BRP-Like WHIFF Factor

[PencilBoy99]

I ran my first long campaign using a d100-ish system and my players and I really disliked how wiffy it was. An average score in a dice pool system (because of the distribution) is more effective than a flat system (d20, d100). My players even after a while were at best around 60% in things. When you combine that with situational penalties OR even worse an opposed roll (like a parry), they failed quite frequently. While I don't mind failure, this was more uninteresting and frustrating than fun. Even when mostly only calling for rolls when "it mattered" the problem didn't go away (and in an "it mattered" situation you're more likely to have situational penalties).

 

[Fenris-77]

I've had this same experience with d100 systems. Most recently in a Delta Green game where my very competent character looked like one one the keystone cops session after session. I'm not sure why these small sample size rolling trends feel so much worse in d100 systems but somehow they do.

 

[SpellObjectEnthusiast]

I feel like situational modifiers should be used extremely rarely. Most tasks should be standard difficulty for the given skill, if you roll them at all.

Which game were you using? Most games' baseline assumes you will be playing basically Normal People. Mythras for example, by default caps pretty low for starting skills maximums and total points, and characters are similarly fragile. Characters will be far more competent when built with the Pulp or Paragon rules.

 

[ffilz]

Which system? I'm curious about the situational modifiers. RQ (at least 1st and 2nd) has relatively few situational modifiers.

One thing that is hard to get past with RQ style attack/parry is that if the parry isn't pretty effective (i.e. increases attack whiff) at the more experienced level, it can feel like a character can too easily be taken out since it just takes one decent hit to take a character out. This BTW is the advantage of escalating hit points, they reliably increase staying power of characters while reducing the feeling of whiffing. On the other hand, it can still feel whiffy when it takes 10 hits to put down an opponent, which extends to 20 attacks if the chance to hit is around 50%.

As to skill rolls, even a 5% chance of failure is going to be huge. Imagine you had a 5% chance of a car crash for every trip, adding in going out to lunch and various errands, you would come close to an expected rate of one crash per week. Since 1990 I think I've had 4 crashes (only one of which I had any level of responsibility for). Assuming an average of 15 trips a week, that's one in 15,000 or so, 0.0067% chance (hmm, looking at Cold Iron's normal distribution, that's 0.000067, which is almost off the chart).

The solution really is to only call for a roll when a real life person would almost certainly fail. Of course most people want to roll more often than that for anything other than a cinematic campaign, so we put up with some Keystone Cops moments. It can help if you cast most rolls as an opposed situation. For example, stealth is an opposed situation. You can't have it both ways that the PCs succeed in most stealth attempts AND PCs succeed in most perception of someone else's stealth attempts. Unless the opposition is always incompetent compared to the PCs.

And then understand that if combat is a meaningful part of the game, that sometimes the PCs have to fail. And that means either you use a single roll to resolve combat (compare two ratings and make some kind of roll, for example a RQ resistance roll) or if multiple rolls will be involved, understand that some or a lot of them won't meaningfully move the needle of resolution, or there will be lots of swinging back and forth of the needle.

 

[Brander]

I have a similar problem with most systems that have flat randomizers and low skill levels that are somehow magically supposed to represent professionals or even talented amateur's. I recall Unknown Armies had only roll under duress (meaning "in danger and adrenalized" to me) and flip-flopping for obsessions to help deal with this (and was the first d100 game I thought could avoid this to some extent). Savage Worlds cures this by having PCs roll Wild dice, use bennies, AND most people being assumed to have an edge in their field of expertise that gives a +2 which turns the usual target number from 4 to 2 meaning they really only fail on a roll of 1 before a wild die on most tasks, so even non-wild cards with no bennies are decent at what they do in their area of expertise (assuming a roll is even needed, I tend to apply "under duress" here too). Otherwise I tend to prefer bell curves of some kind as long as the odds are still in favor of success, though I'm somewhat OK with combat on a flat randomizer since combat should be wilder (which is why Sine Nomine games use 2d6 for skills and d20 for combat which I can appreciate).

 

[Raleel]

It's honestly very straight forward as to why it's like this - the chance to effect is smaller if you try to reach parity between attack and defense.

Consider, a 5e goblin has an AC 15 (https://www.dndbeyond.com/monsters/16907-goblin). Human fighter level 1 pregen downloadable (https://media.wizards.com/2015/downloads/dnd/HumanFighter1-10.zip) from WOTC has a to hit of +5. Which means he hits the goblin on a 10+, which is 55% of the time.

