Josh Sawyer talks about stealth mechanics

[Nonek]

Rogues were introduced in 2nd edition AD&D weren't they? Or was that the group term for both Thieves and Bards, I forget.

[IndiraLightfoot]

Here's me hoping that Josh pops in and enlightens us on this greatly appreciated feature! 

[WorstUsernameEver]

It's funny that people say that. The reason why AD&D Thieves (remember when they were called Thieves?) were renamed to Rogues in 3rd Edition is BECAUSE they were becoming a bit "Fightery" with their non-stealth Sneak Attack. But all that did was train people to think that Rogues and Thieves were the same thing.

It doesn't really help that in some translated D&D manuals (the Italian ones, at the very least) they're still referred to as thieves.

[Infinitron]

Or was that the group term for both Thieves and Bards, I forget.

 

Yeah.

[Pipyui]

I think I like the relative perception circles better than absolute derived from stealth/perception scores.  It could mean the difference between playing this game

 

and this game

 

 

 

It's a terrible example, I know - I just want to illustrate the point.

[SunBroSolaire]

So your stealth radius will be different in every area, in a totally unpredictable way?

 

 

Just give enemies and players more circles to work with. Characters should have a very large outer radius that never changes sizes - when this intersects with an enemy's perception radius, player stealth is compared to something like 25% enemy's stealth. Then there should two inner radii which compare player stealth to 50%, 75% the enemy's stealth and 100% if the enemy's detection circle touches the player themselves. No need to have per-area craziness, no gimping high level enemies with averages, no problems with scaling. You're welcome Obsidian.

[ItinerantNomad]

I think I like the relative perception circles better than absolute derived from stealth/perception scores.  It could mean the difference between playing this game

absolute.jpg

 

and this game

 

relative.jpg

 

 

It's a terrible example, I know - I just want to illustrate the point.

 

BOOBIES!!!

 

Seriously though, I think Josh should reconsider the light/dark mechanic proposed.

 

Instead of increasing or decreasing the circle sizes, perhaps the light/dark in each area can adjust the rate at which circle sizes increase or decrease (from movement to standing still). It would be nice if it would take longer for the circle to decrease in size (regardless of how little the circle size changes) in light and it would occur faster in dark. That way, the devs don't have to worry about players worrying about too many things at once and it plays as a sort of bonus for those who do. If they are moving in the dark though, their circle would just be as large as if they were moving in the light and at the time they stand still the final circle size is the same in light and dark - it's just a matter of how quickly it gets there.

[teknoman2]

he did not answer a question though... do the enemies start looking if your outer circle touches theirs, or if your character goes into their circle?

[JFSOCC]

Your stealth circle allows you to enter his perception circle.

http://forums.obsidian.net/topic/64268-stealth-or-guerilla-warfare/

http://forums.obsidian.net/topic/63102-commandos-a-stealth-perspective/?p=1298295

Edited January 17, 2014 by JFSOCC

[Lephys]

I'm so confused now... I humbly request an official visual aid. Of course, I will calmly wait until the official Stealth Mechanics bi-weekly update. 8P I'm still unclear on how our circles and enemy circles interact. I would think it'd be easiest to just have the character avatar touching/crossing the circle be the trigger. And why shouldn't certain enemies be "trying to" sneak up on us? (Not talking about super-elaborate AI that is launching covert operations on you from across the map or anything). Then, based on YOUR perception, you can't detect stealthy enemies unless they cross into YOUR circle/cone/whatever.

 

Except... Ohhhh, I think I get it. You'd have no way of applying the Stealth Skill bonus without shrinking the opponent's perception circle. So, if you have one character with 1 Stealth, and another with 50, the one with 50 is going to have a "presence" circle that's much bigger than themselves, so that, even though they're still 10 meters outside the enemy's perception circle, they're still being detected. While the 50 Stealth character will have a much smaller "presence" circle, thus allowing them to sneak even closer to the same enemy with the same-sized perception circle.

