So, when you retry a previously failed test, you are supposed to get a -2 dice modifier to it.
I find that rule pretty troublesome, really. What kind of test should have that modifier and what kind shouldn't?

The ones where you do not want the pc to just keep trying and trying and trying and trying and trying and trying and trying and trying and trying and trying and trying.

I myself used it to prevent the mage from summoning over and over again to get the max possible services from a spirit. The result was that he might try again once if the result was too bad, and that was it.

Otherwise I'd use it on non-extended tests doing what is in essence the exact same action. Not shooting at a target of course, but maybe Knowledge checks and Survival checks.

The -2 is only for failed tests. So a mage could indeed summon as many spirits as he likes if he gets at least one net hit on each test and copes with the drain each time.

And it's not like that's a free lunch. Conjuration drain can get pretty wild, so he's taking a bigger risk each time than say, recasting a low-force spell.

Yeah maybe I did that, don't quite remember. Although It's probably on successful tests the penalty would be useful, because when they fail it usually means the spirit gets alot of hits and thus drain, so enough punishment.

Most roleplaying games hate the concept of perseverance in attempting to reach a goal. "If you fail doing something on your first attempt, well, that's it. You suck. Keep trying if you want, but you're just going to embarrass yourself. I don't care what it is you're doing, you only get a decent chance on your first shot. Sucks to be you." It's one of the many things I dislike about most roleplaying game's design philosophies. Trial by error, learn by doing... they're all antithesis to most of the systems out there. At best, you can use experience to improve your abilities -- but nine times out of ten you still need an instructor to do so. At least by the rules that hardly any GM ever seems to enforce. And that does jack all when it comes to actually doing something important at the time you're doing it.

I especially love it when they add in silly complications for failure. For example, if you fail to pick a lock, the lock jams and door suddenly becomes impossible to open through any means (even someone else in the party trying to pick it)... until you gain a level and increase your lockpicking skill. Then, miraculously, the lock is no longer jammed and you can try again.

So, yeah, I tend to agree. It is a weird rule in a lot of situations. I'd prefer to have the time it takes to make the next attempt increased if anything. That would make a lot more sense. If you have the basic skills to perform the task and all the time in the world, I don't see any reason why you couldn't keep at it until you succeed. Unfortunately for runners, you rarely have all the time in the world.

Yeah that's true. However in SR there are extended tests for stuff like that. Picking locks, disabling Maglocks, hacking in general, searching for hidden nodes... even climbing. Actually I thought extended tests were too easy in SR to begin with, but in SR4A that has changed.

If there is only a single test then it is a task usually accomplished only in a single turn or a short period of time, and a test represent giving it your best shot. Sometimes trying again won't work, or failure have made it harder. For instance, if you're trying to con yourself past a guard and he doesn't believe you're the maintenance mad who just lost his card, he will be even less likely to believe you're really the security officer in disguise testing the guards. Also, if you fail a knowledge check to know if a certain gang contains trolls or not, making the same test again will probably not be any point. If not, why not just assume a perfect result from the get go?

Or do you really want to give a retry on a perception check they just failed?

Now if the action is such that it can be improved upon gradually and at last succeed, then you might as well make it an extended test or a "take 20" test. Sooner or later though you should force the PC to give it up and say "you can't do this", even if the player wants to roll all night. Maybe they just can't kick down the metal door.

For a passive test? No. But if they're actively searching an area, why shouldn't they be able to keep looking? Just add a time clause to each retry. First one may only take a moment. Second one takes a full minute. Next one takes five minutes. Next one, fifteen. Thirty. Sixty. etc. Or what, do you give up looking for your keys whenever you don' find them where you thought you last put them?

Yeah, I do. I lost my keys in the snow this winter, and after searching for it 2 times (about 10-15 minutes each time) I gave it up. I found them again when the snow melted recently.

That said my point was about passive Perception tests. Active ones could easily be handled like extended tests by the current rules. You seem to make it work like an extended test where instead of lowering dice pool or limiting the number of rolls, you increase the time of the interval instead. Even so that basically means the runners will automatically find anything given enough time. But people tend to give up eventually, and lose things and never find them again (or if they do they find them later when looking for something else entirely).

