Stealth Discussion

[OgreBattle]

The only time I've been satisfied with stealth rules is in war/skirmish games where stealth/detection is part of the deployment process.

So 'stealthy' units get to deploy closer to the enemy, or pop out of the terrain a turn later, and 'detector' units counter the stealthy unit's ability to freely deploy.

D&D's surprise round kind of represents that to varying degrees.

I'm mostly interested in tRPG's for combat though.

[Reynard]

Josh_Kablack:
> The bigger issues with my suggestion would come with transitioning between out-of-combat and in-combat stealth
Yes. Out-of-combat and in-combat actions need to be clearly separated and transition carefully handled.

Even if "balance" angle could be ignored, some actions would require too much rolls/details, if done via combat system. Those actions need to be condensed into single (non-combat mode) actions. For example, one dice roll that declares if lone scout discovers something, chickens out along the way or gets noticed by enemy drastically reduces problems that arise from splitting the party or from lack of reconnaissance.


But this transition problem is not limited to the Stealth alone. Implementing transition for other systems merits some thought too.

Also, I think there are 3 "game modes", not 2 (in- and out-of- combat). Each requires its own granularity to function smoothly. Using musical terms (because tempo), we have Allegro (running pace, when each second counts), Andante (walking pace, when only general actions matter), and Adagio (the slow pace, when only general intentions matter). Or personal scale, local scale, and global scale. Or (using DnD terms) - combat-mode, Take20-mode, and downtime-mode.


Combat-stealth is the highest granularity Stealth (Allegro), then comes the out-of-combat Stealth (when Thief sneaks ahead of the party and GM makes only 1 (one) roll to determine the outcome - Andante), and, finally, there is downtime Stealth of weekly (monthly) roll (one roll for the whole party) to see if their smuggling operation goes bust (Adagio).

Some examples:

Party attempts to sneak unnoticed through the enemy territory. It's a 2-week journey and details would boggle down game (Adagio). One roll for the whole party to determine the outcome is enough. Result of the roll: party travels 5 days unnoticed, until one of the patrols stumbles upon the party and game switches to Andante pace. [insert here one or more checks on whether party notices patrol, how much time PCs have and whatever other starting conditions there are].

Now its Andante and each PC gets their own roll (and choice of action) to determine if they manage to hide (or if they try to do something else), while Ranger and Ninja use their Stealth abilities to buff everyone's rolls.

If patrol notices PCs, it's transition to the Allegro - slowpoke Wizard gets caught in the middle of battlefield, Ranger has good shooting position, while Ninja hides his presence completely and can either stay stealthied or appear anywhere (even right beside squishy enemy caster) after battle begins.

Then it's Allegro combat-Stealth time, with Sneaky stances.


Diplomacy system would include Bard using Puppy Eyes stance during combat (Allegro), Flamboyant Haggling maneuver to sell dragon's skull (Andante), and monthly roll to see how many new members he managed to convert into his cult religious practice (Adagio).

When middle-level PC makes a roll to see how many (dozens of) mooks he killed after fighting for a few minutes - that would be Andante-Combat. Adagio-Combat - one roll decides the result of the battle between King's army and Horde led by party Barbarian (high level implied), while the rest of the party is either busy elsewhere or cannot be bothered by petty mortal squabbles (Wizard).



tl;dr: (also for Prak) 3 Stealth sub-systems and 2 transition methods/sets of variables are necessary to get a complete Stealth system. Same goes for Combat and other systems.

I'm currently thinking along the lines of ToB-like Stealth/Awareness/Subterfuge system with "Disappear into the crowd" Andante-maneuver and "Running Guild of Thieves" Adagio-stance from the Unfettered Rogue style. The problem here is to determine what actions should there be (and sheer amount of work, of course).

[Josh_Kablack]

Also, I think there are 3 "game modes", not 2 (in- and out-of- combat). Each requires its own granularity to function smoothly. Using musical terms (because tempo), we have Allegro (running pace, when each second counts), Andante (walking pace, when only general actions matter), and Adagio (the slow pace, when only general intentions matter). Or personal scale, local scale, and global scale. Or (using DnD terms) - combat-mode, Take20-mode, and downtime-mode.

I submit that perhaps "combat time", "dungeon time" and "overland time" might work as more D&D-specific terms for the game modes you are describing.

In combat time, you in in the combat minigame, and round-to-round positioning at the hex/grid scale matters.

In dungeon time you are exploring at the micro-scale and trying to set yourselves up in advantageous positions should combat happen, and time and location are handled more generally.