 

Still, though, I would very much think that being able to be detected farther away than closer up would only apply to audibility. At least in terms of skill. The not-being-seen aspect is all about camouflage, and movement timing (while the enemy's facing away, sticking to shadows, etc.). You can't just will yourself to be less visible than someone else. You're either within a foe's sight range and distinct from your environment, or your aren't. Sound, though... You could be louder than someone else.

 

All I can say is, I hope that both visibility AND audibility are distinctly represented, however it is done. I'm hardly demanding it, but would very much appreciate directional cones for vision, and radii for sound. It doesn't have to be perfectly simulated or anything, but, it just makes a lot more sense. Factors. If you're sneaking on moss or soft dirt, you're going to be a lot quieter than if you're sneaking on cobblestone, no matter your skill. But, you're going to be JUST as easily seen if you're out in the open. And if it's pitch black, then even if you're heard, if the enemy can't SEE you, then all they're going to do is investigate. And everyone's sight range would just be adjusted according to general light level (again, doesn't have to be perfect, like some amount of adjustment per virtual lumen or something).

 

*shrug*. I know that's a lot, doing sound and visuals separately. That's at least 2 visually-represented-to-the-player areas for each and every character/entity on the screen, when you're in Scouting mode. But... I dunno. If something could be worked out, I think that would be a lot better.

 

Also, what about the old "throw a pebble or small object at that wall over there so as to lure the sentry into an investigation where I'm not" trick? I'd LOVE to see strategic use of that. Maybe not even tied to the Stealth skill? Maybe a Throwing skill? Plus, some enemies would be too smart to fall for that.

[Jobby]

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Surely the enemies perception should define the size of their circle and your stealth skill should define the size of yours, if they overlap "alert", and then another hidden circle also defined by their perception could define "caught in the act".

 

Perhaps the scaling mechanic is just to maintain some kind of "circle consistency" throughout the game in which case I'm happy, someone else spoke about averaging "perception circle size" for all enemies on A level, this really confused me and sounds friggin awful, I hope that was a misunderstanding on their part (or my failure to read their post properly).

 

Regarding light and dark I don't know how hard it would be to implement in Unity, they did mention tools they were using for dynamic lighting in a previous update. If said dynamic lighting is more than just cosmetic and actually provides a light value for each node I wouldn't image it would be all that difficult to scale the stealth skill to said light node value but I really don't know how the engine works and I certainly ain't no programmer lol.

 

Even if it was easy to do so that may throw up problems with game balance, however the positives would certainly outweigh the negatives, in regards to tactics, gameplay choice, character viability etc... 

 

And yes distractions sound like fun, I can image throwing a rock near a powerful mage and him instantly casting "clarvoyance" then owning me for my silly tactics lol.

Edited January 18, 2014 by Jobby

[Ffordesoon]

Well, you can backstab, but you won't be stealthed while you do it.

 

I suppose they could implement some sort of short detection delay so you can try to run in for the stab before you turn visible.

Yeah, my thought was, "Everyone can be sneaky, but only the Rogue has specific moves designed to take advantage of sneakiness." That would make more sense than only Rogues being able to stealth and/or everyone being able to do Rogue stuff.

[Lephys]

For what it's worth, I would love to see something beyond just a "start combat but I'm really close to you now and get the first attack, MUAHAHAHA!" option. It's hardly integral for the game to not be bad, but it'd be very nice to have a "*grab them from behind and put a knife blade to their throat* Make a sound and you'll drown in your own blood!" option, even if it doesn't initiate any complex dialogue with the majority of intelligent combatants. Some of them might even cry out anyway, so you'd only have a good reason to use it in specific circumstances. Maybe you take someone hostage to escape a sticky situation? No one's going to shoot arrows at someone holding a knife to the baron's neck, right? So you take him out to the edge of the wall, then let him go once you're out of bow range. Maybe they'll chase you down, but you didn't have to fight your way out, at least, etc.

