Role-playing Games Stack Exchange is a question and answer site for gamemasters and players of tabletop, paper-and-pencil role-playing games. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I like the simplicity of BRP's percentile-based skill system. (Chaosium's Basic Roleplay, or BRP, is the system behind Call of Cthulhu and Runequest, among others.) It makes it trivial for new players, even ones that just walked up at a convention, to pick it up and go.

The problem I have with it is that I always hate how it makes you basically min-max your skills, because you need a very high skill to be able to rely on it.

Here's an example - one time, I was given a race car driver pregen character. He had a 60% Drive score, which seems decently high. However, in practice, once I was chasing some guy in a car, I'd go to make a corner and the GM would say "roll Driving!" in the 4 in 10 chance I'd fail, I'd wreck into a building or whatnot. I've had this exact same syndrome happen over and over. It's fine to toss a couple points into "Read Latin," but for any "adventuring" skill where failure has adverse consequences, you are not going to bother to attempt it unless you have a skill in the 85%+ range.

Does anyone have any good ways to mod the system to mitigate this?

I am not looking for "GM advice." Yes, ideally a GM will remember to apply modifiers to every single skill check, though in "roll vs skill" as opposed to "roll vs difficulty" games they rarely do, or they would make everything into a complex skill check so that the flat roll problem would get normalized. But the system itself doesn't really promote that.

After all this time, I am still unconvinced. I have run into other people that similarly note that BRP lends itself to swinginess, and though a GM might well be on the ball enough to prevent that all the time, the system sure isn't helping them do it. Still hoping for an answer besides "sweat your GM about it whenever they don't soften failures to your liking."

I was just listening to a Role-Playing Public Radio podcast and they were griping about the same thing. "A professional mountain climber with a 60% climb?!? I'd never even try to get up on that mountain with so low a skill." The default mechanic lends itself to "90% or nothing."

share|improve this question
As this is a system implementation problem among a reasonably large percentage of people who dabble in BRP games, it strikes me that no matter what you do you will end up having to teach/train the GM to use it effectively and appropriately. The best mod I can think of is to keep a record of rolls and judgements so that you can show the GM their ratio of success vs catastrophe so that they can learn not to treat failure as a botch, and to focus on adjudicating the task not the circumstances. Ex. Failed chase roll = failed chase, not crash. – Runeslinger Apr 13 '12 at 0:35

15 Answers 15

I think the best advice comes from Burning Wheel...

A failed roll doesn't mean a failure to perform, merely a failure to perform well enough.

A failed drive in that example chase should be merely that he opened up the range. Only a fumble should be a crash.

In your opposed drive rolls, it should be:

  • Both make, range stays the same
  • Both fail, range stays the same
  • Target makes and Chasser fails, Range opens
  • Target fails and chaser makes, range closes

Just reduce the scope of failures. What you've been using, make instead a fumble, and what you've been using as a fumble, delete...

share|improve this answer
Again, "GM advice" is fine, but I'd like to make it part of the system. There's a reason I've had so many BRP GMs that don't do this, even though it's the "right thing to do." – mxyzplk Jan 30 '11 at 3:56
@mxyzplk The trick is that the "GM advice" in Burning Wheel are rules, not ignorable advice. You can consider this answer as a new rule to add to the system. All systems tell you what a failed roll means; change that part of the system and it will perform differently. – SevenSidedDie Jan 30 '11 at 4:13
I kinda think this is the real solution, and that in the examples given the problem is that the GM is being, well, a little bit of a douche. Maybe not to the full-on Gygaxian adversarial-gamemastering "hurr hurr you rolled low YOU DIE NOW" thinks-Tomb-of-Horrors-is-the-greatest-module-ever extent, but a smidge toward that overall direction. – chaos Jan 30 '11 at 8:20
Yeah, the problem is that the system tends towards this because otherwise you have to have a complex set of these outcomes per skill... It's not that there are douchey GMs that do this all the time, it's that even good GMs misstep into it frequently out of sheer time-and-effort limitations. – mxyzplk Jan 30 '11 at 16:32
Given that most versions of BRP do NOT explicitly tell you the effect of non-combat fails, it's a matter of interpretation. And borrowing the approach from BW is in this case, merely a "wake-up" call that the mode the GM was using is likely not the mode Greg & Sandy had in mind when putting together the system originally. – aramis Jan 22 '14 at 5:20

Yes, these GMs are undoubtedly doing it wrong, as others have said.