I'd handle it like an extended test with a basic interval, say 1 or 10 minutes (depending on how hard it is to find or if the runners know what they're looking for and not just searching for clues). Then I'd apply the -1 per additional test, so in the end the runners run out of dice and give up. If they have some sort of revelation while doing it I might consider resetting the test (in which case they'd also lose accumulated hits).

That's the thing. I don't see the point in lowering the number of dice in these situations, but extending the time between checks makes a lot more sense. Let's use your example. The key is there. You have the capability of finding it. The only reason you gave up was because you got tired of trying to find it. It wasn't due to a lack of an ability on your part to find the keys; it was just taking longer than you wanted to invest in finding them. If you had stayed out there longer and began being more meticulous in your search patterns, you would eventually find them. It's just a matter of time.

Additionally, that's not an Extended Test. An Extended Tests are used for actions that only take time to do, such as writing a program. You're all but guaranteed to finish, you just have to invest a huge chunk of time to get there. No real luck is involved. Instead, these are Success Tests that you have a real chance of failing at each and every time, but that doesn't mean you can't succeed given enough tenacity. But with the diminishing -2 dice on subsequent tests, that means you will eventually run out of options. Which doesn't make a whole lot of sense when dealing with something like looking for a set of keys. Or most situations where this rule would apply, though I can think of a few where it might.

Well in that case the keys were probably frozen at some point, but theoretically If I had the time and inclination I could find them. Problem is that players don't really spend the time trying to find things. They can say they search for one hour in a single sentence, and thus will never get tired of it. Thus you need a mechanic that represent boredom and giving up.

Also you're contradicting yourself. If finding the keys is guaranteed given time, then it IS an extended test. As you said "You're all but guaranteed to finish, you just have to invest a huge chunk of time to get there. No real luck is involved." There is little luck involved, the only luck is how quickly you find it, in which case how many hits you get on your first rolls.

Otherwise, I don't agree that extended tests are always infallible. Climbing a sheer cliff, hacking a system, getting open a maglock ... all endevours that can and should eventually fail. Programming is another matter though, but can anyone program anything? Should a person with Software 1 and Logic 1 always succeed in programming a rating 6 Hacking prog? No of course not. This guy will probably glitch, making bugs inadvertently, or a critical glitch which could mean ruining the whole program. Even a professional programmer should not automatically be able to program the very best programs. After all, there is a reason why we have ok programmers and excellent programmers, and why corps pay alot for the really good ones.

Exactly. Extended Tests don't cover that, but allowing repeated attempts in increasing time increments does.


No, because an Extended Test requires massive amounts of time. When looking for those keys, you could get lucky on your first, second, or five millionth attempt. It's a Success Test, just not one you're guaranteed to succeed at on your first attempt.


If you're attempting to climb a cliff and you have the skill and resources to achieve it, there's no reason at all you shouldn't be able to eventually do it. Same thing with opening a maglock, or even hacking a system. Though the latter has opposition forces and time constraints; the longer you attempt it, the more likely you are to get noticed. But that's an Opposed Test, not a Success Test.

Here's another example. Say you go to a firing range to play with a new pistol you bought. There's a bullseye on the range, and you want to practice hitting it. You have a perfectly fine gun, plenty of ammo, you know how to use a gun fairly well, and you're in a comfortable, relaxed environment. Are you really going to get worse at hitting it with each subsequent shot you miss (-2 diminishing rolls)? Are you really guaranteed to NOT hit it the first dozen times you roll your dice (Extended Test)? Like looking for your keys, it's not something you that becomes impossible to do, and it's not something that requires massive amounts of time to do in and of itself, so neither rule fits at all. The best you can do is ignore the diminishing -2 penalty and keep at it.

Didn't you get my point or did you simply choose to ignore it? Increasing time increments does not cause players to give up. In situations they are on the clock they may run out of time, but otherwise players can simply say they search forever. Unless you actually play out the actual time spent in game. If the players had to wait a whole hour just to get a retry, then YES they would get bored and give up. But searching 24 hours on a mission that they have a week on? No problem.
Extended tests on the other hand has the new mechanic of people loosing their edge and actually performing worse and worse at the same repetitive boring task. Once they hit 0 dice pool they will automatically give up for the time being.


