3E Stealth

[ACOS]

[...] If Stealth is a "roll until you fail" situation, then iterative probability means that you will fail and stealth is useless. If Stealth is a "roll a bunch of dice and succeed if you get middling results" situation, then iterative probability works in your favor.

Okay, I do back off just a little -- I mistakenly read that as you taking the same angle on iterative probability as Roy from back in the day. You're actually taking the opposite -- fair enough.

On that note, I'll also grant you that very few GMs are actually going to have it in them to satisfactorily adjudicate such a scene such as those described above (whether that be an issue of motivation or cognition); because crafting complex/complicated scenes and pulling them off as intended is, as it turns out, also very hard. I'm unsure as to whether or not that is something that can be rule-designed away; or if it may/may not be easier to teach people how to better conceptualize and analyze scenarios in order to better chain sequences together - in part, through asking the proper dramatic questions surrounding particular types of action sequences.

[Username17]

On that note, I'll also grant you that very few GMs are actually going to have it in them to satisfactorily adjudicate such a scene such as those described above (whether that be an issue of motivation or cognition); because crafting complex/complicated scenes and pulling them off as intended is, as it turns out, also very hard.

The issue is that none of the scenarios being described are particularly complex. We haven't gotten into extra teams or reserve troops where it definitely matters whether someone notices target A but not target B. The Museum Theft and the Forest Encounter are trivial stealth challenges, and a halfway decent stealth subsystem should be able to crap them out on autopilot.

Throwing up your hands and saying that DMing them is always going to be hard is simply conceding that you're never going to have a set of stealth rules that isn't fucking terrible or absent altogether in any game system you ever play. And I just can't accept that that's true. Yes, the very best Stealth systems in use today are "roll the dice and tell the MC how stealthy your character is being, using the results of the numbers as an improv prompt, then the MC will riff some shit off the top of his head." And that's extremely close to not having any rules at all, though of course that's exactly what we used in every edition of Shadowrun, and most other games as well.

The thing I memorised from Anatomy of Failed Design: Skill Challenges was how they were so close to at least being mediocre but usable and not being a dumpster fire, and that part of why you can tell 4e was poorly designed was that none of the iterations implemented any obvious fix to make the system actually meet its nominal design goals.

That's exactly it. The basic concept of Skill Challenges would be very easy to adapt to work with the Museum Heist scenario. Different numbers of successes and failures cause the alarm to go off at different times, and you hypothetically have enough skills and enough mid-challenge events that there are choices to make other than spamming "Stealth" and "Thievery" over and over again. Fucking done.

I mean, that would simply be mediocre. And it would be difficult to make such a simple system handle more complicated scenarios like "three gangs meet up for the meeting, and all three of them have brought secret reinforcements because they don't trust each other." But tweaking it so that it handled the simple shit quickly and without fuss isn't particularly difficult.

We don't have to and should not despair.

-Username17

[ACOS]

Yes, the very best Stealth systems in use today are "roll the dice and tell the MC how stealthy your character is being, using the results of the numbers as an improv prompt, then the MC will riff some shit off the top of his head."

And that is actually the preferred mode of play for many people. Getting bogged down in crunch-heavy systems has its own set of problems -- problems, frankly, I'm simply tired of contending with. But I get it, there are also plenty of crunch-monkeys out there (I sometimes still get those moods as well). The more I interact with rules-heavy systems, replete with robust mini-games, the more gamey the play feels (especially when you've got people at the table who just can't be bothered with remembering all that shit, so you have to hit the pause button every 5 minutes to coach them through it ). I'm in a place right now where I'm just inclined throw up my hands and use GURPS Lite for everything. The major downside of that being the finesse work needed to actually pull that off.
It's plenty of work either way. For me, it's a compromise of which end to focus that work. I've come to find it easier to focus on framing and communication rather than rules-hammering. Though, that probably doesn't do anything to address disparate incentives towards modes of play.