But if you want more predictability and less failure, just roll percentiles but allow the player to choose which die to take as tens, and which as units.

This would give your character with 60% Drive only a 16% chance of failure or thereabouts.

A relative beginner with 30% skill would have about 50:50.

Won't suit everyone, but might be worth considering.

share|improve this answer
I like that... GMs remember to apply penalties/bonuses more if they are broken out into very rough difficulty classes. So maybe: Easy task - don't roll; Typical task - roll and swap favorable; difficult task - roll straight; ridiculous task - roll and swap unfavorable. – mxyzplk Jan 30 '11 at 15:42
@mxyzplk: Ahh, the UA flip-flop mechanic. I thought of it, but you'd already been unhappy about losing percentile odds. What UA lets you flip-flop is the skill your character is obsessed with, but I like your difficulty tiers. – chaos Jan 30 '11 at 21:00

The base skill in Speak Own Language is 30%. That doesn't mean that you fail to come up with words to say what you want to say 70% of the time in routine conversation (though there is a special mechanic here, see below). Instead, a 30% skill means that you are competent but unarticulate in conversation, and will perform better in conversation than most non-native speakers.

The 30% should be seen as setting what you expect the character as doing, and it should set up expectations for what happens with success and failure. You might roll if the character is expected to digest complex instructions without a chance to interrupt.

This is really to say that aramis is right when he says that A failed roll doesn't mean a failure to perform, merely a failure to perform well enough, and that the use of judgement here really is part of the game system as it stands.

Aside on language mechanics

In fact in cooperative situations like talking, the two talkers add their skill, so two inarticulate 30% speakers would have a 60% joint skill. But this doesn't mean that they fail to understand elementary utterances 40% of the time.

share|improve this answer
Though I agree again with another poster that this is how it "should be," the system lends itself to GMs just interpreting failure on a roll as failure. It requires the GM to have a complex mental model of what failing by 10% vs 20% vs 30% vs whatnot is on Driving vs Climb vs Read Latin vs... As a result, in play, it just gets ignored. BRP isn't a game system that pulls in the crunch-lovers so there's a certain deliberate lack of focus on doing lots of rolls and making sure you've got all fiddly modifiers applied anyway. I have yet to play a game of BRP where "fail roll=fail" didn't arise. – mxyzplk Jan 30 '11 at 15:47
The problem is the GM being a dick, not the rules being broken. That's not curable other than by confronting them about it. – aramis Feb 1 '11 at 3:52

This is basically why systems like GURPS go for randomization mechanics that approximate a normal distribution.

You could mimic GURPS and rescale the skill numbers from the 1-100 scale to 3-18, and then roll 3d6 against them. Or perhaps it might be interesting to try keeping the skill numbers the same and rolling 5d20 instead of d100. (Super bonus points if you can work out a way to roll 3d33. Given your username, I don't see why this answer should limit itself to the boring kind of geometry, the sort that's physically possible.) Or, 3d30+2d8-5 gives a (centralizing) range of 0-100.

The Unknown Armies approach of, rather than having a big laundry list of potential modifiers that pretty much guarantees the GM is never going to apply any of them, having three levels of criticality of the skill check, the less severe of which give you a nice hefty bonus, might also be useful. You're much more likely to get your +20% to effective skill if all the GM has to decide is that cornering a car is a Meh-grade task.

share|improve this answer

The old FASA Star Trek had a list of things that didn't need rolls.

at 20-30, one could, without rolling, perform all routine duties.