A strict opposed roll requires a 61 to hit and a 31 to defend to match that. https://anydice.com/program/35d46

So, your average heroic fantasy from some other game is MUCH more heroic. You must kick up the skill level to match what we have in D&D (and frankly, other games).

The other side of this is that generally successful effects are MUCH stronger in BRP family games. While it doesn't apply to your goblin (7HP, he's got a very strong chance to drop that goblin with his d8+3), hit point accumulation will impact this at later levels. It won't with BRP.

So you end up with a whiff whiff whiff SMACK without some careful guidance.

 

[ffilz]

I have a similar problem with most systems that have flat randomizers and low skill levels that are somehow magically supposed to represent professionals or even talented amateur's. I recall Unknown Armies had only roll under duress (meaning "in danger and adrenalized" to me) and flip-flopping for obsessions to help deal with this (and was the first d100 game I thought could avoid this to some extent). Savage Worlds cures this by having PCs roll Wild dice, use bennies, AND most people being assumed to have an edge in their field of expertise that gives a +2 which turns the usual target number from 4 to 2 meaning they really only fail on a roll of 1 before a wild die on most tasks, so even non-wild cards with no bennies are decent at what they do in their area of expertise (assuming a roll is even needed, I tend to apply "under duress" here too). Otherwise I tend to prefer bell curves of some kind as long as the odds are still in favor of success, though I'm somewhat OK with combat on a flat randomizer since combat should be wilder (which is why Sine Nomine games use 2d6 for skills and d20 for combat which I can appreciate).

Using a bell curve doesn't automatically fix the problem. The problem is when the probability of success is less than 90% the player is going to experience whiffing. Even with 90% they might still experience whiffing. Some players might not see 70% chance of success (so basically succeeding twice for every failure) as whiffing.

The problem is if you have any opposition, then the PC having a 70% chance of success means the opposition only has a 30% chance of success. Now some players want that. They want to always sneak, and never be surprised. That's a super-hero. And that's fine if that's the genre.

But I can make any of those probabilities work with a flat randomizer, in fact easier than a bell curve.

With Cold Iron's normal distribution bell curve, 70% means your rating has to be 3 or 4 higher than the opposition, not too bad. 90% means your rating has to be 8-9 higher. That either means the opposition has below average attributes and the PC has to have on the high end of what I look for PCs (absolutely average across only humanity would only net a 5-6 difference in rating) OR the PC has to be 5 to 6 levels higher. That means the enemy soldier is a recent recruit and the PC is well above starting level (fighter PCs start at 4th level in my campaign, a seasoned soldier would be 3rd level), so of the soldier is seasoned, the PC is a hero. Make the soldier 3rd level AND reasonable stats (should be above average), and the PC still has to be 3 or 4 levels higher. He isn't a starting PC.

It's not quite so bad for stealth vs. perception (level plays in at 2 x level instead of the 1.5 x level for combat skills).

What bell curves do accomplish is they put more numbers at the extremes of the curve so there's more granularity of ability that would produce at upper range chance of success. They also make an average result more likely which can help improve predictability if the chance of success falls in the middle, though mostly that only actually matters if margin of success or failure matters. Recall that ANY success/fail random result can be reduced to a chance of success that can be rolled on percentile dice (or finer resolution if desired). In fact that's what Cold Iron does. It uses a chart to convert a position on the normal distribution bell curve into a probability that can be rolled using decimal digit dice (percentile dice are decimal digit dice where you always roll 2 decimal digits, Cold Iron requires you to roll 2 or more digits depending on where on the curve the roll is landing). A 3d6 roll can be mapped to percentile dice, but is better mapped to d1000 (or 3 decimal digits).

 

[Fenris-77]

Frankly, I think a big part of the problem is people's mental treatment and associations with percentages (marks, dowloads, etc etc, so lots of associations people don't have with d20 results). D20 game are just a swingy in a lot of cases but also get less press for the same issues. It's got very little to do (or at least less to do), IMO, with comparative math or the actual chances of success.

 

[robertsconley]

I ran my first long campaign using a d100-ish system and my players and I really disliked how wiffy it was. An average score in a dice pool system (because of the distribution) is more effective than a flat system (d20, d100). My players even after a while were at best around 60% in things. When you combine that with situational penalties OR even worse an opposed roll (like a parry), they failed quite frequently. While I don't mind failure, this was more uninteresting and frustrating than fun. Even when mostly only calling for rolls when "it mattered" the problem didn't go away (and in an "it mattered" situation you're more likely to have situational penalties).