Massive amounts of time? Interval can be as low as a complex action. Which means, a wired 2 character can do a ton of tests in a single minute and even unaugmented ones have 20 tests in that time theoretically. You can get lucky on extended tets too, especially if you use Edge. But it's more likely to take a few tests. If the threshold of an extended test is less than 4 (easy) I'd just say it's automatic success. Like finding a set of keys you just dropped on a clear floor.




Skills and resources is what gives you your dice pool. The mechanics decides if you're good enough, not the GM beforehand. If I did decide beforehand that climbing (say a ladder) would be very easy, then I'd not require any climb test at all. The climbing rules are broken anyway, always have been. It should be a success test with success meaning you climb X meters (based on net hits). Each new test would be a new action, so no dice penalty. Not that it matters according to Dakka Dakka, the penalty applies to failures. Failing a climb check could potentially have disastrous results anyway, at least on critical glitches. The only real danger in hacking maglocks is the anti-tamper system. The interval is low anyway so spending 4 turns instead of 3 will probably not matter anyway. But I like to think there exists maglocks that people COULD fail at hacking. Yes even without glitching.



Shooting on a firing range wouldn't require a test at all. It is training you do in your spare time. No extended test, no test at all. If it was a tournament it would be opposed tests probably or success tests with very high thresholds. And as said, no diminishing -2 pool when you actually succeed. The -2 on failures is optional IIRC and only meant to prevent players from abusing the rules and auto-succeeding on tasks that are supposed to be hard.

It's really not controversial, is it? The rule has a 'cooldown' left up to the GM's discretion; after waiting 1 minute, 5 minutes, an hour, any retry penalties evaporate. This is basically what the suggested 'increasing time penalty' idea above is, and it's already right there in the rules.

This seems to make sense, and we're clearly not talking about doors that are 'locked until you level up'. It's more of a 'frustration' penalty, because many people do get decreasingly effective after several failed tries at once.

The example of trying to con someone multiple times in a row fits with what my understanding of this rule has always been.

Obviously, all rules are optional per GM/player preference, but this one doesn't seem so terrible. Neither does the idea of a runner spending 24 hours on a task to get it right; who exactly is hurt if someone scours the matrix carefully for the obscure info, 'wasting' a whole day? If the task is not one that can be realistically repeated for that long, let the game reality stop them: you can't keep picking the same lock, because a guard patrol will eventually come along, etc.

Burning Wheel does trial by error. In order to advance your skills, you have to keep track of how many times you use them, the difficulty of the test, and your successes and failures. When you reach a certain ratio of successes to failures, you get better.

So why would that be bad?

Because it is unrealistic and detracts from my gaming experience.

Why if I REALLY wanted to find something in a building. I woudn't be able to repeatedly search rip out the floors the walls ect ect till I find it. If I tried that in your game as PC you'd just go fuck you? Sure I as a PC RPing would like give up after an hour or two in game if its not major but if its something important why woundn't I be able to search till there's nothing left to search?

Exactly. And the increasing time requirements means you, as the GM, have ample opportunities to bring in outside influences to interrupt the action. Be it a random encounter, the bad guys tracking them down, the cops showing up to investigate, or whatever else is appropriate.

Unlike your argument where you're trying to say it's "unrealistic" and a detriment to gaming, Yerameyahu makes a really good point. The rules do sorta-kinda account for it, and looking at the -2 as a frustration penalty until you take a break to catch your breath works pretty well for me. Especially since the net effect is the same. I think I'd rather there be a grace period between repeated searches, the penalty, and the cooldown though. Shooting a bullseye at the firing range isn't something that should give you a penalty immediately after your first shot, for instance, nor should looking for your keys. But after a token amount of time, sure, I can see frustration kicking in and penalizing you as a result.

When do you know when there is nothing left to search? Well after rolling your dice X times, when X= your dice pool (probably secretly), I would tell you you've searched until you're sure there is is nothing there. However unlikely, it is a possibility that no matter how long you search, you might still fail to find something hidden.

DR funk: I kinda forgot about the cooldown period, but that's not quite the same as your increasing time interval. "Let's come back tomorrow" might work fine in my game.

If I shoot someone and I miss, do I have -2 on my next bullet?

hi, I think that -2 should be added each time a test could got a penality...failed or not.

Physical action
Climbing once and trying it again, right away when you are tired needed a penality.

Same thing with trying social action
intimidation once ... whatever if failing or successing show that your point was bad, and you are trying sumthing new.