share|improve this answer
That definitely helps, though only clips the problem on the low end. The problem is less "I start the car and drive to the market" - "Roll Driving. You failed? OH NO YOU RUN OVER A GRANDMA!" (though I've had crap like that happen) but more the "your skill is 60% and if you fail you or your car or whatever is completely trashed." – mxyzplk Jan 30 '11 at 4:03
That's the GM being a real jerk, Mxyzplk... For drive skill, in FASA-Trek, 20 means you can drive without rolling unless it's freakish weather, or the first snowfall of the year, or a damaged car, or being shot at, etc. No roll, just automake. But those kinds of issues are why I gave up on BRP decades ago. – aramis Jan 31 '11 at 8:40
Hence my question about actual game mods - you are admitting the same thing I've experienced, which is that dice failure=task failure is the most common way of running BRP. – mxyzplk Apr 12 '12 at 16:12
@mxyzplk It's the most common, but (having reread RQ3 and EQ lately) "dice failure=task failure " is also the WRONG way to run it. It's neither the designer's intent, nor actually how the game says to run it. Newer versions may differ. – aramis Apr 12 '12 at 17:37
@mxyzplk - Aramis has the right of it, and newer versions do not differ: the message in the rules has been fairly consistently clear. The problem is one of not actually adopting the rules of the system, but instead assuming that the rules of other games (roll a D20 for pass/fail) apply. The solution is to address the lack of improvement running the game, not changing the game. – Runeslinger Apr 17 '13 at 9:14

I don't know if you'll interpret this as "GM advice", but it sounds like the people you've played with have come up with their own interpretations of the rules, or followed some wrong-headed advice themselves.

The general rule of rolling skills in Call of Cthulhu is, don't, unless you need to. A broad spectrum of actions is covered under the "Automatic Actions" clause on page 52 (of the 6th edition rules):

There is no need to roll dice to walk or run, to talk or see or hear, nor is there reason to roll dice for any ordinary use of a skill. (Emphasis added.)

But a chase, of course, is not an ordinary use of Drive. That still doesn't mean that failing the roll means the car crashes.

In general, the rules are silent on what the outcome of a failed skill roll is. Unless the skill mentions a specific failure result (e.g., failing a Climb roll means you fall), the consequences are up to the GM. Usually it's wasted time or resources--nothing that can't be surmounted. Many skills can also be re-rolled, but that's up to the GM as well. So a mere 40% chance of, say, Library Use, means that in a 12-hour day, you can check three times, which would give you a 78.4% chance of success.

As for your case in point, Driving does not indicate a specific penalty for failure. And there's an optional Vehicle Chase set of rules on pp. 284-5 that runs somewhat like a d20-style complex skill check. The notes are a little murky, but under "High-Speed Turn", it specifically states (p.284):

Success allows the turn without decelerating. (Emphasis added.)

...which implies that failure means that you have to slow down. A bad consequence when you're chasing someone, but not catastrophic. Nor, for that matter, a guarantee of failure. And even if you interpret the rules to mean that you have to roll on the Trouble Table, it has some results that don't indicate a crash, like Fishtailing and Skidding.

In short, the rules, as written, don't punish characters with skills under 90% as much as they appear to at first blush. If you keep the percentile rolls limited only to where they're needed, and the GM doesn't specifically have it in for you, a character with middling percentages in many skills can be as successful as a min-maxed one.

share|improve this answer
I find everyone harping on this unconvincing. There are all kinds of "adventuring" skills - climb, jump, drive, etc. - all of which are likely to have significant consequences when failed in an adventuring context. Also, opposed checks like the chase thing don't really help. Let's say you "don't crash," but if your Drive skill is less than 50% you will just end up losing chases or other stuff and if it's more you'll end up winning. Very high variability over the breakpoint, and CoC chargen isn't really tuned to having many skills that high. – mxyzplk Apr 13 '12 at 12:56

Roll thrice, take the median (i.e. middle value). Reduces variance, but keeps the mean and the range fixed, and does not involve too much extra rolling.

For stronger effect, roll five, 7, or some other odd number of dice and take the median.

share|improve this answer
Could work. Somewhat similar to "make all checks complex checks." – mxyzplk Apr 12 '12 at 16:16

60% driving..... think what that means. It obviously doesn't mean you can only drive 60% of the time. It might mean you're 40% 'better' than a driver who only has a 20% skill, or it might mean you're 60% able to fulfill those really critical tasks that crop up in emergency or stressful situations.

