrubs, low wall, tall grass, a ditch, etc.), your characters would have the opportunity to move in prone/crouch. Maybe you could have people prone-hide in tall grass as a set up for an ambush? In a related note, I think climb is under-used in CRPGs, and an archer should be able to (at least in certain areas/circumstances) climb a tree and go equally as undetected as the person in the grass. Maybe... Ideas ideas... 8P

 

Back to crouch, it would be pretty cool if, say, an archer would, in the midst of combat, contextually kneel when you place him near a big stump or a low stone wall or something, and fire from there, with some kind of environmental bonus. Cover, pretty much. But, I wouldn't want it to be a super-active "take cover" system, like in military-style tactics-RPGs. Just... a very fluid, passive, contextual bonus with accompanying "here's why you're getting a bonus" visuals. And it would just be simple things, like kneeling when you happen to be placed beside something you could fire a bow over. Heck... murder holes might even be used, in towers and keeps and such. Just a simplistic tactical factor. No levels entirely designed for everyone to just take cover and have one big fire fight. The real-time combat, especially with magic and swords and such, simply isn't set up for that.

 

The main thing I'd want cover to affect would be visibility/stealth.

[Osvir]

No crouching? Why doesn't crouching get any love?

 

It's not that crouching is bad per say, but Obsidian has stated (I'm too lazy to find it) that crouching is not feasible. I would personally love to get crouching, but I'm just following the official word (improper sourcing for the laziness! )

 

It's the Tim Cain QA video I believe, update... #16 or something like that.. early early stuff.

 

EDIT: As for your climb idea, sounds great! Some Tree props could possibly have some sort of utility like that. No "climb animation" but more like "You will now climb this tree" *pop* "You have climbed this tree"~ imaginative climbing.

 

Perhaps doesn't even need an archer artwork if climbing a tree either. Oooh I like this...

 

The tree prop could get the "selection circle" and visually it could be a tree shooting arrows. If an enemy would hit the Archer/Ranger/Bow-User in the tree (practically & visually an enemy shoots the Tree) you could "fall down" (visually spawn at the bottom of tree with knockback-effect) from the tree (high-risk/high-chance of doing so) and take extra damage.

 

[Clean up list]:

- *pop* Climb tree: "Thou hast become a tree" = The character model disappears

- Tree gets selection circle and gets "locked" range modifiers (A tree can not move)*

- Enemy shoots at tree

- Character "falls" out of tree = Character model spawns beside tree and falls to the ground a la "Knockdown"

- Fin

 

It also gets you on that "1 more door" boat.

 

1 House, 1 tree next to house, *pop* climb tree and *pop* jump into window.

 

EDIT EDIT:

 

More stealth stuff~

 

Faction Clothing (Fallout: New Vegas inspired). Take down a Guard, take his clothes and/or armor/banner stuff and get through the gate and similar shtuff.

 

EDIT EDIT EDIT:

Added [Clean-up list]

* Can a tree move if a Druid makes it so?? :D

Edited August 31, 2013 by Osvir

[Lephys]

Yeah, it could be as simple as some trees being more easily climbable. Just like if you were to go into a forest right now, you might find SOME trees with lower limbs than others, that would be a lot easier to climb (without some kind of climbing spurs/axes/equipment; I'm just talking hands-and-feet agility/strength climbing here... simple stuff). Instead of, you know, every tree in existence having various climb checks or something. Could also possibly be used for scouting. Climb check gets you up into the tree, thus grants a greater sight range.

 

It would really only be good for scouting/hiding/ambushes, and not so much for cover. Even though a tree MIGHT provide difficulty in hitting you, with lots of limbs in the way and such, you're also a lot less capable of moving/dodging when up in a tree, especially holding yourself stable WHILE firing a bow. I would think that once people spotted you and decided to target you, the tree would lose most of its usefulness. *shrug*

 

Annnnywho, . As for faction clothing, YES!. But, even sort of along the lines of "diguises" (appearing to be a specific someone else) could be simply identity concealment. Maybe even if you're spotted by some people while being sneaky in some situation, when you bump into them later, in a non-sneaky setting, they're just all "Hey, you should be careful... there's sneaky folk afoot! o_O," having no clue that it was you. *shrug*

 

Again, that's kind of something that may not be super feasible to implement, depending on how everything else works/is designed. It could, though, be as simple as clothing/equipment that conceals you. Maybe even magically/supernaturally. That could even be a use for a potion. A sort of doppleganger potion, a la Harry Potter and various other fictions. You've got to get something from the target you wish to take the form of, so you have to subdue them in close range and "activate" the potion. Then, you drink it, and *poof*, everyone thinks you're another person for the next 30 minutes. Might even make for a fun "did you do your homework?" sequence, where you had the opportunity to learn a lot of various details about the target beforehand, and you must then walk the walk, so to speak, or you'll still arouse suspicion ("Heyyy... Herbert the Guard is ALLERGIC to cats! He wouldn't be petting one like that! IMPOSTOR!").