In overland time you are journeying between locations and/or exploring at the macro scale and resources like rations and hours spent traveling per day are the ones that matter.

And while I'm not sure how precisely to relate these back into the stealth minigame, I think it could be pretty smart to design a game with all three modes made explicit in the rules and supported by various abilities.

[RadiantPhoenix]

Would this be something like AD&D "rounds", "turns", and "hours or days"?

[Reynard]

OgreBattle:
> I'm mostly interested in tRPG's for combat though.
I strongly suspect that's because only combat system is somewhat developed.

If RPGs could tackle trading/travel adventures a-la Marco Polo, feudal intrigues, or Genghis Khan campaigns of uniting tribes and taking over the known world via weaponized bureaucracy, I'd guess you would be interested in those too.

Same goes for stealth, heists and the whole lot of other activities.


Josh_Kablack:
> "combat time", "dungeon time" and "overland time"
Perhaps. But "dungeon" and "overland" are misleading. It's not like you won't routinely get "dungeon time" outside dungeons or "overland time" while staying in one place, even if the game is extremely DnD-esque.

Granted, Adagio/Andante/Allegro are a bit confusing. I took them from Italian-style heartbreaker about city-state politics I tried to make, so they fit there (imo).

I just don't want to see "social combat" all over again, when people pigeonhole themselves into "social HPs" and "social AC".


RadiantPhoenix:
> Would this be something like AD&D "rounds", "turns", and "hours or days"?
I don't think so. Not as an exact transition 60 rounds = 6 turns = 1 hour, at the very least.

For example, GM doesn't get to have a turn (not as specifically, as during combat, at the very least) and there is no set sequence of player turns. Players simply tell GM what they do (try to sell loot, or poke at the trap) and GM resolves their actions without noting "Rogue spent 32 minutes 3 turns and 2 rounds poking at the trap, while Paladin prayed only half of that time" and feeling the need to ask Paladin what he was doing the rest of that time.

Bigger scale should not be about bigger numbers (most common flaw of all mass-combat systems). Bigger scale is about less accounting. I.e. more quantum bears approximations. (Properly implemented) transition from discrete "gold pieces" to more abstract Wealth system would be better example of changing scales/tempo.

[tussock]

In BECMI, there's a more solid turn structure for exploration. Not spelled out anywhere, but almost everything outside combat has a duration in turns (in multiples of 6 for longer ones, so 24 turns for longer spells, 12 turns for a lantern, 6 turns for a torch, a few spells last 1d6+1 turns or something), your movement and exploration speed is per turn, doing shit like setting up surprise takes a turn, searching takes a turn, combat takes a turn after you rest up, smashing something takes a turn, wandering monster checks are every 6 turns (give or take), and once every six turns you have to rest.

Turns are the pace of the exploring part of the game. The overland movement is just miles/day by terrain, with an encounter check and a getting lost check, so hex-crawls go very fast.

But BECMI stealth is almost completely abstracted out, there's just extra random damage (and you lose your surprise chances and stuff) for PCs acting without due caution and stealth, and a few PC types get stealth modifiers to starting combat. You can't do much with it, but it does what it does, clean and fast.

[Prak]

Without trying to find norse terms for musical tempos (hahahahahafoolsfuckingerrand), my mind turns to the "merc./adventuring company" theme that I put into Midgard when it was more heavily influenced by the Hobbit movies and I get "Encounter," "Operation," and "Campaign" as terms for the three scales. I think they would almost work for D&D except that in D&D campaign has connotations of "the entire story, and the next isn't necessarily connected" as well as connotations of being focus.

[...You Lost Me]

If your ability to disappear exists as a combat power and you don't let players use it in places outside of combat, players are going to be very angry at you. Do not make a system like that.

[Prak]

A three phase system doesn't necessarily have to be that way. Power writeups can include rules text for each phase.

[PhoneLobster]

Every additional phase you add, and every mini game you add, multiplies the potential transition mechanics significantly.

I've tried writing systems that transitioned between Combat/Social/Stealth/Chase minigames and others that worked with transitioning between Combat/Exploration/Travel time scales. One thing I learned was that almost ANY discrete transitions that require major rules to handle the conversion are too many transitions.

The other thing I learned is that no one fucking likes distinct and separate mini games, no one wants to learn a stealth or chase minigame that is separate to the combat minigame they want to learn a fucking compatible mechanic for stealth and chasing that occurs IN the combat minigame.

Seriously people, less with the distinct minigames, the moment your proposal is close enough to earn the name "minigame" be it due to a theme or time scale change it is probably a bad idea to continue with the proposal in it's current form.

