I think it should be a modifier to the Observers, because sometimes Observer A is invisible, and Observer B isn't. The Infiltrator rolls one test, and A gets a bonus to the opposed test because the Infiltrator can't see A.

I think the current general system is sound - one roll for the infiltrator, which all the observers compare to. Any modifiers related to specific observers should be applied to the observer's dice pool, thus not unduly influencing the odds for any other observers.

So Infiltrator rolls once, this sets the base threshold for all observers, which gets modified up or down based on other things (such as the infiltrator not being aware of them).

Ah, good point. I forgot about the 'setting the bar' system Infiltration usually uses.

That's the system now, and we should keep it, since it's a good system.

Really, the only thing we need to add is a bonus for undetectable (undetected?) observers, and that solves so many arguments.

So… are we having infiltrators roll Perception, then? And then the spotters roll infiltration. But we need to know if they can see each other to determine the Threshold, so they then each roll Perception and Infiltration, and then Infiltration and Perception…

I'd assume if someone isn't actively using infiltration, then the threshold to see them is 1. So the infiltrator peeks his head around the corner, trivially gets 1 success, sees the obvious guards. Then rolls infiltration to try to sneak past them.

Infiltrator gets 5 hits on his test, and has a chameleon suit, giving the guards a +2 threshold to see him, for a total threshold of 7.

Guards have 3 intuition, 4 perception, and +3 enhancement, for a dicepool of 10. None of them succeed. There is however a hidden camera, with a sensor rating of 6. Since it is hidden, there is a -3 threshold to spot someone unaware of it (such as the infiltrator in this scenario) and it rolls 12 dice. It rolls 4 hits against a Threshold of 4, succeeding at spotting the infiltrator. An alarm is sounded, and the infiltrator is in a bad spot.


I'd also say if the Infiltrator is aware of something but can't see it (example: Hidden Camera pointed out to him by the hacker, Watcher Spirit pointed out by the Mage, Invisible/Infiltrating Guard seen at an earlier point but no longer aware of position), then the threshold reduction is only -1 or -2, rather than -3. So in the above example, if the Infiltrator had been warned by the team hacker that there was a camera node in the room and gave a rough location, the infiltrator could try to avoid it better, taking a lesser penalty, and ultimately succeeding.

But you didn't give him a chance to spot the hidden camera. Luckily, it *can't* use Infiltration, aborting the loop. But, if the guards were hiding?

The Infiltrator's perception roll goes automatically against whatever the threshold of the hidden camera is. I handwaved it and assumed he failed.

Then the guards are using infiltration against a target they can't see. Whether they take the -1 or -3 penalty to threshold is the difference between them just trying to be generically sneaky (ie "I'm gonna hide in this corner and hope nobody sees me"), or if they're specifically trying to lay an ambush (ie "I know there's intruders that way, let's sit here and wait"). Typically guards don't do a lot of #1, but #2 you should see -after- an alarm has already been sounded.

Things like hidden cameras would have a perception threshold based on the type of sensor (a camera, needing visual line of sight, being one type; a heat sensor could be disguised more easily, etc.), it's size (a laser tripwire would be smaller than a camera) as well as a few other things, such as position relative to the infiltrator.* Objects are passively hidden and thus cannot move to get a better position if they are "aware" of an intruder.

As for hiding guards, it's too situational.

There are going to be situations where the guards should have the drop (i.e. spot the intruder before he spots them) and vice versa.

I would say that it goes in order of who does what first. The guard hide (1), the intruder hides (2), the intruder spots (3), the guards attempt a spot on the intruder trying to spot them (4), the intruder continues infiltrating aware (or unaware) of the guard's position (no roll), guards get another spot attempt (5, with bonus if the intruder is unaware of them).

*i.e. if it's in a corner of the room that the infiltrator can't see without sticking his head around the door and thus be in its view radius it would get a bonus.

Hehe, indeed, Draco18s. What a mess!

Perception and Infiltration are both skills that can be used either generally or specifically.

Perception can be used generally to spot something abnormal or to keep watch, or specifically to try and spot a specific person or thing.

Infiltration is the same. You can camouflage yourself, keep to shadows, and be generally sneaky, or you can specifically try to hide from an individual person.

In both cases, the specific use is actually easier than the nonspecific one. As such, perhaps a specific use of either skill should have a conditional bonus.

Now, when trying to sneak past something you don't know is there, you don't get that bonus.

Any GM can randomly call for Perception or Infiltration tests when they aren't needed just to keep a player on his or her toes.