Firstly, I would add % values to tasks based on their difficulty. Walking down the street is something most of us do without problem, but occasionally we trip up. Walking down the street under the influence of drink is something I manage to do regularly with only minor variance from a straight line, Walking a tightrope is something I'd pretty much fail to do every time.

so, add % modifiers to the skill rolls based on how difficult you think it is, though I agree its probably better to treat the task as having a skill itself and you roll against the 'enemy'. So hiding in shadows is your skill roll v your opponent's roll for observation. Climbing a wall is your climb skill v the wall's 'steep and slippery' skill. Everything gets an opposing skill roll, which can increase the tension as you and the player are rolling dice, not just the player for success or failure. And if both succeed or fail, then you keep rolling, which in itself is great as it allows you to show things taking more time than expected.

This is already in the rules - combat works like this, if you attack and succeed, the opponent gets to roll defence and if he succeeds too, no damage is taken as your sword crashes against his shield. So why not use the same mechanism for every skill roll.

One thing to think about in BRP: Imagine 1000 warriors at the top of a hill ready to run down into battle, shouting and waving their swords in the air. 30 of them fail to make it having cut their own, or their neighbour's heads off due to fumble rolls.

That doesn't happen in reality. It doesn't happen because waving a sword in the air is a easy difficulty task. You might get 1 person who smacks the guy behind him in the face, but that'll only happen if the opposing 'waving' skill rolled a critical and the warrior failed his skill (or the opposing roll succeeded, and he rolled a fumble).

share|improve this answer

The skill ratings on BRP are meant for when the character is acting under pressure, either in a stressful situation like combat, or under a strict deadline. If the character has plenty of time to resolve a task, or the task itself is trivial, the skill rating should be doubled (this is called an Easy skill roll on the BRP rule book). That way, characters with skills higher than 50% will only fail if rolling a fumble on trivial tasks and/or when they have plenty of time to accomplish that task.

share|improve this answer
This is similar to the approach taken by Unknown Armies, which I was about to suggest myself. Skills are rated at the "someone's trying to kill you" level, not at the "just chillin' in my driveway" level. Upvoted! – Colin Fredericks Jun 1 '12 at 19:17

On BRP central, there is a nice 'wild dice' idea. Each session, each player gets 5 additional percentage dice. If they roll a fail and really want to succeed, they can roll one of their wild die and replace the previous roll. E.g. Player rolls 72, a fail. they roll one of their wild % dice, get 30 and replace the 70 to give a 32 score. The player can spend one or many wild die on each roll until they run out of their 5 wild dice.

share|improve this answer

As GM, I'd only ever ask someone to make a roll on something if success or failure matter, or make things more interesting.

Also, as stated before, "failure" does not mean "catastrophe". A failed Driving roll might mean "catastrophe" if you're attempting an action which could result in such -- trying to outrun a train at the crossing. But that failed Driving roll could also mean "Your nerve fails at the last second, when your driver's instincts tell you that you're not gonna make it; you hit the brakes, your tires squeal, and the locomotive rages past mere inches away from your headlights".

I know you don't want GM advice, but that's really the only good answer I can think of.

share|improve this answer
I didn't want gm advice. – mxyzplk Feb 4 '11 at 2:53
@mxyzplk The problem is you've given us examples which aren't supported by the BRP rules, and barely supported by the Call of Cthulhu rules. While the rules don't mention exactly what happens on a failed check, they do have a disastrous failure at 100. You've given us a disastrous failure on 61. A failed drive roll means that you fail to win the car chase, not that you die in a fireball. – Canageek Jun 2 '12 at 15:05

Get rid of the dice and use a computer to pick a random number from a Normal distribution. Pick what you consider to be the mean skill for a professional. Something around 30 or 40 I would say with a sensible variance depending on how extreme you want the results. It could maybe be linked to stress/sanity. Alternatively, the variance/mean could be linked to the difficulty of the task.