[tussock]

@Phonelobster, no.

People might want mass combat to use the small-skirmish combat rules of the PC's scale, but that doesn't work at all, so under your concept there we don't even get to have mass combat rules, and that fucking sucks. We should have that in core books and also rules for when to transition to the big scale.


I also don't see how there's "too many transitions". There's basically conditions for starting each system, which can just kick in from any other system by just doing that. If you can pass the condition to enter a chase then you're in a chase, if you agree to parley then you're in the social system, and if you go to stab someone you roll initiative for combat. All of that can happen during overland travel or while resting.

You may also like to write options within one system to force everyone to switch against their will no another, but you don't need them and they'll probably just throw rubbish at you anyway (force the enemy to talk for three days while reinforcements arrive, when they just want to stab you and leave instead).


So very few games only use one system for combat and everything else. There's mini-games everywhere, in everything that sells numbers you'd care about, and people love them because doing that shit with the combat system would be frustrating and stupid and maybe not even in the game otherwise.

There's a lot of mini-games that are terrible, that's a good argument, I'd agree with that. What players really want is something that just fucking works. But seriously dude, the moment you propose doing diplomacy with the 6-second round UGOIGO combat system, let alone multi-month sieges or year-long cross-continent fetch-quests, it is probably a bad idea to continue with the proposal in it's current form. Because you need a mini-game for resolving that shit.


Also for scouting, or infiltrating an enemy base on the quiet if you want that in your game, because combat-stealth with multiple checks for every actor every six seconds doesn't work there even in theory. Likewise, whatever rule lets you spy on a conspiracy against the king for their hour-long meeting can't be allowed back into the combat system.

[Ghremdal]

My solution is to get rid of the perception skill entirely.

A character that wants to stealth rolls the stealth skill at the beginning of the scene.

Guards start at awareness rating of 0. Depending on the actions the sneaky character takes the guard awareness rating rises. When their awarness rating is equal to the sneaky characters stealth skill, he is found out. A sneaky character can also get perception resistance (PR) to reduce the rate the guards gain awareness rating. The default rate of awareness rating increase is 0 (awareness gain = AG), and it increases on a per round basis.

Example of modifiers for a dice pool system, where the expected skilled sneaker has around 8 hits.

Sneaker is standing still: +1 PR
Sneaker is at close range: +1 AG
1 guard on scene: +0 AG
2-3 guards on scene: +1 AG
4-7 guards on scene: +2 AG
Sneaker is moving: +1 AG
Sneaker is moving at full speed: +3 AG
Sneaker is running: + 7 AG
Sneaker is at long range: + 2 PR
Heavy cover: +2 DR

So for example if a sneaker is at close range, standing still, and there are two guards on the scene. The guards AG is 2, and the sneaker has a PR of 1, so the guards effective awarness gain is 1 per round. This means the sneaker, if conditions remain the same, can stay hidden on the scene for 8 rounds.

[PhoneLobster]

I also don't see how there's "too many transitions".

Then go ahead, waste your time making a piece of shit game no one wants to play.

[Irish]

Not that I'm particularly well suited to discussing entire rules systems, but IMO any stealth system that can't be resolved in 2-3 dice rolls MAX between the player and the DM would make me pretty annoyed. Time consumption on something like this should be kept to a minimum, not abstracted out to nearly 30+ die rolls.

Sorry if this point has been brought up already, I've mostly just been skimming through the thread thus far.

[...You Lost Me]

A three phase system doesn't necessarily have to be that way. Power writeups can include rules text for each phase.

That's certainly possible, but if I can disappear as a power and that translates to +20 Stealth HP or 10 Stealth DR or whatever I am going to put down your game immediately.

[pragma]

My solution is to get rid of the perception skill entirely.

A character that wants to stealth rolls the stealth skill at the beginning of the scene.

Guards start at awareness rating of 0. Depending on the actions the sneaky character takes the guard awareness rating rises. When their awarness rating is equal to the sneaky characters stealth skill, he is found out. A sneaky character can also get perception resistance (PR) to reduce the rate the guards gain awareness rating. The default rate of awareness rating increase is 0 (awareness gain = AG), and it increases on a per round basis.

This sounds a lot like Dean's "shallow HP pool" in the other thread. I like Dean's idea and, consequently, like this one. Just wanted to point this out so that we don't lose the other discussion. It starts about halfway down the page here.