That's something I can get behind: give a bonus for avoiding things you know about. That's the whole reason we do legwork, after all.

But then if you were crossing a room with hazards (A, B, C) you know about, and one you don't (D), would you roll two separate Infiltration checks, the A,B,C-check with a bonus? That's awkward - it's more rolling, but also you'll occasionally get the really disappointing situation where you ace the roll against A, B and D, but blow it against C.

So I'd rather have one roll for the infiltrator, and apply all modifiers relating to specific observers, to the rolls by those specific observers.

---

And yeah, there's the classic problem with ninjas A and B, where A and B are both initially unaware of each other, and trying to stay hidden. I'd go with an initial Infiltration vs. Perception test for both of them, with both getting the bonus for being an unseen observer. A succeeds in seeing B, then in the future B won't get a bonus for being an unseen observer.

Infiltration would be a bonus to the observer, rather than a bonus (or lack of one) to the infiltrator.

Statistically it's almost the same (for normal-sized pools, the difference is something like 0.1% due to the different odds of a glitch). But close enough for our needs.

Essentially, if you're going to be making multiple rolls:
If (legwork), then (bonus dice), else (not).
If you're going to be making one roll, with multiple opposing rolls:
If(legwork), then (no bonus dice), else (enemy gets bonus dice).

I don't know that I want Infiltration to be just a single roll. For a stealth character, it's what they do, and deserves some spotlight time.

(though multiple observers in one area might be handled like a group)

That's a point. Hm. Infiltration is one of those tricky ones. You could potentially roll it once for an entire 45 minutes of infiltration, skirting walls, avoiding guards, everything (I don't recommend this). Or, the GM can have you do it for each separate place/action (possibly too far in the other direction, but it feels like what you're doing matters). It's as if there were Matrix rules that just said 'roll to see if you get the paydata… success!'.

Hmm.

True.

Damn.

I'd say roll once per "level"; some dramatically appropriate part of the facility you're trying to infiltrate.

For example, the parking garage to get into the building CHECK, then the staircase system CHECK then the executive offices floor CHECK, then the office of the particular manager you're here to assassinate CHECK TO AMBUSH.

I have the idea that for every check that you don't win, there are different consequences; if you get spotted in the garage, you might be able to flee the building. If caught in the staircase, there'll be only low-grade guards. On the executive floor, the guards will be discreet but deadly. In the target's office, you'd lose the surprise attack.

Agreed, and I think that's roughly how people really play already? So, for each given step: we want them to roll a single Infiltration, presumably knowing how many hits they score. The GM compares that to his secret info about the Threshold of the task, and then against the Perception of each observer? At this point, the GM can apply a moderate bonus (+2 or 3?) to unknown observers, reflecting the PC *not* specifically avoiding them.

On the issue of infinite regression ninjas? If a PC and an NPC are literally playing cat and mouse in the dark? Hmm. Maybe just roll both ways and winner wins. Bleh. The 'screw it' method.

I don't get why you need to apply a bonus to dice pool for PCs not trying to avoid them. The same thing is accomplished by lowering the threshold, but with one big difference:

If the situation is reversed, and PCs are looking out for someone infiltrating, they aren't aware of that bonus being factored into the threshold.


Basically Threshold Mods keep that sort of thing in the realm of GM info, without having to deal with random dice pool bonuses. It also makes things a bit easier on the players, since they can almost always know what number of dice they are rolling, without having to worry about modifiers increasing or decreasing their dice pool.

We're having thresholds *and* opposed rolls here? I mean, we can, but personally I prefer to just oppose it, give observers a bonus as appropriate (with "threshold" of +X resulting in a bonus of +3X dice) and demand that the infiltrator [win/not] lose the test.





Ninja A and B both just roll simultaneously, and both get the Perception bonus for observing a target that hasn't noticed them yet.

Suppose they both notice each other; now they try to hide again from each other, and both don't get a Perception bonus, because their target is aware of them and actively trying to avoid them.

It's not difficult at all.

I'm going on the assumption that all tests of all kinds have Thresholds, per our earlier discussion (e.g., ranged combat). What if you're Infiltrating while X, Y, and Z negative conditions are in effect? Currently, it's -DP, but we'd discussed certain penalties working better as Threshold bumps.

Yes, Seerow, either way. However, you can't lower Thresholds below 1 or 0 very elegantly (usually, convert to auto-hits?).

How can they try to hide from each other after they've already spotted each other? It's not not difficult at all.

If the Threshold goes below 1, it's an auto success unless the character is incompetent at the skill. Yes that is possible, and yes it will suck if you're stuck trying to infiltrate and do badly enough you have 0 threshold to be spotted. But that happens sometimes.