So, a drive of 60 in a car chase: pick N(30, 10). The same driver fleeing from a shoggoth, N(50, 30). Note that the numbers are not carefully thought out

Sure, it removes dice rolls. But now a days, most of us have a computer device capable of generating good enough pseudo random numbers.

Note that I still like Aramis's answer the best.

share|improve this answer
Potentially workable. Seems like you might as well dump the entire existing skill system and GURPS it though at that point. – mxyzplk Apr 12 '12 at 16:13
@mxyzplk: That is a fair comment. To be honest, I would just drop the system altogether but that's just me ^_~ – Sardathrion Apr 13 '12 at 6:13

I think part of the problem is that you're taking any failure in a disastrous way, when my understanding of the BRP rules is that above your skill is only a failure, while a 100 is a dipterous failure. However, as you "Don't want" GM advice (Even though this is a very simple fix);

Success tracks. Don't make everything hinge on one roll. Take combat for example: You don't just roll one die to see who wins. You roll your skills a number of times. So if you are in a car chase make a track: Lost other car <> Distant from other car <> Middle Range <> Close <> Caught up You start at middle range. Each success moves you up a range, each failure moves you down one. Each impale either moves you up two places. This way someone with a 60% won't fail on one bad roll. However, the longer you take to succeed the more times you have to roll, and the higher your chances of getting 100 and a crash are.

You can make similar tracks for any other event: Safecracking, climbing and so on.

share|improve this answer
Yeah, that's what I call a complex skill check in the question, it definitely helps. – mxyzplk Jun 2 '12 at 18:35
I'd say based on what you say, that this, a different game (As I'd count rewriting it to be non-%) or DM advice is your only option. – Canageek Jun 3 '12 at 4:26

I think you've run into the same problem nearly every failure-intolerant player reaches in a strict pass/fail RPG system. I have the same issue with a player in my D&D 4th group who feels he must min/max ever possible way into his hit rating because he can't stand to use a powerful encounter power and miss the enemy.

No matter what your min/max stats are, however, there is a very basic rule to any strict pass/fail system: the only two possible outcomes of a roll is a pass or a fail.

It doesn't matter if you have 85% skill, you can still roll a 14 and fail.

There is "GM" advice about handling failures, but that can feel unsatisfactory to a player who has expected outcomes of specific rolls. Instead, it sounds like you want a modification to the system to lessen the strictness of the pass/fail mechanic.

What I suggest for you is a pre-defined set of rules for variable passes and failures. With this rule change, the strict pass/fail is changed into a range of possible passes and failures with results the player can understand ahead of time.

For example, when the roll is:
25+ = Major Success
10-24 = Normal Success
0-9 = Minor Success
-9 - -1 = Minor Failure
-24 - -10 = Normal Failure
-25+ = Major Failure

You can spread these values out further, contract them, or add more tiers of them, to suit your system's need; but the basic principle is that it clearly defines what roll clearly describes a "failure" of epic proportions. Thus, what roll a driving check equates a crash, likely the Major Failure zone. This means that you don't need to min/max to 85% to feel comfortable in a car chase, you just have to define the "crash and spin out" to a comfortably small zone.

share|improve this answer
The problem isn't a fail-intolerant player, it's GMs who make every fail into a botch. If I was playing with a GM who would not let me fail but made me botch every time, I'd be walking away too. Just imagine if you fell on your sword every time you missed in D&D. That would suck some epic brass monkeys. – SevenSidedDie Jun 2 '12 at 18:14
Well, what is "fail intolerant" in a game like 4e is "realistically attached to your life" in Call of Cthulhu, where any opportunity to take damage is likely fatal (car crash, fail a jump and fall - not minor inconveniences as in a level based system). But yeah, a degree of pass/fail would help. – mxyzplk Jun 2 '12 at 18:39

Roll 2d12-2 and multiply by 5. Or take your chance of success divide by 5 and add 2. Roll equal or under using 2d12. It not perfect as you can roll 110% (22 times 5) but should do what you want.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.