[RadiantPhoenix]

"Stealth HP" is apparently the system used in Skyrim and Fallout: New Vegas

[Username17]

If you insisted on doing things like attacks and hit points, you'd actually want to model the stealth as the attack and damage. The amount you could get away with being determined by how much stealth "damage" you generated. But that's because the stealthy person is the one is actually doing things to stay hidden. Dean's observation attacks only make sense when the guards are aware and actively searching for a ninja. In the vast majority of cases, it's in no way clear what a perception attack would even mean.

It's a difficult design problem. You're trying to generate answers to the question of how alert each character is, and also track how aware each character is of each other character. And while that's a lot of information, you also need to do it in just a few literal die rolls and also demonstrably be reusing the core game mechanics of the game.

-Username17

[Ghremdal]

-first off I hate perception as a skill. Its so overused in any one of games I play it makes no sense anymore.

The system I posted above I playtested a bit It works fairly ok for a single sneaker vs a group of identical guards. These are the problem points I found:

Two+ different perception type cooperative guards.

Two+ uncooperative groups of guards.

2+ different skill rating sneakers.

Highly perceptive stuff like dragons and ogres detecting expert skill sneakers at long range through cover in a matter of rounds.

[Username17]

The problem of course is that you just listed almost every scenario that a stealth system is likely to be called for. So at this point you don't have much of anything.

Your stealth and scouting system needs to output:

• Encounter distances.
• Who is surprised at the beginning of an encounter.
• What the aware characters can do without starting the encounter for the unaware characters.
• Whether alerted guards detect a sneaking character they are looking for.
• Whether alerted guards detect a sneaking character they are not looking for.

And these are all really useful things to have be non-deterministic, but you need to resolve this stuff with a small number of rolls and calculations because the action is often fairly one sided and the game can't afford to drag in this area.

-Username17

[czernebog]

If you're talking about a fully-fledged stealth subsystem, suitable for infiltration and escaping unnoticed, being hidden during an encounter entails that PCs fulfill the objectives of:

• Obscuring knowledge of the party's presence: when you're in someone's secret base, it's better if they don't know you're there (or at least can't pinpoint where you are).
• Obscuring knowledge of hostile action: obviously, if no one realizes you're bumping off guards, you're doing stealth right.
• Obscuring knowledge of the party's connection to hostile action: if you can pull off looking like an innocent day laborer who just happened to be next to that guard who was poisonedsuddenly took ill, you're doing stealth right. This includes not leaving evidence (so that you can't be recognized on sight or by a fingerprint database/magical aura).

This involves controlling beliefs of the party's opposition:

• Does the opposition know that the party is present and operating locally? How precise is that knowledge?
• Does the opposition know that the party is hostile?
• Does the opposition know that the party is responsible for hostile action?

Is this too fine-grained? It can work in conjunction with a coarse measure like a shallow pool of HP that represents the party's stealth presence, with some sort of wound levels on that condition monitor corresponding to discrete changes in the opposition's knowledge.

This requires having an abstract view of the situation, and once battle music starts a different set of mechanics have to kick in. Once you're in tactical combat mode, stealth presence can feed into bonuses to stealth-powered abilities and battlefield positioning, and after a battle is over stealth presence is adjusted based on how many explosions went off, if the enemy got word out, etc.

If prolonged stealth missions are run as a series of (potentially recurring) obstacles that the party has to sneak, fight, or otherwise bypass (for obstacles like "the guards in the tower," "the moat," and "the hellhounds patrolling the castle interior"), then the persistence of the stealth monitor provides a means of representing how well these components of the opposition are communicating and whether the party is doing a capable job of keeping a lid on their progress infiltrating hostile territory.

It's a difficult design problem. You're trying to generate answers to the question of how alert each character is, and also track how aware each character is of each other character. And while that's a lot of information, you also need to do it in just a few literal die rolls and also demonstrably be reusing the core game mechanics of the game.

I'd like to think that backing off to obstacle vs. party helps with the awareness problem outside of the combat subsystem. On a combat grid, adding additional shades of "NPC knows where PC is" doesn't help (and actually makes it worse). Is it satisfying to declare that the tactical combat system is what happens when hostile agents interact with the knowledge that they're within shooting distance of each other (or else you fall back to skill checks)? You still need to be able to deal with a situation like a bunch of invisible rogues trying to flank ogres who are flailing around wildly at the PCs they know to be somewhere within arm's reach.

[Username17]

On a combat grid, adding additional shades of "NPC knows where PC is" doesn't help (and actually makes it worse). Is it satisfying to declare that the tactical combat system is what happens when hostile agents interact with the knowledge that they're within shooting distance of each other (or else you fall back to skill checks)? You still need to be able to deal with a situation like a bunch of invisible rogues trying to flank ogres who are flailing around wildly at the PCs they know to be somewhere within arm's reach.