As for trying to hide after they've spotted each other, I'd imagine that's a pretty large threshold mod for trying to hide from somebody actively watching where you are, and a lesser penalty for hiding from somebody who is aware of you but isn't paying much attention.

I assume it's impossible to hide while spotted. But sure.

Anyway: ninjas. Ascalaphus, you really just repeated what I said: the 'screw it' method, both get the same (lack of) bonus. It's a hack, but I think we can live with it.

Sure you can. You can hide behind objects.

The other person can't see you, but knows where you are.

It's not the same thing, but from a position behind an object you can, if the circumstances allow, you can make stealth checks to Go Somewhere Else.

Imagine being in a non-empty parking garage. You've been spotted, but you duck behind some parked cars, and quickly scramble away. Now they've got an idea of where to look for you, so hiding again is harder, but it's still possible






Yeah, but I don't think it's a hack, I think it's a genuinely reasonable and workable system.

Ah, see, I'd describe that as 'no longer being spotted'. You broke LOS, or hearing, whatever. Perception is 'do I see/hear/etc. them?', not 'do I have a general idea of where they used to be'.

I think that's really a sliding scale, depending on net Perception hits.

Really? You get a net hit on Perception, you don't see them? :/ I'm concerned about that. It differs a little for non-LOS senses, but still. The book says it means 'you have noticed something'; to me, noticing that guy there means looking at him.

It could be "1 hit: you have a general idea there's someone in that dark corner" and "X hits: you see the ninja in the dark corner"

Infiltration should be scene based. Look at it this way: Each roll you have a chance of failure. If one failure likely means that you've failed the objective, then each time you roll, you dramatically increase your chance of failure. That is the "forcing a roll until failure" method of dick-GMing. The only FAIR method is giving everyone even chances: spotters and infiltrators, everyone gets one roll per scene. Now the definition of scene is of course questionable, but once you know you need it, you can define something.

Also, I don't entirely believe all this difference between knowing the observer is necessary or not. Either you WANT a game of stealth, or you don't. You can theoretically make every run un-stealthable, and that is very un-desirable from a gameplay perspective. Making it hard to use infiltration will just reduce the game to MORE planning, because the previously highly abstracted infiltration skill is now useless, or you reduce it to more mohawk, because if you can't sneak, anyway, you might as well just shoot your way through.
Sneaking is IMHO already hard enough, and stacked observers give enough chance of being seen. (Unless you have so many successes that no present observer can indeed spot the infiltrator.)

About spotting hidden things, and hiding from hidden things:

An observer can't "observe in detail" while he doesn't know someone is there. He can't specifically keep a lookout for someone specific, unless he has an indication that someone is coming. He might hide in a certain place in order to get a better view about a certain avenue of approach, but in order to do that, he has to know about those approaches. A security spider watching 20 cameras which are all hidden to be nigh invisible will still have to watch those 20 cameras at the same time, until he knows which one he should be watching in more detail. Hence, giving an observer a bonus to spot a hidden character while he himself is hidden ist rather difficult.

So the only sensible method is giving a penalty to the Infiltration check. And it's also the transparent method - the GM says "ok, make an infiltration check at a -3 penalty", or whatever. There may be other conditions that cause that penalty: It's too bright, it's just a hard place to sneak about, etc.

The other transparent method would be to tie it to some other rolls or mechanics: For instance you roll a legwork/gather info or data search in order to find security measures in place, and then get a penalty for each threshold you fail to meet:
For instance:
Successes on data search/whatever:
1 (or x): General layout, floor plan, etc.
2 (or more than x, etc.): security doors, obvious security
3: security features (cameras, pressure plates, etc.)
4: magical security in place

The problem is that there are several ways to get all this info, and not all are easily gameable. One could also do an extended test, and see how many hits (with a different table, obviously) one has, and then apply modifiers. One might also combine results from several different tests in order to find the resulting final penalty.

Of course, simply ticking off the actual information aquired and then applying a penalty based on the actual situation is also possible, if more complicated and less transparent.
(So, for instance, you fail to find one camera and a patrolling spirit, and hence the GM decides to apply a -3 modifier for each, or just a -3 modifier, because they don't occur at the same time.)

The ninja question is best solved (as was said) without modifiers at all. You don't get to re-hide until you notice that you have been spotted, and then re-hiding works normally: You make another infiltration roll, and the spotter gets the benefit of being able to specifically look for you, that's all.