The bell curve versus straight line probabilities. Which I preferred using GURPS and had fun with Fantasy Age as they both used 3d6.

 

[robertsconley]

D20 game are just a swingy in a lot of cases but also get less press for the same issues.

Not with my group. The difference between d20 or d100 versus 3d6 and other bell curve probabilities has been remarked on numerous times.

 

[Fenris-77]

Not with my group. The difference between d20 or d100 versus 3d6 and other bell curve probabilities has been remarked on numerous times.

Yeah, again, I'm sure there's a huge range here from group to group.

 

[CaoCao]

In my opinion the probem is:

(Lack of shades-of-gray/partial success/success at a cost mechanic) + (lack of fail forward mechanic)

I.e. the problem is being basically a binary success/failure rule set from the late 70s.

The reason why 7e CoC tried to address that with pushed rolls, for example.

 

[robertsconley]

Using a bell curve doesn't automatically fix the problem. The problem is when the probability of success is less than 90% the player is going to experience whiffing. Even with 90% they might still experience whiffing. Some players might not see 70% chance of success (so basically succeeding twice for every failure) as whiffing.

I don't agree. While 70% odds are the same regardless, how that is achieved is very different in feel. With 3d6 or 2d6 you know that most of the time the numbers are going to be in middle of the range. GURPS this means between 9 and 12. This makes planning more straight forward.

However in your defense, the nature of the system can influence one's perception. In GURPS, dealing with things via skill rolls like combat or social interactions often takes multiple rolls. So, while you whiff in GURPS, the consequences are not always as severe as they are with an all-or-nothing setup like D&D's.

This was evident in GURPS versus Fantasy Age. Fantasy Age is set up more like D&D. So a whiff at 70% (and other) odds generally had the same consequences as D&D. But still, the fact that 3d6 meant that most of the time your rolls were 9 to 12 did have a small but positive impact.

 

[ffilz]

I don't agree. While 70% odds are the same regardless, how that is achieved is very different in feel. With 3d6 or 2d6 you know that most of the time the numbers are going to be in middle of the range. GURPS this means between 9 and 12. This makes planning more straight forward.

However in your defense, the nature of the system can influence one's perception. In GURPS, dealing with things via skill rolls like combat or social interactions often takes multiple rolls. So, while you whiff in GURPS, the consequences are not always as severe as they are with an all-or-nothing setup like D&D's.

I agree that there is a difference in feel for different randomizers, and 2d6 (which is NOT a bell curve, it's a "roof" line) and 3d6 do have some properties that can be leveraged if the math is done right.

How margin of success/failure is used can also play in and makes more use of the shape of the curve. How modifiers play in also can leverage the shape of the curve. But any single roll where all that matters is success or failure of the roll has a probability of success. Decimal digit dice can always be used to directly roll that probability.

But back to multiple dice... If you can influence the numbers such that middle values are part of the success (so 12 or less on 3d6 or 6 or higher on 2d6), then you can easily get to that 70% or higher chance of success. It may only take a +1 or +2, whereas needing an 11+ on d20 takes +4 to get to 70% chance of success. Going from a 10 or less on 3d6 to 12 or less goes from 50% to 74%.

Note that d100's attack/parry rolls turn the resolution into a 2d100 system...

Now some d100 games you start with skills well under 50%. That can feel very whiffy. But note that a 1st level OD&D character needs a 14+ to hit AC 5, that's only 35% chance of success.

TFT and GURPS with 3d6 makes shooting for a 10 or less on 3d6 easily in the realm of a starting character. I think even higher in GURPS.

Traveller with 2d6 requires an 8+ for success most of the time, with skill adding to that (so a character using skill is probably 70% or better chance of success).

So d100 systems really should target a starting PC having some 50% to 70% skills. RQ1's previous experience system will do that, though if you only qualify for militia, maybe not so.

 

[robertsconley]

So d100 systems really should target a starting PC having some 50% to 70% skills. RQ1's previous experience system will do that, though if you only qualify for militia, maybe not so.

Yes RQ/d100 starting skill levels always seemed to be on the low side. Especially when compared to Harnmaster.