You're right, neither of the above scenes are particularly complex; and as far as I can tell, MGuy and I (more so MGuy) explained how the current systems at hand can be perfectly serviceable for said scenarios (though, the way one followed the other, it seemed that you were clearly leading towards increasing complexity en route to proving a point). The types of complicating factors that you've been describing seem like narrative framing issues. Yes, there is obviously a critical mass of complexity at which point the current system(s) simply fail. You say that this can be fixed, and I say good luck to you and yours. Just mind exactly what these extra rules are trying to fix -- which are you trying to manage: complexity, or complication? I ask because I'm not sure, and those 2 things probably require different strategies.
Speaking of Shadowrun, RE:"literally naming your game Cloak & Dagger" ... while I'm dubious of the idea that D&D would significantly benefit from the type of sub-system you propose, I do believe that Shadowrun might ought to demand it.

Throwing up your hands and saying that DMing them is always going to be hard is simply conceding that you're never going to have a set of stealth rules that isn't fucking terrible or absent altogether in any game system you ever play. And I just can't accept that that's true.

It seems that you are much more the optimist than I. But fair enough, it has only been 40 years.

[Username17]

The more I interact with rules-heavy systems

I'm going to stop you right there, because what the actual fuck are you talking about?

There isn't a dichotomy between having a stealth number on your character sheet that does absolutely nothing except allow you to report a higher number while doing an improv stealth scene and having a massive set of charts with thick mini games and exhaustive line of sight rules and shit. Rules don't have to be heavy in order to exist. You can still be a rules-lite game and have some kind of actual procedure for handling simple stealth questions.

If you wanna shrug your shoulders and tell me that your system can't give you unambiguous answers to situations like The Triple Cross or The Great and Powerful, then sure. Go ahead and tell me that scenarios with dozens of individuals in multiple groups with different agendas attempting to hide from and deceive other groups is something that's going to require the MC to "just wing it." I can accept that. But The Museum Heist is a fucking simple scenario. There basically isn't an excuse for an RPG to not have a procedure there.

It doesn't have to be a complicated procedure, it just has to exist. And the Gygaxian "roll to move silently, roll to hide in shadows, the DM makes something up based on how the numbers make him feel" system developed in the 1970s doesn't fucking count.

-Username17

[ACOS]

Dude, chill. Seriously. That thing you seized on wasn't designed as an argument; that was me trying to bow out gracefully. Untwist your panties and move on.

[shinimasu]

Is you say "rules light" three times while looking in a mirror Frank Trollman will manifest physically in your house to swear profusely at you for an hour.

[Lokathor]

Me: *Looks at the first sentence of the After Sundown hack I made today*

Altered Dragon is a (relatively) rules light fantasy adventure game based upon the mechanics of dice pool games and the stories/ settings/ themes of classical fantasy adventure games.

Oh no

[Username17]

At the risk of getting this thread dangerously back on subject, let's talk about a very simple stealth challenge that probably should be handled with a single check: Guards Playing Cards.

Guards Playing Cards
There's a hallway, there's an open door from the hallway into a room. There are Guards in the room and they are doing something that is distracting them. Playing cards, watching TV, eating dinner, whatever. There is a Spy who wishes to sneak past the room to the end of the hallway.

That's an extremely simple scenario, and one which should probably just be a single test against an easy difficulty to do without making any noise that alerts the Guards or get spotted by the Guards on your way through. The problem is that in 3.5 D&D, the Spy cannot succeed at this task. The act of moving through the doorway leaves the Spy momentarily without any cover or concealmeant, and the game has no rules for facing so the Spy automatically gets seen by not just one Guard, but every Guard.

-Username17

[maglag]

At the risk of getting this thread dangerously back on subject, let's talk about a very simple stealth challenge that probably should be handled with a single check: Guards Playing Cards.

Guards Playing Cards
There's a hallway, there's an open door from the hallway into a room. There are Guards in the room and they are doing something that is distracting them. Playing cards, watching TV, eating dinner, whatever. There is a Spy who wishes to sneak past the room to the end of the hallway.