 

I'm even fine with it only being available in situations where it would be significant, even though that's abstract. I realize that Josh said that, even when you're detected, in the stealth mechanics, the detector will either initiate dialogue or attack. I just hope those are our options, as well, from the sneakee's perspective. But, you know... with some kind of incentive for the other person to not make a sound, rather than "Hi there! I've snuck up on you, but I'd rather talk, so I'm just going to start talking to you, even though you're just going to flinchingly call the rest of the guards in a fit of surprise and duty fulfillment! 8D!"

 

And yeah, Jobby. Don't get me wrong. I mostly analyze things in isolation, just 'cause I can't help myself. I fully understand that it might be REALLY tricky, code-wise, to actually facilitate in-game. I'm not trying to demand it be done, but I just want to emphasize what I see to be the major benefits of actually representing these things in SOME way. What with the limited knowledge I have, there might even be a simpler/less-tricky way to implement these things, so I'm just trying to come up with what I can, in case it's useful. And, at the very least, as mere feedback, in case a lot of other backers feel the same way about it. *shrug*

[ZornWO]

The system sounds fun.  It might be a little gamey to have your own circles depend on an average of other creatures in the level (the thing I dislike about it, if I understand it correctly, is that it gives you a metagamey hint about the kind of foes you face on the level even before you've scouted).  If they make no changes to it though, it'd still be fun I expect.

 

I'm wondering though abt something else - what are the benefits to scouting ahead?  Is it just to avoid combat, position yourself for combat you do have, and eavesdrop?

 

In the IE games, it was fun not to prebuff unless you reconnoitered the area to find the enemy.  With the SCS(2) mod, the enemies started buffed too, so it made it a level playing field - one you had to earn.  It's not a dealbreaker (particularly w/ the surprising eavesdrop mechanism), but it'd be great to have a game that had that, without it being a purely self-imposed restriction.  Since PoE buffs are combat-only spells, the question's really just - when does the combat condition start, when you see an enemy or when you both see each other?  It's related to Lysen's issue regarding backstabbing - can you make one initial combat move before the enemy's alert to your presence?

[Wombat]

As for possible scaling, considering various things, I personally think reasonable level-scaling is for plot-required areas make sense (especially in a system, where the strengths of characters are roughly majored by levels).  However, the topic is rather touchy and the word "level-scaling" could give a wrong impression, ending up with inefficient inter-desinger-backer communication.  Also, I think detailed balance-out task should come after, at least, basic features are on the table.  So, I guess Sawyer is not prepared to talk about it yet both socially and technically.  At least, considering the info revealed so far, I think Sawyer deserves some trust, along with his past records, in terms of his history of both released games and communication with his players.

 

 

All I can say is, I hope that both visibility AND audibility are distinctly represented, however it is done. I'm hardly demanding it, but would very much appreciate directional cones for vision, and radii for sound. It doesn't have to be perfectly simulated or anything, but, it just makes a lot more sense. Factors. If you're sneaking on moss or soft dirt, you're going to be a lot quieter than if you're sneaking on cobblestone, no matter your skill. But, you're going to be JUST as easily seen if you're out in the open. And if it's pitch black, then even if you're heard, if the enemy can't SEE you, then all they're going to do is investigate. And everyone's sight range would just be adjusted according to general light level (again, doesn't have to be perfect, like some amount of adjustment per virtual lumen or something).

I wonder if there is a differenciation among creatures which are perticulary sensitive to specific information such as sound, vision and ordor.  There is already a creature, which is probably almost impossible to avoid by stealth skill, though.

 

Update by Adam Brennecke, Executive Producer and Lead Programmer
 
Skuldr have poor vision, but they use a form of echolocation to perceive the spirit world. This allows them to “see” souls, making it difficult to use stealth to avoid them.

 

 

I'm wondering though abt something else - what are the benefits to scouting ahead?  Is it just to avoid combat, position yourself for combat you do have, and eavesdrop?

