My other thread about turning success tests with high thresholds into extended tests didn't go where I wanted it and I'm not sure it ever would. I realized, though, that the problem may not be the rules for rolling dice, but rather my execution of them. I have a different question, now, about opposed tests outside of combat. I repeat for clarity: combat related opposed tests are not what I'm talking about here.

In my games, I always let the players roll their dice for stealth (as an example), and use their hits to define a threshold by which the security attempting to see the stealthy players must beat. In other words, if the players roll stealth and get 4, 5, 3, 4, and 2 (these are hits for five fictitious characters as my table), the goons attempting to see them would need to beat one of those thresholds. It's likely they could beat a threshold of 2 but the threshold of 5 might be outside their capabilities. This seems to speed things up a bit because the PCs only have to roll once to represent an entire abstract amount of stealthy actions.

However, this does seem to create situations wherein characters (be they NPCs or PCs, doesn't really matter which) have to attempt to beat a threshold that doesn't change that could be quite high.

Do other GMs (or if you're a player, does you GM) ask the PCs to roll more than once for these sorts of things? Thus, if there are three goons attempting to see the characters, do the PCs roll three opposed tests or do they just roll once and then the GM rolls three times? I've always done the latter to move things along, but the former provides better opportunities for the dice to come up weird.

I wrote it already in the other thread. I have them roll stealth each time a guard has a chance of detection.

If they have to reroll occasionally, with rather certain odds of winning, noone knows when the unlikely will happen. But if you keep such a difference between runners and their opposition that they will usually not be noticed, you have to have so many rolls that the occasional failure happens. Ensure that everyone rolls stealth, the mods for distance are not so great that someone in the next room can´t be overheard. (For extreme uses, see Cains Shot in the Barrens).

Bear in mind, repeated rolls work against the player. Sneaky McNinja is virtually guaranteed to fail to sneak past a couple rent-a-cops if you make him roll for every step he takes. If he gets a good roll, he gets to make another roll. If he gets just one unlucky bad roll, he FAILS.
Okay, obviously that is an overly exaggerated case. But it makes the point. To put it another way, say you feel that a professional sneaky guy (6 dice) should be able to sneak past a professional lookout (6 dice) roughly 50% of the time. (please, please, PLEASE recognize the specific numbers here are not the point, just try to see where I'm going) If you give professional lookout 10 chances to see professional sneaky guy in 1 episode of sneakage, sneaky guy will fail virtually every time.
The solution for repetitious tests: Don't repeat them. Just make one test and then let it ride until someone does something genuinely different, not a new test every time McNinja moves behind a new bush.

I do one of two things in my game:

1) Player rolls one stealth, opponent (as a group) rolls one opposed perception. They either see him or not, and I use this when its not a key encounter and want to just move fast. I may apply teamwork tests, if the situation warrants. And obviously, they need to get one net success to "see" the player.

2) Player rolls one stealth, each opponent rolls one opposed perception. Again, needing a net success to see the player (as ties go to the iniatitor).

Now I will say that I don't roll dice in-game. I have a bunch of pages of random number sheets printed out and I just cross them out as I use them during my sessions (As a GM, rolling the dice is a chore, not as much fun as being a player).

But I guess, either way, I am only asking for the player to roll once...it just moves along quicker.

I would have everyone in a group roll (a group in this case is anyone directly traveling together). The lowest roll is the threshold (in this case, your 2 hit character) is your weakest link. Anyone that is stealthing around with him is at a higher risk of being detected. If you have the team split into two groups, anyone in the position to notice both groups would get two rolls, one for each group.

Now, that said... If a group takes their time, slowly sneaking their way into a position (extended stealth test), they'd get to accumulate hits (always based on the lowest hits roll for the group). I think it would be reasonable to say that anyone actively searching for them would be getting extended perception tests to try and find them.