 

[Raleel]

So d100 systems really should target a starting PC having some 50% to 70% skills. RQ1's previous experience system will do that, though if you only qualify for militia, maybe not so.

this right here. and most of them start you off as a regular dude. "Heroes" probably are professionals in their field, in the way we talk about them in RPGs. Pro normally starts around 50% in d100.

 

[robertsconley]

How margin of success/failure is used can also play in and makes more use of the shape of the curve. How modifiers play in also can leverage the shape of the curve. But any single roll where all that matters is success or failure of the roll has a probability of success. Decimal digit dice can always be used to directly roll that probability.

Except that mainly matters over time. Over multiple dice rolls the rolls will settle into the average predicted by probability.

But for an isolated single roll, a bell curve roll can be counted on falling into the middle range. That makes the difference for planning purposes. Especially if your skill odds of success lie beyond that middle range. It is not entirely rational, but some I observed in my decades of using GURPS and other systems.

If I had to guess it, both use a similar linear scale in terms of rating things, i.e., roughly 1 to 20. In GURPS there is a big difference between evaluating a 8 or less roll versus a 13 or less roll. Using a d20 13+ versus 8+ is not as dramatic. Now we know behind the scenes the bell curve makes 13 or less way better than 8 or less compared to d20's 8+ versus 13+. But at the table, with limited time to decide, it is an easier call to make with 3d6.

 

[robertsconley]

this right here. and most of them start you off as a regular dude. "Heroes" probably are professionals in their field, in the way we talk about them in RPGs. Pro normally starts around 50% in d100.

Except in real life Pros don't fail 50% of the time under normal conditions. Basically, in the last few years, as I get older and have more experience under my belt, I think most RPGs' way of handling skill difficulty is bullshit.

The fix is to qualify in detail when that 50% applies if it is to be considered a "professional" level of skill.

In my book, the time span in which to accomplish something has to be very limited, the danger level of failure that high, or resources are very limited before a professional (a journeyman) skill level is reduced to where the character only has a 50-50 chance of accomplishing the task.

If the professional employed at the factory shop I worked at had a 50-50 chance of failure for day to day skill use then it wouldn't be in business long.

 

[ffilz]

Except that mainly matters over time. Over multiple dice rolls the rolls will settle into the average predicted by probability.

But for an isolated single roll, a bell curve roll can be counted on falling into the middle range. That makes the difference for planning purposes. Especially if your skill odds of success lie beyond that middle range. It is not entirely rational, but some I observed in my decades of using GURPS and other systems.

If I had to guess it, both use a similar linear scale in terms of rating things, i.e., roughly 1 to 20. In GURPS there is a big difference between evaluating a 8 or less roll versus a 13 or less roll. Using a d20 13+ versus 8+ is not as dramatic. Now we know behind the scenes the bell curve makes 13 or less way better than 8 or less compared to d20's 8+ versus 13+. But at the table, with limited time to decide, it is an easier call to make with 3d6.

Right, I think GURPS (more so than TFT, but TFT to some extent) leverage the curve so a typical skill roll feels good and your observations about intuitive feel about the chances all work together. And yes, when the curve is being leveraged in that way, 3d6 works really nicely and it's much easier to get that feel than with d20. In comparison, to take another d20 system, Bushido feels very whiffy, I don't even feel good about making a 12 or less roll on the d20, and many rolls are 8 or less.

 

[Lofgeornost]

I think it's largely a matter of perception or maybe socialization/expectations. A few years back, I did an analysis of some of the games from the start of the hobby, say up to early '80s, and found that they tended to give starting characters fighting typical opponents a hit-chance of roughly 40%. Nowadays that's considered whiffy. I'm stuck in the past, so it's o.k. with me. I also have a little experience with sports where a 30% success rate would make you a paragon.

I don't think that percentile systems have any more issue with whiffiness than bell-curves, or dice-pools, except that it is blindingly clear what your chance of success is. If you don't find a 37.5% chance of success enough, you won't like the outcome anymore as 9 on 3D6 than you will on percentile dice.

 

[ffilz]

Except in real life Pros don't fail 50% of the time under normal conditions. Basically, in the last few years, as I get older and have more experience under my belt, I think most RPGs' way of handling skill difficulty is bullshit.

The fix is to qualify in detail when that 50% applies if it is to be considered a "professional" level of skill.

In my book, the time span in which to accomplish something has to be very limited, the danger level of failure that high, or resources are very limited before a professional (a journeyman) skill level is reduced to where the character only has a 50-50 chance of accomplishing the task.