When trying to hide from astral observers that you can't see, I've actually gone through all the modifiers in the game at present, and found that actually, if you just have the astral observer roll perception (=assensing) normally, you are only making very minor (situational) mistakes of maybe 1-2 dice total (but this should be different for mages or others with active spells). So overcomplication is hardly necessary, and giving the infiltration penalty is enough. I would even go so far as to check what it is that is on the astral: Watchers have 2 dice (unless part of this revision makes them stronger), and as such, the simple act or intention of hiding means that they might not see anything. (0 hits vs. 0 hits and defender wins.) So basically, a watcher should not apply a penalty to infiltration at all. A smarter spirit, for instance one that could buy a perception hit (4 dice), should apply this penalty.


To reiterate:
The design goal here is to make infiltration reasonably feasable. This is achieved by:
- not stacking probabilities against the infiltrator
- not giving overly large penalties for infiltration
- not giving overly large bonuses to spotters, especially if they are just on normal non-increased alert.
- not making sneaking past astral observers impossible, and I would like to limit this to mundanes. Mages should have to buy masking or go in with their foci and spells off. There are perception modifiers in the game that also apply to astral perception/assensing rolls which already enable this mechanic, when applied properly.

I disagree. Intensely. It works badly with multiple observers: because Observer A is hidden, the Ninja gets a penalty to Infiltration, making it harder to hide from Observer B?

Remember that SR is also a game of double crossing. Its rules should certainly accomodate a PC spying on a clandestine Yakuza/Mafia meeting, where the Yaks and the Mafs have different odds of seeing the PC, and the hidden Yak doesn't make it more likely that the Mafia sees you.




And this then is the inverted way to do it. I agree that legwork should help, but instead of risking penalties, you remove the opposition's opportunities. You also don't know that you've missed something because the GM just told you that you get a -3 for something secret - if you missed it, you don't know about the bonus they're getting.






Maybe a 5th Ed change should be to give Watchers a bonus to Perception, so that they can, you know, watch. That aside, the matter of invisible guards has come up so often, and is really similar to hidden cameras, that it should simply be handled in the rules, instead of left as a reason for arguments.





I agree with all these statements, but I dislike your methods.

Giving infiltrators a penalty for ignorance, as you propose, makes glitches more likely, and also makes it harder to hide from everyone. However, giving a bonus to better observers doesn't make it needlessly easier for the less-elite observers.

The general table of modifiers for stealth should indeed not be stacked against the infiltrator. This basically requires that either Infiltration dice are easier to get than Perception (through skills, gears, attributes etc.), or that there are tactics and modifiers a ninja can exploit. Because if pools are roughly equal, and the ninja has to win Infiltration against five observers, his chances become pretty slim. This basically requires a good table of situational modifiers (that mostly make it harder to notice someone), and limits on Perception-boosting gear (the augmentations are particularly generous...)

I like your idea of making mages hard to hide on the astral. Might be interesting to make Watcher spirits specifically useful to warn about Awakened, while remaining rather blind to mundanes.

I'd take watchers back to 3rd edition. Their rules were not the same rules other spirits had to use.

I understood that hiding (astrally-active things) on astral is already really, really hard.

Yes, watchers are totally different and wonky.

In 3E, Watchers had 100% chance to find anyone anywhere. The only limitation was how long it took compared to their summon duration.

Other spirits had to make an opposed check (which they would likely make, but it was still opposed).

Well, 100% chance to find sounds a bit extreme, just like only 2 dice is a bit extreme. Maybe meet in the middle somewhere?

Okay let's try to codify infiltration rules:


1) A scene is defined as a place where things happen. You could have an entire building be a single scene, or each room can be an individual scene. Typically, a higher security area will have more scenes in a smaller area. So breaking into some guys house, one scene no problem. Breaking into a high security research facility, you're probably treating each room and hall as an individual scene.

2) At each scene, all parties get a passive perception. For every 4 dice they would roll on perception, they count as having 1 perception success. (Note: If the game switches to TN4, that should be for every 3 dice instead). A character not trying to hide, but not announcing himself counts as a threshold 1.

3) Any character may choose to actively perceive at any time as a free action.

4) Any character actively trying to hide rolls their infiltration skill, and adds hits to the threshold to see them. This must be rolled independently per scene.

5) If a character's perception fails to meet the perception threshold by only 1, they are aware that something is there, but not sure what without investigating further.

6) If a character who is infiltrating is spotted, they may attempt to roll another infiltration check, but to do so they must find cover, and suffer a -3 penalty to their threshold for being spotted again.