Also worth noting: As soon as the team stops to do anything else, they compromise their stealth and have to start re-accumulating successes. And yes, that means that against someone that's been actively pursuing them characters, they will almost certainly be spotted. About the only active skill that wouldn't interrupt the extended stealth test would be a perception roll.



I clump NPCs into groups of similar ability to speed things up.

Say I have three security guards. They're all basically the same except for their names. Same gear, same skills, same attributes, etc. I'll make one Perception roll for that group. They'll get a teamwork bonus if they're actively looking for someone. Now, suppose that their boss is with them. He has better attributes and skills (that's why he's the boss). Depending on my mood and the situation, I'll either 1) roll him and the guards seperately, 2) have the generic guards give him a teamwork bonus, or 3) allow only the boss or guards to get a roll.

Every guard the team sneaks past is a potential roadblock on the way out. The chance that the team will be able to exercise the same amount of stealth on the way out is slim to none. And as they say, Slim left town.

Its fair to say, I'm probably not following RAW. At the end of the day RAW doesn't (and can't) cover everything, IMO. And that's okay by me.

I agree with everything you said; I'm not sure it answers my question, though. Let's make a team of 5 players and they're sneaking into a facility watched by three guards. Currently, I would have the following occur:

1. Each of the 5 players roll their infiltration test with all appropriate modifiers.
2. Then, I'd roll a perception test for each guard comparing their hits to the hits by the players.

This means that we'd only roll 8 dice pools (one for each player, one for each guard). What I'm wondering is if other GMs do this:

1. All five players roll against guard A.
2. All five players roll against guard B.
3. All five players roll against guard C.

Thus, there would be 15 total pools rolled. Probably doesn't take up much more time, though, since they wait for me to roll dice, in either case. However, because each player makes a greater number of rolls, there's an increased chance of glitches and failure for any given player.

I've always done the first but I'm not sure it's what I should be doing. And, we've noticed at the table that GM glitches occur far more often than player glitches (and dice pool size has nothing to do with it) but I notice that in the first scenario, I roll three times while they only roll once which could lead to this result.

I remember reading in the book that for perception checks, you do a group check by the goons. Take the highest rating perception amongst them, and add a +1 per goon, capped at his skill. Though for ridiculus numbers of goons, I'd maybe give him another +1 for every 3 or 4 goons beyond his skill.

I can't remember page, but basically lots of goons have a better chance of detecting you.

Team Teamwork Stealth test Versus Guards Teamwork Perception test.

NPC Guards Buy hits at 4 per 1 by default.

Agreed. Do a teamwork test for the guards. And 9 times out of 10 I probably wouldn't bother rolling for the guard team anyway, I'd just have them buy hits based on their full teamwork pool. Then each member of the runner team would roll against that threshold. Then my notes just look like: "Guards detect with 3 hits" or something.

Ah, ok. I count one encounter as one chance to perceive the runner. The guard only gets a new chance at perception if the players do something thats easy to detect. I´d just reroll if another guard came along, or the same guard another time. The perception roll will likely have some negative mods, and the stealth roll will have positive equipment/in case of runners augmentation mods. The relative advantage in mods changes the DP size ratio to 2:1 for the stealth-guy.

Edit: Within your options, each guard against the runners. One roll per guard for everyone at the table. Pretty fast.

I say you should make each player roll once, and then make all the guards roll together. If they're all in a cluster backing each other up, you might help them out with teamwork bonuses - 6 pairs of eyes are better than one pair. Using teamwork bonuses would also streamline play without screwing the guards -- if all the players get to roll for themselves, and the guards collectively roll once, it's very unlikely that the guards could ever, in a million years, spot all of the players. But a single test with teamwork bonus evens it out a little. Though I should emphasize that a teamwork bonus would only apply if the guards are working together in their guarding, not if they're individually patrolling, or watching in completely different directions.

As for whether players and guards needs to reroll perdiodically, I think that's up to the GM. Like if a player was hiding in one place for 5 hours, I might make them roll off against the guards once each hour. Especially long periods might even impose penalties, since the longer you try to hide the more likely you are to get fidgety or what have you. Though this could be negated with massaging liners, or digging yourself a nice hidey hole where nothing but the top of your head sticks out, or something.