If the professional employed at the factory shop I worked at had a 50-50 chance of failure for day to day skill use then it wouldn't be in business long.

I absolutely agree with you. I think rolls really should only be made in adventuring situations where there are stresses that maybe actually require heroic effort to succeed. Then a 50% chance of success logically is reasonable.

Part of the problem is we went from the OD&D 3LBB where the only rolls were for combat (and dealing with some extreme things in dungeons) to adding skills, and deciding PCs should be able to be carpenters, mechanics, and accountants. And then we wonder why the low chance of success that worked OK for combat in a dark and dangerous dungeon doesn't work so well for the mechanic trying to fix a flat tire. The mechanic fucking just fixes the flat tire. Even if it's dark and raining. In order to roll, the spare must have already been used, it's dark and raining, and you only managed to get a 5 minute head start on the bad guys.

 

[SpellObjectEnthusiast]

Except in real life Pros don't fail 50% of the time under normal conditions. Basically, in the last few years, as I get older and have more experience under my belt, I think most RPGs' way of handling skill difficulty is bullshit.

In BRP you don't roll under normal conditions. You only roll a skill under extraordinary conditions, like duress, extreme time pressure, lack of proper tools, or other similarly adverse scenarios.

 

[Raleel]

while i do agree that d100 only rolls when there is stress, I also don't think it is communicated well and that 50% starts to lose it's intuitive meaning when you say "well.... that's only under stress".

then on top of it, adverse conditions are also often listed as penalties!

 

[PencilBoy99]

The pushed mechanic is cool wish that was in more brp like systems.

As soon as you get into combat you have to roll, just not when it's challenging.

Also, the whole only roll if it's a difficult situation doesn't fit with those same systems having difficulty modifiers. If you're 50% good in a difficulty situation, why do the rules have me apply a difficulty penalty. Shouldn't there only be eases? Otherwise, that same person is really only 25% or whatever good.

 

[robertsconley]

First off, GURPS I feel has the same issue as I am about to criticize BRP for.

In BRP you don't roll under normal conditions. You only roll a skill under extraordinary conditions, like duress, extreme time pressure, lack of proper tools, or other similarly adverse scenarios.

Here is what the rules state





The problem, what is considered East, Average, Difficult, or Impossible, is defined either incorrectly or not well enough. And this problem is not limited to BRP either.

From GURPS


I find both sets of advice (and other system's advice) on when to make rolls and assigning difficulty unsatisfactory. My fix isn't to radically change how a system assigns skill levels. But to think very differently about what is trivial, easy, average, and difficult.

For example, over the past decade, I read or watched a lot of stuff on extreme mountain climbing as well as deep cave exploration. For the most part, my impression is that people quickly develop competence in the basic skills needed to handle these situations. When things go wrong, in many cases, most times* they screwed up the planning. They got themselves into situations where no amount of success on skill rolls would have gotten them out of because of a bad plan.

Now I admit I am biased in the way I referee. I tend to use skill rolls to resolve specific actions done as a character in pursuit of a plan created by the player. Rather than as a way of bundling up both planning and execution into a single abstract roll.

There is a high rate of casualties and injuries among even experienced mountaineers, and cavers, so a 50% "professional" rating may make sense if you just using skills in a very abstract way to resolve in one fell swoop whether the mountain was climbed or the bottom of a miles deep cave was reached.


*The other big source of screw ups are random events that are hard to account for. Unexpected flash floods in caves, the stratosphere is lower than normal when above the 8,000m mark, and so on.

 

[Raleel]

There is a high rate of casualties and injuries among even experienced mountaineers, and cavers, so a 50% "professional" rating may make sense if you just using skills in a very abstract way to resolve in one fell swoop whether the mountain was climbed or the bottom of a miles deep cave was reached.

as I've played more, I've begun to think that is the intension to some extent. We aren't going to get into the details, we'll just call that task X skill, which entails planning or ancillary tasks, and make one roll to cover all of it.

 

[ffilz]

The pushed mechanic is cool wish that was in more brp like systems.

As soon as you get into combat you have to roll, just not when it's challenging.

Also, the whole only roll if it's a difficult situation doesn't fit with those same systems having difficulty modifiers. If you're 50% good in a difficulty situation, why do the rules have me apply a difficulty penalty. Shouldn't there only be eases? Otherwise, that same person is really only 25% or whatever good.