That's an extremely simple scenario, and one which should probably just be a single test against an easy difficulty to do without making any noise that alerts the Guards or get spotted by the Guards on your way through. The problem is that in 3.5 D&D, the Spy cannot succeed at this task. The act of moving through the doorway leaves the Spy momentarily without any cover or concealmeant, and the game has no rules for facing so the Spy automatically gets seen by not just one Guard, but every Guard.

-Username17

That's just the start of your problems.

If the guards have a guard dog, scent auto-detects the spy even if they have +texas hide/move silently.

If the guards are grimlocks, they have blindsight and can perfectly hear the spy even if they roll infinity move silently.

If one of the guards has a wand of alarm, then the spy is also auto-detected.

Trying to be a spy in 3.5 is suffering.

[Trill]

That's just the start of your problems.

If the guards have a guard dog, scent auto-detects the spy even if they have +texas hide/move silently.

If the guards are grimlocks, they have blindsight and can perfectly hear the spy even if they roll infinity move silently.

If one of the guards has a wand of alarm, then the spy is also auto-detected.

Trying to be a spy in 3.5 is suffering.

As someone who's not familiar with 3.5's actual gameplay: How big of a change (both for the stealth minigame and the rest of the game) would it be to remove all those autodetect features (or at least make them rare)?
Because to me it seems there are lots of ways to instantly detect someone, but not enough ways to counter it and not enough ways to be completely indetectable.

[Username17]

Actually the interaction between Blindsight and Move Silently is a null pointer error. Blindsight does not say whether it beats Move Silently, or loses to Move Silently, or has to make rolls. You don't make Spot or Listen checks to perceive creatures and you ignore most forms of Concealment. But you don't actually ignore cover and you don't ignore the Hidden condition. So um... wat?

The primary issue with Blindsight and Hiding is the Guards Playing Cards problem generally. Hiding doesn't work if your opponent can see you, which is fairly tautological. The game doesn't have facing and only tracks peoples' locations to 25 square foot areas - whether people have a clear line of sight to each other is the question the Stealth system is supposed to answer, but instead it's the first question the Stealth subsystem asks.

Blindsight doesn't actually say how it works with respect to people trying to be undetected by it, which is rather crucial bit of information for a fucking fantasy sensory power to omit. Generally speaking though, it works when you have "line of effect," which gets to the same "hiding behind your own tower shield" issue that was brought up on page 1.

-Username17

[Kaelik]

"autodetect features" are not actually the problem. Alarm is a magical spell, it literally glows if you have the right abilities, and then you can see it and bypass it. There are items and spells and actions that make you scentless, and also ones that make you immune to or should make you immune to blindsight.

But Darkstalker, a feat you take to make all the detection methods not work at all, is a terrible thing.

Hide and MS are skills. And in 3.5, that means if you specialize in them you have +60 on a 20 point RNG over people who don't. If you say "ah, but Hide works against all the methods of detecting you, not just sight and hearing." Then what you've actually done is declare that whenever you get to roll it, Hide is an automatic success.

Having a bunch of auto failures to stealth that you can detect/anticipate and counter with in game actions is a hell of a lot better than just having one auto success that you invest in at character creation and then makes you permanently undetectable.

None of which is to say that Hide actually works for anything in D&D, since the reason people do dumb stuff like Cast Blur and then roll their 1d20+50 Hide check at level 10 is because of the other failings of Hide, like that you can't stealth across a doorway past guards who are distracted, or how you have to roll every round with varying penalties until you fail on the 3rd one, so you need to have +20 over their spot to have any ability to sneak past them at all.

[Orca]

The stealth budget idea works for the basic scenario. Probably at a minimal cost. Whether the autodetects autodetect or cost varying amounts against your stealth budget (probably less if you disable the alarm or have counter-countermeasures) is a question to be answered.

[Kaelik]

My overly complex 3.5 compatible stealth budget system takes up 5 points for detection that tells you someone is there, like scent, 10 points for something that tells you exactly where they are, like blindsense, and 20 for something that is sight, like Blindsight. (If you are within their range and they are able to detect you.)