I thnik scouting/occational looting is good enough for all the maps but, depending on the area designs, there can be a chance of subotarging (disabling alarms and/or other defense systems, oppening doors for other party members, maybe setting traps or make use of existing ones/environments, and talking to possible friendly NPCs bypathing guards).

Edited January 18, 2014 by Wombat

[TrashMan]

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Instead of circle, use half-cirlces. That leaves a blind spot for vidual detection.

 

Audio detection is another matter.

[anameforobsidian]

I can see the utility of a circle that bulges out

 

 

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Instead of circle, use half-cirlces. That leaves a blind spot for vidual detection.

 

Audio detection is another matter.

 

 

I can see the utility of something that bulges out, perhaps like this:

Red is noticed you, yellow investigating.  It's circular near the human since that's affected more by hearing, and vision extends conically from the entity.  You could even have a formula which made the yellow decrease faster with sneak ability than red.  

 

For ninja death attacks, you could have characters dash in with sneak attack ready before their target raises the alarm.  If they succeed and kill the mob, the dying character makes a ping, which hits the other detection shapes and interacts normally.  That way it isn't DX:HR levels of ridiculousness where you can muffle guys from their buddies five feet away, but it still allows the option in the gameplay.

 

The problem is that you would then have to make / test / debug several different types of perception cones.  While the first models human perception, this might model skulldr perception:

 

Eventually that becomes a lot to implement, test, debug, etc.  They could do archetypes, human, beast, dragon, undead, etc.  But that still turns into a substantial amount of work for a less important subsystem.  That would be an amazing job for a modder though (although this is going to be pretty hard to mod from what I understand).  Or, it's work for obsidian to do in a game based on this technology.  The great thing about Project Eternity is that it's just the first, and other games can build substantially on its capabilities.

Edited January 18, 2014 by anameforobsidian

[Jobby]

 

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Instead of circle, use half-cirlces. That leaves a blind spot for vidual detection.

 

Audio detection is another matter.

 

And in which direction would the half-circles face? That was the difficulty he was mentioning in regards to line of sight as far as i could tell. 

 

Don't get me wrong line of sight would be a great mechanic but I think the circles are a huge improvement over I.E. systems and in my mind they represent audio and visual in an abstract way (the semi-circle behind the facing direction of the enemy could be considered audio and the semi-circle in the facing direction could be considered visual).

 

It's not perfect but its certainly not bad.

[Wombat]

Yeah, while it could be ineteresting to have independent zones for vision, sound, scent and magic (soul power like in the case of Skuldr), I guess it would be overkill for the scope of the game...

[anameforobsidian]

 

 

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Instead of circle, use half-cirlces. That leaves a blind spot for vidual detection.

 

Audio detection is another matter.

 

And in which direction would the half-circles face? That was the difficulty he was mentioning in regards to line of sight as far as i could tell. 

 

Don't get me wrong line of sight would be a great mechanic but I think the circles are a huge improvement over I.E. systems and in my mind they represent audio and visual in an abstract way (the semi-circle behind the facing direction of the enemy could be considered audio and the semi-circle in the facing direction could be considered visual).

 

It's not perfect but its certainly not bad.

 

 

It's a matter of interpretation, but I read that as him saying that it was too complex for the player to deal with six separate entities and however many guards.  Then there's the weird problem of mixing 2D and 3D, which would make lighting a difficult (but certainly not impossible technical challenge).  They would probably have to have a small group of artists (or whatever) make / place invisible objects over the pre-made maps, which would add a significant amount of work for a game with 20 devs.  Then It gets even more difficult when you take the problem of moving objects on maps with baked lighting (which is why the Torment promo vids looked weird, the spinning wheel didn't have a shadow).  

 

The actual movement part would be ridiculously easy technically.  You would make a 2D cone (or whatever), and make it a child of the mob object.  Then, rotating an object toward another object is only a line or two of code (vector3.lerp etc.)  Showing all that in real time in  a way that doesn't overwhelm the player in a game that is ostensibly RT is much more difficult.  