Interestingly difficulty factors were added for many systems later in life. Classic Traveller (mostly) doesn't have difficulty factors (a few skills have a situation or two that is much harder to succeed). RQ1 and RQ2 don't have difficulty factors (other than a 20% bonus or penalty to combat rolls when advantage/disadvantage is at play, but there you've got an opposed roll situation). On the other hand, back in the day, we also made lots of skill rolls that probably were uncalled for given what the skill percentage represents.

 

[Mankcam]

The D100, whilst more granular, feels only a touch more swingy as a D20 to me.
However in modern D&D the PCs often have pretty good bonuses to off-set this in their favour
Most percentile-based systems are less suited for heroic and pulpy play, but seem to hum well for more serious-tone genres.

Things feel more challenging, so negative modifiers tend to have greater impact and bonus modifiers tend to be much rarer, depending upon the situation.

I find percentile core mechanics work well if paired with grim, gritty, or otherwise arduous genres. If these genres have Magic, then it tends to make a BIG impact on PCs rolls, rather than just adding or hindering a character's innate heroism like in more pulpy games.

That's my experience from most BRP games (and now also WFRP), but it's not backed up with anything empirical, it's just my phenomenological experience
(translate that as 'vague memory and gut instinct', heh heh)

 

[ffilz]

There is a high rate of casualties and injuries among even experienced mountaineers, and cavers, so a 50% "professional" rating may make sense if you just using skills in a very abstract way to resolve in one fell swoop whether the mountain was climbed or the bottom of a miles deep cave was reached.

What's the rate for mountain climbers when you remove the "tourist" climbers of Everest and such?

I haven't read the reports on cave accidents in recent years, but back when I was reading them, almost all of the fatalities were non-trained divers entering underwater caves. Now I did know someone who died in a dry cave, but I suspect perhaps his mental faculties were diminishing because several years before his fatal accident he told a story of nearly stepping off a ledge to rappel with his rappel device clipped into a belt loop on his pants rather than his harness and then his fatal accident was a failure to properly secure his rappel device to his harness. That strongly suggests to me that he was starting to lose his mental faculties.

Another chunk of the caving fatalities are due to having a medical emergency while in the cave. The rise of those incidents is probably due to the advanced age demographics with a pastime that didn't start being a big thing until the 1950s. And then also a part to blame on the decreasing fitness of the population overall.

When I was actively caving in the 1980s and 1990s, it was generally suggested that so long as you weren't cave diving, you were more likely to have an accident on the car trip to and from the cave as in the cave itself. But exploration HAS become more extreme since then.

 

[Mangoplum]

Except in real life Pros don't fail 50% of the time under normal conditions.

Depends on what situation you're talking about, and how you define 'normal conditions'. A quick look for statistics regarding police accuracy in armed shootouts gave me an accuracy rating of 35%. https://daiglelawgroup.com/new-study-on-shooting-accuracy-how-does-your-agency-stack-up/

 

[SpellObjectEnthusiast]

First off, GURPS I feel has the same issue as I am about to criticize BRP for.

Here is what the rules state

View attachment 80587

View attachment 80588

The problem, what is considered East, Average, Difficult, or Impossible, is defined either incorrectly or not well enough. And this problem is not limited to BRP either.

From GURPS
View attachment 80589

I find both sets of advice (and other system's advice) on when to make rolls and assigning difficulty unsatisfactory. My fix isn't to radically change how a system assigns skill levels. But to think very differently about what is trivial, easy, average, and difficult.

For example, over the past decade, I read or watched a lot of stuff on extreme mountain climbing as well as deep cave exploration. For the most part, my impression is that people quickly develop competence in the basic skills needed to handle these situations. When things go wrong, in many cases, most times* they screwed up the planning. They got themselves into situations where no amount of success on skill rolls would have gotten them out of because of a bad plan.

Now I admit I am biased in the way I referee. I tend to use skill rolls to resolve specific actions done as a character in pursuit of a plan created by the player. Rather than as a way of bundling up both planning and execution into a single abstract roll.

There is a high rate of casualties and injuries among even experienced mountaineers, and cavers, so a 50% "professional" rating may make sense if you just using skills in a very abstract way to resolve in one fell swoop whether the mountain was climbed or the bottom of a miles deep cave was reached.


*The other big source of screw ups are random events that are hard to account for. Unexpected flash floods in caves, the stratosphere is lower than normal when above the 8,000m mark, and so on.