Logic action
Modifying/opening a maglock if failed need a penalty considering the fact that the pc is going to look at the instruction closely.
Also lets say he have 10 lock to open ... got it 3-4 times ... then got just tired of doing the same thing over and over.

And for magic, its harder to apply the -2, but lets just say that considering a willpower test, the pc think he should try again or that hes alright with less services,etc.

Using frustration as a reason, on a retrying test as to be proven by a willpower test. If the pc have 6 will ... he might just spend the night trying to do his stuff (as important as it can be !!!), if failed he just gave up really soon.

As you can see, i think there are a lot of reasons to used the -2 retrying penalty. And if the DM dont mind about a success, he just have to say that the pc need a few minutes to calmdown (or taking a few MORE minutes studying and analysing the thing) and then try again with no penality.

I would like to know what you think about that and any ideas how to apply retrying test ?

If you shoot sumbody and miss ... again and again, then you should and might get a penalty, giving up attacking an impossible target..... BUT ..... if your doing sumthing succesfully over and over, you could get a positive modification on you roll ... sound good to me ! Retrying test can be negative or positive.

No, it only applies to success test, not opposed tests.

IIRC Climbing is an Extended Test, so the -2 does not apply either.

As a GM, I might be bored too, if what the players want to achieve isn't important. At that point you could just handwave it and say "you take 20".

On the other hand, why is it unrealistic? If the character is that stubborn, well, that's the way they are. If there's no time limit, and success is important, why is it unrealistic?

Because it's not the character that is stubborn, it's the player taking the easy way out. A player will easily give up if something is taking a long time or he gets tired.

And why, as the GM, are you not bringing that up? And why, as the GM, are you not introducing ongoing difficulties and pressure for particularly long endeavors? I mean, if the character is at home fighting with his cable box so he can get free porn, who really gives a flying fuck that he's being stubborn? If he's doing something that actually means something, there should probably be time constraints and outside pressure. Even if that pressure is a completely random encounter.

And how are Extended Tests any better, especially when trying to do something impossible or extraordinarily difficult in the process, such as coding your own AI-level Agent with every software option in the game? Don't they get tired and frustrated, too?

Retrying may have implications (like you'll really screw up this time by glitching or critically glitching).

With extended tests, the -1 die for each interval makes extended checks a challenge (like slow hacking-which in SR4 was an auto win) rather than an automatic thing. It also leads to the possiblilty duplicated in real life where you may not know you failed until you have devoted significant time and effort into the project.

Yeah, but why do you want the player to give up? Why would you want to bore a player into giving up?
Roleplaying isn't just about realism, one of the nice things is that I can play a character who's less lazy than I am myself. That's not a flaw in the game.

If it's something that doesn't necessarily get harder if you keep trying it, and the character can succeed, and the player thinks it matters - well, why not let him brute-force it to success? Would it mess up the story?

Personally, I'd just go "It takes you all night, but eventually you succeed" if it bored me.

There might be difficulties, and there might not be. And no I don't use nonsensical "random" encounters just to interrupt the party's actions.

See Warlord's posts below about extended tests. The mechanics are already there.




I don't want to bore my players. Instead I let the mechanics decide the outcome of "attempt until success or..." Just because something is theoretically possible doesn't mean the players will automatically succeed even given time. Also, runners can be lazy too. They probably are usually, since they're doing crime and "get rich quick" schemes instead of honest work. Although it's not just laziness, it's also despair and fatigue.

Usually if the runners can try again without penalty if they have a new approach of attacking the problem. Something else than just "I rolled badly, I want to roll again." If that's all it is to it, I would not force them to roll AT ALL as I have already said.

The hell? What's wrong with random encounters? Shit happens. If you're pounding through walls in an abandoned building looking for something, who's to say you don't randomly hit a nest of devil rats or something? Who's to say a neighbor or passerby doesn't call Knight Errant about the disturbance? Who's to say that building isn't being used by a random street gang as a storehouse for drugs or whatever, and they're not too happy to find a bunch of punk asses tearing it up? etc.

I can't imagine a more boring GM than one who doesn't account for random events. Especially if the players are doing something suspicious or out of the ordinary, like obsessing over things like that. Doubly so when it's pretty much your only argument against allowing them to do it.