It gets worse than that of course. You also have circumstances where just one of the PCs is an invisible rogue trying to do some kind of flanking action while the rest of the party are Aragorn, Gimli, and Legolas facing down the Trolls head-on. What the invisible rogue is able to accomplish with their flanking action is very important, both to Frodo's player and the rest of the people at the table.

Having one character be "stealthy" in a combat where another character on the same side is not is a pretty common occurrence in combat. It's the default state for the "four person party" that only one or two of the characters are stealthing once arrows start flying.

-Username17

[PhoneLobster]

I feel like this is touching very closely on a lot of the stealth work I've done in my mousetrap system.

So I'm going to ramble on about it, because that's what I like to do regardless of a bunch of killjoy grognard wankers around here.

Time Scale Issues
The stuff about "combat time cannot be dungeon explore time" is kinda dumb.

I've tried that before (the combat/explore/travel model) and even aside from the transition issues (which are major) it very quickly became apparent that it conflicted with a major goal I wanted in the game.

Whole of dungeon encounters was that goal. The whole "fight each room one by one as an encounter" is and will always be a bad thing. There NEED to be correct mechanics, from the foundations of the system upwards, to support the possibility of combats being detected from other rooms, information propagating through the dungeon and the inhabitants appropriately responding.

Requiring that to transition through an Exploration scale DURING combat is basically totally unacceptable, which is OK because it is actually also totally fucking unnecessary.

But anyway, in vague terms, having your "detect a fight" mechanic say "each X combat time units, characters within ? combat distance units may detect actions in combat time and once detected may respond by approaching with combat actions during the same combat turns as the fucking combat" is way better than saying...

"each X combat time units, characters within ? exploration distance units may detect actions in combat time and once detected may respond with actions in exploration time to move towards the combat until a transition point (which we need to convert between combat and exploration time units to figure out when it happens and combat and exploration distance units to determine where it happens) and only then their actions become combat actions in synch with the combat mechanics without elaborate transition/conversion rules."

Fortunately as mentioned you can just do the first thing.

And so I did that. There... isn't too much more to be said about it. nicely abstracted time and distance units, fairly large granularity positioning, and notably a bunch of actual fucking mechanics for stealth, and the detection and spreading of alarm information among inhabitants of a dungeon were key parts of making that work, and of the many crazy things I've tried over the years the "whole dungeon encounter" scale (and designing the system from the ground up with that actually in mind instead of tacking it on like bullshit at the end) has been I think one of the really successful ones.

Perception, active detection, alertness, stealth defense
Or whatever you want to call all that mess.

Having an active perception roll is not automatically a bad thing... but having a system with active stealth rolls AND active perception rolls... is certainly at the very least very strongly at risk of bad things.

As a general rule unless your entire base mechanic for your system is opposed rolls (probably a bad choice on its own) you really should just pick which side of the mechanic gets the active action/roll and damn well stick to it.

I went with the stealth roll being active. Perception was just a defense value, and since I happened to have a Social Defense value in my system it does double duty as a vaguely "awareness" themed defense against social AND stealth actions, not a perfect marriage but certainly close enough to work out well what with it making sufficient sense while preventing the system having 3 separate defense dependencies and adding value to the "secondary" one.

As for becoming alert/alarmed/active searches, well the game also has an "Alert" bonus for the social defense score when physical attacks are observed, which already applies to social defense against stealth actions already. So hiding in the middle of combat is already harder, throwing in a few extra triggers for alert (like observing a failed stealth check) and that side of things was looking pretty solid.

Deciding there really should be a "combing the area" active action of some form worked out ok as long as it required some sort of cost (actions), some sort of justification (plausible information that something might be found there) and NO ACTIVE ROLL. So, with appropriate justification/excuses characters can burn their actions on becoming alert vs specific areas in order to "search the place".

Imperfect Detection
"wait did you hear something over there?" needed to be a thing at some level.

The way that worked out was fairly simplified.

Failed stealth checks within Line Of Sight provided full information on who failed at what where.

Failed stealth checks outside of Line Of Sight (there are bonuses to stealth checks for being behind barriers, potentially big enough for auto success, but potentially small enough to fail while being out of line of sight) provide imperfect information on the precise details of who or what caused some sort of vague commotion over behind the wall or in the next room behind that closed door or whatever.

Just One Guy Stealthing
So. One guys stealths everyone else stabs? No problem. As an emergent upshot of using the same time scale and mechanics as combat we know exactly how it works, and without the Alert bonus being too punishing it is a perfectly viable strategy, especially for a stealth specialist (which the one guy will usually be, and good for him too).