 

I do think it's an area they could profitably pursue in an expansion / sequel.

[Sabotin]

Not sure if this applies as LoS, but I hope the circles aren't going through walls. It would be awkward to get detected by enemies from a room you didn't even know was there.

[Mor]

Question to the guys who have been following on this

1. Does stealth mechanics are only limited to the player party, or we might get ambushed by some critters? it might be nice if it worked like the hidden object spot, that high perception will show us a red circle of enemies a head who are waiting to ambush us once we are closer.

2. Does stealth provide any kind of surprise attack bonus? at least for the first round.

[rjshae]

When you're standing still, a circle is a convenient abstraction for being generally attentive and glancing about. But when you're moving forward, your attention is generally focused that way. Thus I might be tempted to offset both perception circles in the direction of movement. When the character returns to rest, the circles can drift back to a centered position.

[Jobby]

 

 

 

Given what Sawyer said about how difficult line of sight would be to implement I doubt that will be present, Don't get me wrong it would be awesome but it just seems unlikely, the circles are a decent compromise, don't like the idea of scaling to level tho? (don't know if that has been misinterpreted on my part)

 

Instead of circle, use half-cirlces. That leaves a blind spot for vidual detection.

 

Audio detection is another matter.

 

And in which direction would the half-circles face? That was the difficulty he was mentioning in regards to line of sight as far as i could tell. 

 

Don't get me wrong line of sight would be a great mechanic but I think the circles are a huge improvement over I.E. systems and in my mind they represent audio and visual in an abstract way (the semi-circle behind the facing direction of the enemy could be considered audio and the semi-circle in the facing direction could be considered visual).

 

It's not perfect but its certainly not bad.

 

 

It's a matter of interpretation, but I read that as him saying that it was too complex for the player to deal with six separate entities and however many guards.  Then there's the weird problem of mixing 2D and 3D, which would make lighting a difficult (but certainly not impossible technical challenge).  They would probably have to have a small group of artists (or whatever) make / place invisible objects over the pre-made maps, which would add a significant amount of work for a game with 20 devs.  Then It gets even more difficult when you take the problem of moving objects on maps with baked lighting (which is why the Torment promo vids looked weird, the spinning wheel didn't have a shadow).  

 

The actual movement part would be ridiculously easy technically.  You would make a 2D cone (or whatever), and make it a child of the mob object.  Then, rotating an object toward another object is only a line or two of code (vector3.lerp etc.)  Showing all that in real time in  a way that doesn't overwhelm the player in a game that is ostensibly RT is much more difficult.  

 

I do think it's an area they could profitably pursue in an expansion / sequel.

 

 

 

If that is the case then LOS should be in in my opinion, Stealth is in no way essential to complete the game so i wouldn't say it's too complicated for the player, I never thought about what you said regarding 2d/3d and the lighting situation until now, that may make that aspect considerably more difficult, which is a shame, I love the concept of waiting till nighttime before you go prowling about.

[anameforobsidian]

I misspoke.  LoS, where the enemy's vision is blocked by objects would have the same problem as lighting.  It would be fairy easy if you had a strictly 2D perspective, like this game, but this game doesn't.  For this, you would have to put isometrically rendered objects on top of the scene as well as have someone sketch LoS barriers on top, turn the whole thing invisible, and shoot rays from the mob to the object to check LoS.  

 

So you're duplicating the work that a 3D game does once (albeit with no modelling etc.), you're wasting spending a bunch of money / time less than optimally.  It could be doable, depending on what the artist's do before they render (like if they make pillars before hand).  But it's certainly not trivial, and it would take artists / bugfixers doing mostly that for months.

 

What I was proposing was far more simple than that.  They keep it 2D, they just change the shape and track orientation for mobs (or maybe make a 3D cone object rotated to the camera angle and show where it intersects the plane).  That's still more work, but not much, and it offers some basic LoS stuff.