I also think Ryu's approach is sensible. I would add that if the players, say, stop hiding and start sneaking, or vice versa, that would require a new test, since though covered by the same skill they can effectively be seen as two different modes of stealth. That isn't to say that sneaking somewhere and pausing to hide along the way as a patrol goes past would be two rolls, more like "I am hiding in the bushes" and then "I am now sneaking across the field" would be two seperate tests.

I say for an opposed test, each character, NPC and PC alike, makes one roll.

Also, it is not like only one Infiltration test is rolled on the PCs part for the whole run. That would be foolish. Instead, you roll a new test when you start to move or come out from hiding (or the situation changes). The infiltration roll for jumping the fence is different from the roll when crossing through the shrubs to the facility. Once at the facility, roll again to quietly climb the wall, etc. Each of these tests will have different situational modifiers and so each should have a different threshold for the guards/security system to notice.

The guards (if there are any in the area) should also get a new chance each time to spot the runner. This is a decent amount of rolling and can get quite boring, but I feel it is the most correct way to do things.

I allow buying of hits and I only ask for another opposed test when the modifiers change. If Bob the guard is standing in the middle of the garden and Joe Sneak is trying to infiltrate, as long as Bob and Joe in the garden, 1 opposed check is made. If Joe crosses the threshold into the building(modifiers changed), another opposed check is made. Most of the time, Bob will be buying the hits. If his buddy Pete is there with him, Bob may get a teamwork bonus from Pete who will also be buying hits. So unless sometime happens to trigger Bob and Pete from passive(simply buying the hits) to active(rolling) like an triggered alert, as long as Joe's hits are more than theirs, he goes through.

This may be totally against the rules and shit for stealth but I have an idea that seems like fun. This is from the matrix section and kinda goes back to what you were first thinking Dashifen.

Opposed extended tests!!
- Sneak Threshold: ninja rolls Stealth + Attribute and adds hits to stealth skill for enemy guard threshold.
-Ninja rolls Stealth + attibute (i forget heh) (X, some random number you create based on the situation) 1 combat turn
-Guard rolls Perception+ Intuition (Sneak Threshold) 1 combat turn.

This is cool because it adds pressure to the scene. Its not about whether or not the ninja gets past the guards totally unnoticed or sets off every alarm in the building. Its about can the ninja make it to the next area before he gets spotted.

a complication.. movement speed. first inclination is to hand out a die modifier to the ninja according to movement speed (sprint, run, walk, half walk). but that doesn't really make sense. a lot of stealth situations call for sitting tight for a minute until you get a window and then dashing to the next bit of cover. you could try changing the time to distance or "moves". so the guard would have 3 rolls to spot the ninja moving 30 meters. or the guard would have 3 chances to spot the ninja because there are three tricky spots that the ninja is likely to be spotted.

a thought about the guards perception tests.. angle of view/method of perception matters. lots of ways to model it, but maybe something like this: three guards are hanging out on a balcony trying to watch the ninja sneak through the garden to the house. ninja rolls a stealth threshold and each guard rolls a perception against it (or you can teamwork the guards). but there's another guard walking around the garden. this guard has a totally different perspective. so the ninja rolls a new stealth threshold for the patrolling guard. also, since it's harder to stay hidden from more than one direction, you might give the ninja negative modifiers. maybe -1 for each perspective over one. or the -1 could stack, so the first perspective is at -0, the second at -1, the third at -2, etc. kind of like dodging.

this brings up another point... what about modifiers depending on whether or not the ninja is aware of the guards. the ninja knows about the guards on the balcony, so he makes sure to stay in dark shadow or covered from above. but if the ninja doesn't know about the patrolling guard, he might not pay so much attention to moving silently. it's definitely easier to hide from someone you're aware of than not.