7) Perception Thresholds are handled on a per character basis. If you have two characters trying to spot the same infiltrator, they can have different thresholds, if different modifiers apply to each.


Bonuses and Penalties to Perception Threshold: (target references the creature/object/drone attempting to be spotted, for lack of a better term.)




Example: You have 3 guards in a room standing guard, and a rating 6 camera that has been hidden with a threshold 4. Each guard and the camera has 12 dice of perception. There is an infiltrator coming through this room with 12 infiltration dice, 9 perception dice, and wearing ruthium polymers. Our infiltrator rolls 5 successes, which gives them a total threshold of of 1+4+2 = 7. This far exceeds the guard's passive perception of 3. However the infiltrator is completely unaware of the camera, as the infiltrator has a passive perception of 2, and the camera's got a threshold of 4, giving the camera a -2 threshold modifier to spot him, making the threshold 5. The camera having nothing better to do rolls its active perception, and gets 4 successes. The camera is now aware that something unusual is inside the room but not sure what, and alerts the guards to be on alert. The guards now make a teamwork test, rolling a total of 16 dice, against a threshold of 6. They get a slightly above average roll of 6, and manage to successfully see the infiltrator.

The infiltrator, realizing he's been spotted, dives behind a nearby desk and makes another infiltration check, spending an edge and getting 7 successes, making his base threshold to be spotted 10. Unfortunately the guards are still looking for him, and know where he went, so they only need a threshold 7 to spot him. While he can probably make it out of this with a fair chance at not being caught, the area is now on high alert and it will be much harder to go on, the infiltrator decides to fall back and try again another time.

Ee, I dunno. AFAIK, the GM should roll their 'passive perception'. And it seems pretty complex anyway. Also, your Threshold labels cause cancer.

Mind, while the Watcher had 100% chance to locate, it still had to roll dice. Number of successes reduced how much time it took, and a low-force watcher would expire before that timespan had elapsed. There's a nice example in the magic book for 3E, where the watcher ends up having just enough time to send a message, get a reply, get back to the summoner and expire.

"Found her, she says, mreep!" *pop*

It seems complex, but it basically boils down to everyone gets to buy successes on perception except when they want to actively perceive, and infiltration gets rolled once per scene.

As for the threshold labels causing cancer.. what?

See, and I don't know if we want hit-buying a standard mechanic at all, for one thing. I know I'm not offering an alternative at the moment, time constraints.

They're the size categories from D&D.

Actually the D&D size categories are named a bit different and have different sizes for each category. I don't know them inside out but I know that my little chart wasn't an exact copy. But yes, you do need to identify size to some degree when talking about something like stealth and seeing things. That 20 foot tall anthropomorphic car is going to have a much harder time hiding than a regular guy, who in turn is easier to spot than a pixie.


On the standard hit mechanic, I like them as long as they're pretty much exclusively used for things that are passive or very low risk. It beast the GM having to roll 10 sets of dice every time the players enter a room to determine who sees who, or having guards who don't see anything unless they take a specific action to observe, which no human is realistically going to keep up every turn.

It occurs to me: If you dont like distinct size categories you could do it like this:


When perceiving an object larger or smaller than yourself, size matters. For every multiple of the perceivers size what he is trying to perceive is bigger than him, add a -1 threshold modifier. If the target is smaller, every 50% interval smaller than him adds a +1 threshold.


For example, you have a typical 6' tall human. Something between 3' and 12' has no modifiers for him to see. But at 3', it is now half his size, so it receives a +1 threshold. At 1.5', it is another 50% smaller, adding a +2 threshold. At 9", it is another 50% smaller, adding a +3 threshold. At 4.5" it is another 50% smaller, adding a +4 threshold. And so on.


The benefit of this is you don't have set size categories, which apparently cause cancer. It also means that two beings the size of a grain of sand could still see each other just fine, which is a problem with the size categories if you don't account for it. The downside is that this needs to be calculated out for everything perceiving, which is tedious, and has weird issues with cameras which are typically pretty small, which means unless they have a specific exception, they have a pretty big bonus to see most intruders. Personally I think size categories are simpler and easier but there's an alternative.


Of course you could just ignore size altogether and say that it is just as easy to see that pixie as the dragon.

IIRC, current edition has a rule where a bunch of people can roll Perception as one, instead of rolling separately, with the best observer getting a bonus for the other guys.

Do we like this mechanic?

On the one hand, it creates one very strong observer. On the other hand, rolling five times against a ninja increases the odds that one of the observers gets lucky. On the gripping hand, it shortens die rolling, particularly for NPCs.