Reasons To Stealth
Any system that lets you stealth should give you reason to stealth. Mine rewards surprise assassination attacks, like, ALOT. And lets be honest, why shouldn't it. In a system where 2-3 successful attacks might take down most targets a mechanic where you need to succeed in at least one stealth roll, possibly penalized by in combat alert bonuses on the target's observation defenses, AND an attack roll to perform your assassination action it probably should be about that rewarding.

But also using a "whole dungeon encounter" scale can provide some pretty good reasons to try and be stealthy. You ARE going to want to try and get to a good position before the whole complex is alerted, ideally you may NEVER want the whole complex alerted if you can avoid that and just get what you want to done.

Better yet with specific mechanics to determine the location and state of readiness of major characters in a dungeon (I totally have those) AND a simple standard set of mechanics to determine if they notice things going on elsewhere in the dungeon (have those too), and how quickly it takes screaming messengers (or fleeing routed victims) to run through the base spreading information about attackers (hint, it happens in combat time using the standard combat actions and mechanics), one VERY good reason to stealth around, and EVEN to have a stealth expert separate from the group and go do useful things on the other side of the encounter map, is that succeeding in finding a major enemy and taking them out before they are alarmed or equipped is VERY REWARDING.

Causing A Disturbance
To facilitate the whole of dungeon thing it was clearly important to provide a means of determining when characters detect actions, like say someone just walking around, or someone fighting, or someone fighting and shouting for reinforcements. And for that to be a thing that works like stealth but for it to happen even when characters are NOT trying to be stealthy and even when they actually WANT to be detected.

This indirectly resulted in what is arguably the ONLY action in my entire set of mechanics where the active rolling character may, depending on context, actually WANT to roll low.

It works exactly like a stealth action, only without action cost and it is involuntary. Any action other than actual stealth actions like say, the vast majority of attacks and even social actions, hell even roll and action free communicate to allies type actions, trigger an involuntary stealth type roll to determine if potential observers detect you (with basically the same persistence and best observer rules gone over later).

Actions with the "silent" keyword get to do this roll without any notable penalties pretty much at about the same value as a sneak action.

Actions without the "silent" keyword (ie just short of all attack actions for a start) score themselves a big juicy penalty on their Cause A Disturbance roll, making detection more likely.

Actions with the "Loud" keyword (explosions, guns, or simply anyone who wants to scream for help during whatever else they are doing at no additional action cost) net another stacking big juicy penalty.

This penalty can rack up to more than enough to count for a few walls and doors and/or a fairly significant distance in value.

And this gives us a way for information about "Disturbances" (ie attackers) to propagate through the dungeon.

Killing The messengers
And then the secondary way for information to propagate is that some guy can damn well just run off and tell people, (or at the very least run off screaming to shift the source of his cause a disturbance rolls).

This gives characters good reasons to retreat and regroup, AND good reasons to try and stop enemies doing that (somehow).

There are other information propagation mechanics, but they fall into special skills and equipment and stuff like that (alarm devices in certain base rooms, telepathic communication, etc...). But hearing loud noises and having messengers running around gathering reinforcements are actually a pretty solid base to start from.

Hiding Vs Sneaking
As a minor eccentricity of the system I flat out decided hiding without moving should be easier than hiding while moving, so I split the most basic stealth action of the game into two variants, and gave the immobile hide action a significant inherent bonus over the mobile sneak action.

It helps among other things give low level characters a reliable way to interact with the stealth rules with other low level characters without too much early investment in stealth bonuses.

It also means that when characters get really desperate and try to hide it may become easier to guess where they are for searching purposes (especially if they just hid right in front of you and you have a pretty high observation defense). And if gives characters a reason to suddenly jump into a closet and wait cowering there while the guard patrol or marauding ogre lumbers past.

To leave a little extra guess work in the active search action it's explicit that when a hide or sneak action succeeds observers do not get to tell the difference or destination location of the sneaker/hider, I mean we don't want them to say "well it was a hide action so they are STILL IN THIS ZONE" or "well it was a sneak action so they are NOT STILL IN THIS ZONE" or "they successful snuk to that zone over THERE!". Because then they can just search it, or fire bomb it, and we want THOSE things to be potentially unproductive guess work.

Lots Of Rolls Issues
I don't care what your time scale for your sneaking around a dungeon mechanics are, lots of rolls issues are NOT as some idiots seem to think automatically associated with the mechanic being a "combat time" mechanic.