There's nothing random about it. I usually decide beforehand whether a certain building has devil rats infestation or not. If they are crossing the Barrens they are likely to encounter gangers. If they try to hack a maglock on a highrise located in a busy street, some people are probably gonna see it. If however they try to find a secluded spot, wait until night, and have lookouts to warn the team, then they'll probably be able to pull it off as long as they can beat the lock itself (which is not automatic).

Shouldn't players be the judge of how their characters' personalities work? If they want to play an absurdly motivated person, so what?

Is the thing they try to do important or meaningful? Then instead of stubborn, think "determined."

Is it unimportant? Does it bore you? Handwave the time it takes them to succeed; don't get them stuck on that issue. Give them something else to do.

What kind of things are your players trying to brute-force that gives you such grief?

No not 100% just like they can't just decide to resist poison, always be composed, always remember everything etc.

Otherwise. short answer, they're not. I just happen to use the rules in this aspect. I don't have to have a reason other than "it's the rules" but I do.

But usually it doesn't come up at all. This is not a problem in my groups and never has been. I think you are exaggerating the effects of a -2. it's not like it makes tests impossible unless they're really really bad at it.

But in other games I've seen a tendency of players to brute force their way through anything. A solid door? We bang on it until it falls down. Steel door? We destroy the wall instead. Nothing is an obstacle, and if we fail we just hit the "try again" button until we succeed.

Sometimes that works in my game. If the players have the right tools for the job and the obstacle isn't too hard, then they will automatically succeed, no roll required.

Those tend to be special cases.

Poison is mostly a physical thing, like damage.

Composure is something I'd be careful with as a GM. It can really take away the fun if the GM tells you how to feel, but on the other hand you might want to simulate surprise, or very aggravating NPC behavior.

Memory is another tricky one. In SR you can choose to remember everything if you have a simrig or something like it.
I've found repeatedly that expecting players to keep track of everything doesn't really work so well. RP is meant as a diversion, not as a second study.




In those cases there's a flaw in the system. It would be better just to state a minimum strength+tools to break down the door; if you don't have it, it's not going to work. If you do, a check is only necessary if you want to determine a secondary factor, like speed, stealth, or whatever. It doesn't make sense that you can only open the door with N successes, but that you need to try several times if you have N+k strength.
A "try again attitude" is actually the most sensible response when something that should be dependable - breaking down a door - becomes random, based on die rolls instead of simple capability.

Compose, Memory, and Judge Intentions are all handled by the rules. Simrigs, trideo and microphones of course matters, but that has nothing to do with memory and everything to do with recording.

I only use tests when it isn't clear that the PCs should be able to break down/open the door. And in this case it is a matter of chance. Think of Gandalf before the gates of Moria, it wasn't just a matter of try until success, it was a matter of saying the exact right phrase. They could have ended up spending a day there, and giving up trying to go another way.

IIRC, composure has notes that you shouldn't overuse it as a GM. In Vampire, frenzy tests have a lot of meaning because problems with self-control are a major part of what you are. In SR, I'd reserve it mostly for when people take specific negative qualities - I'll trust players to roleplay confusion, panic or rage when they feel it's warranted for their characters. If I disagree, that'll cost them the potential karma awards for good roleplaying.

A Simrig, cybereyes and cyberears and such would allow you to search the recordings for anything related to the matter at hand, effectively giving you advanced memory options. It certainly allows you to exactly recall NPC speeches from three sessions back (anything that was relevant, the GM should remember/have notes of).
Mind, I like it when the players remember what the plot was about, but this isn't school, there shouldn't be exams. I'd call for a memory test only when the PC wants to relate something long and involved he witnessed without recording gear, like the bad guy's diary he glanced through. The might be the kind of things that I didn't write/play out because they'd be long and boring, and only catching the important detail mattered.

Judge Intentions is sometimes tricky, if the GM is a poor liar - or too good. Generally my players just talk with NPCs, they rarely call for tests. If they do, they'll get a test, but I'll roll it in secret. (I generally roll tests to detect something unknown to the players in secret. If they called for the test, they can declare Edge.)



In LotR, it was a mental test to find the answer to a riddle; brute physical force wasn't going to cut it. So you'd tell the players that they just don't have the physical force to open the door that way; solve the riddle or take another route.