However ANY stealth mechanic is at risk of iterative complexity and probability if you do the dumb things and don't attempt SOME sort of mechanic design to reduce the numbers of rolls.

With the right mechanics you can just spend your action costs and NOT require rolls for every interaction and every turn of action, and without them you are going to have issues on ANY scale.

For me roll inflation comes from three main sources...

Too Many Stealthers
With the active roll for your stealth mechanic on the stealthing character every character who stealths could, and possibly should mean an active stealth roll (man that's like, three bad made up stealth words in one sentence and a title).

I decided as a general rule each stealther SHOULD make at least one roll each (not precisely per turn though, see Stealth Persistence below), essentially you have to draw the line somwhere and THIS is where I decided it was "worth it". One roll, per stealther, per stealth "event".

The only notable exception is that the same group mook type characters that get only one co-operative action attack roll together ALSO do stealthing together (so your squads of expendable mook ninjas don't cause massive stealth roll inflation in the same way they should not cause massive attack roll inflation).

The Best Observer Mechanic
So the next "too many stealth rolls" hit list item is too many observers. You almost certainly should not have to roll individually against ALL individual observers.

Comparing a single roll individually against all observers is acceptable, but also very possibly could be improved.

So as a base mechanic for stealth actions I decided that steatlh actions only bothered comparing against the BEST observer.

This excludes any character you were not trying to hide from (so bringing an observant ally doesn't screw you stupidly).

If you succeeded against the best observer, well, then we don't really care too much about the rest, additional comparisons are only made if the best observer does not choose to (or cannot) inform other observers about what they just detected (if so keep the roll and move on to next observer comparison until SOMEONE informs the others or until the roll succeeds against some remaining portion of observers who never get informed about it).

That certainly sped things up and clarified things by just formalizing the basic comparison process.

It ALSO provided a frame work for a mechanic that very slightly reduced the penalty for bringing a dumb crappy stealther with your group when all of you are stealthing together, because I tacked on a "each character can only be best observer once per turn, unless you run out of observers". So if your group is trying to sneak past three guys of differing observational skills your best stealthers can just go first and "use up" the best observers and then let your lame fail stealthers roll off against the lame fail observers. It's not MUCH of a boost, but it was a simple enough beneficial opportunity.

AND it also provided a nice frame work for distraction mechanics which basically allow characters to attempt an action that simply removes effected characters from the potential observer pool, potentially even allowing automatic successes for other characters attempting stealth actions.

The Stealth Persistence Mechanic
The third source of "too many stealth rolls" I decided on was clearly multiple turn iterative stealth. Rolling every single turn to sneak past just one observer or group of observers is just not right. It really should just be one roll, maybe somehow two or three if it the result was "a bit close" or something.

Roll in the stealth persistence mechanic. As long as you are performing the same stealth action as a prior turn with the same best observer you MAY retain your stealth roll result from the prior turn. You MAY reroll and you DO get to know if your retained result would fail before you decide.

So hiding from or sneaking past the same group of guards potentially requires ONE ROLL. (and will rarely produce even as many as three before you are either fully successful or detected).

If you roll "close" and they keep approaching you or you keep approaching them and distance bonuses on your stealth action would reduce your net result using the retained "close" roll to a failure, then you DO get a nail biting additional roll to succeed or fail

If you move on to a new group, its a new roll.

If a new better observer shows up it's a new roll.

If a search action occurs (lets say you successfully hid while in plain sight and searchers decide to spend an action and guess your general location well enough for the search to apply) or someone else otherwise causes the Alert bonus (say your ally jumped out and stabbed one of them) it isn't actually a new roll, but if it boosts the opposing defense your persistent stealth result compares to and you MAY roll again for a chance at success if you need it.

If you switch between (immobile) hiding and (mobile) sneaking, it is a new roll.

And all those implications from that simple initial persistence mechanic.

Which notably you could very well want to apply some sort of equivalent of REGARDLESS of your time or distance scales.

Other Uses Of The Base Stealth Mechanics
Once you have your base stealth mechanics, like which side gets active rolls, some sort of common observational defense, some sort of common alert bonus etc... you can use them for all sorts of things, and a great deal of the rest of my system gets to touch on those things.

But many of the base stealth mechanics also carry over to things like hidden traps/hazards, concealing items, stealing items, mechanics for a disguise action, and a number of social special actions.

Similarly, and people may hate this because they are assholes who hate fun, but even the stuff about information propagation and alert states gets to interact with the social stuff ALSO being integrated into combat time/mechanics to provide options in the social actions for attempts at managing information propagation and preventing or reversing the alert state (and the distraction thing already mentioned).

Also there are some fairly predictable if simplified lighting and fog mechanics
That thing this heading just said. You should probably try to account for those sorts of things in any stealth system, and doing it with mechanics in common with combat means that it can and should roll into or at least share common terms with related issues attacking and defending in such circumstances.

Fuck you Phonelobster and your talking in detail about your somehow still totally mysterious "Secret" rules!
And, you know AGAIN, as previously I've really covered enough here that you DON'T need to see my actual rules set, but, you know, it probably needs preemptive pointing out, again, that isn't actually secret and you can totally find the damn thing in a thread in the It's My Own Invention forum.

The thread over there is dated, and may not have the entire set of mechanic described here (I don't even remember), but the vast bulk of them have been integrated into my rules for YEARS now so yeah, they are basically all in there.

If anyone inexplicably really wants to go bleeding edge I've been working on converting to a Google Document as part of my latest revision and could readily arrange access to that (I probably will eventually over in My Own Invention anyway even if pointedly not asked).

[violence in the media]

I'd be interested in seeing that Google doc, PhoneLobster.

[Reynard]

Prak:
> norse terms for musical tempos
There are none, but you could use Norse poetry instead (drapa/stef/visa - long poem, refrain, and single stanza, for example). However, those terms are even less known than Adagio/Andante/Allegro (i.e. ergonomics goes down the drain). Using Saga/Tale/Visa (or Verse) instead would be better, even if less exotic/correct. [/smartass mode]

Anyway, generic terms are better. It's not like calling AoO "riposte" would drastically improve game.

> "Encounter," "Operation," and "Campaign"
I was thinking more along the lines of Combat/Encounter/Campaign.

Combat - because we don't really use any highly detailed non-combat systems.

Calling Combat phase Encounter is not really good idea, since encounters could be resolved without combat. In fact, even combat-related (assassination, for example) encounters could be resolved without switching (fully) into Combat phase.

Also, see below.



PhoneLobster:
> Every additional phase you add, and every mini game you add, multiplies the potential transition mechanics significantly.
Your point being ..?

You don't "add" phases. You already have them.

You can either create mechanics for phases or pretend there is no such thing and hope that it all somehow works out. But in the latter case you are forcing GM to create all those subsystems + transition mechanics on the fly once players will get to another phase.


The question here is how many phases should you recognize.

3-phase system essentially has Campaign-based-stuff-lumped-together phase, Encounter-based-non-combat-stuff-lumped-together phase, and Combat phase.

For example, you can easily split Campaign into Economy phase (crafting/management/logistics), Personal phase (healing/training), and Intrigue phase (politics/conspiracies); Encounter - into Exploration (sailing/overland travel/dungeoneering)[1], Awareness (stealth/detection), and Negotiations (bureaucracy/persuasion). And I didn't even mention mass-combat.

I don't think there is too much phases in the 3-phase system. So what's your point?

[1] - travel & sailing - keeping ship afloat, scaling mountains, not getting lost in jungles, and finding water in desert. Stuff like that. Logistics is in Economy.


> I've tried writing systems that transitioned between Combat/Social/Stealth/Chase minigames and others that worked with transitioning between Combat/Exploration/Travel time scales. One thing I learned was that almost ANY discrete transitions that require major rules to handle the conversion are too many transitions.
Or, you know, someone sucks at game design. That's also possible, no?



tussock:
> What players really want is something that just fucking works.
This. So many bad systems and crazy assumptions it literally makes my teeth hurt.

There are dozens and dozens of good board games. If someone can't design system, why can't he just steal stuff? I know, that might look a bit unethical, but peddling crap to players is objectively worse.




P.s. I don't think I like Stealth HPs (if I got them right).

If we are talking non-combat Stealth:

Having party's Stealth decrease doesn't make much sense. Alertness of guards has very few discrete states: 4-6 imo (Non-existent security/Relaxed/Alert/Possible Intruders/Intruders confirmed) and transition from Sleeping Guards to Searching for Intruders will not be gradual. But some sort of Detectable Presence rating could be numbers, of course.

If we are talking combat Stealth:

Having numbers for characters current Stealth is not something you can get away with in board game. As I already pointed out - 4x4 ninjas fighting each other will produce 32 variables (who is aware of whom). Keeping track of all those variables is hard. Making those variables gradual is simply impossible. That's why using "tags" (sneaky stances/detection manuevers/perception traits) is so much better. Not only they allow not to keep track of who is hiding from whom and how successfully he does it, you can even get away with having no (or very little) dice rolls.