You omitted the description of automatic actions, which is very important for contextualizing the others:

"Any activity that is so mundane, routine, or under the most favorable of circumstances and without any drama or conflict can be assumed to be Automatic, with no roll necessary to determine whether it succeeds. Everyday physical and intellectual actions attempted under average conditions always succeed, unless there is some reason they should carry the chance of failure. Your character should be able to perform Automatic actions at reasonable levels of competency in their chosen profession without needing to roll each time they wish to succeed—rolling for skills are at dramatic or difficult times, when success or failure is of importance to survival or destiny. Automatic skill use never yields an experience check."

 

[Acmegamer]

I ran my first long campaign using a d100-ish system and my players and I really disliked how wiffy it was. An average score in a dice pool system (because of the distribution) is more effective than a flat system (d20, d100). My players even after a while were at best around 60% in things. When you combine that with situational penalties OR even worse an opposed roll (like a parry), they failed quite frequently. While I don't mind failure, this was more uninteresting and frustrating than fun. Even when mostly only calling for rolls when "it mattered" the problem didn't go away (and in an "it mattered" situation you're more likely to have situational penalties).

This part of the reason I tend towards GURPS over RQ/BRP/Mythras family of games. Don't get me wrong I still am a big fan of that system mechanics, but I do recognise exactly your point which we were running into back in the late 1970s and into the 1980s from players.

 

[Acmegamer]

The bell curve versus straight line probabilities. Which I preferred using GURPS and had fun with Fantasy Age as they both used 3d6.

Hehe, I should have read further down the thread before posting.

 

[robertsconley]

You omitted the description of automatic actions, which is very important for contextualizing the others:

"Any activity that is so mundane, routine, or under the most favorable of circumstances and without any drama or conflict can be assumed to be Automatic, with no roll necessary to determine whether it succeeds. Everyday physical and intellectual actions attempted under average conditions always succeed, unless there is some reason they should carry the chance of failure. Your character should be able to perform Automatic actions at reasonable levels of competency in their chosen profession without needing to roll each time they wish to succeed—rolling for skills are at dramatic or difficult times, when success or failure is of importance to survival or destiny. Automatic skill use never yields an experience check."

Still not good enough in my judgment. Which I why use as an alternative.

I assume that the characters are competent in tasks involving their class and adventuring. Given time and the right equipment, they will succeed at what they try. Rolls are made where there a consequence to failure: the waste of valuable material, needing to complete something in a short amount of time, or the stress of combat.

This is a different standard than deciding whether an activity is that mundane, that routine, or that the circumstances are that favorable. If there is time available with the right equipment/resources, then success is all but guaranteed. I have a crit failure exception for certain tasks, but that's about it. Plus, the task be one that somebody of that general skill level would be competent at (apprentice, journeyman, master, grandmaster, etc.).

I understand if folks perceive this as splitting hairs. Think of it as the difference between legal proof standards. A "preponderance of evidence" standard versus "beyond a reasonable doubt" standard. There is not much between the two standards, but there is a difference.

For example, climbing a sheer cliff. If there is time and the party has rope and other climbing gear. Then the only thing I am looking for out of a skill roll is whatever the system considers a crit failure. Even then the consequences would be more embarrassing than damaging.

Most folks I know who runs GURPS, BRP, etc. would require a traditional skill roll.

Again this is not a specific criticism of BRP but how most RPG treat skill rolls. My view on this solidified after playing some campaigns Gumshoe and then trying out my own take on my own campaigns. I did this prior to Gumshoe but in ad-hoc manner not systematically.

 

[RangerRupert]

Frankly, I think a big part of the problem is people's mental treatment and associations with percentages (marks, dowloads, etc etc, so lots of associations people don't have with d20 results). D20 game are just as swingy in a lot of cases but also get less press for the same issues. It's got very little to do (or at least less to do), IMO, with comparative math or the actual chances of success.

Excuse me, but d100 is exactly five times whiffier than d20 ...

 

[ffilz]

Excuse me, but d100 is exactly five times whiffier than d20 ...

How so?

A 35% on d100 is the same chance of failure as needing a 14+ on d20.

On a related note:

The d100 is also no swingier for general success/failure.

In fact, d100 may be less swingy if you have automatic failure/success only on a 01 or 00 compared to auto failure/success on d20 1 or 20.

Swinginess doesn't exactly have to do with the dice rolled, it has to do with the relative probabilities of various outcomes.

A 2d6, 3d6, or other bell curve can be less swingy when the typical roll necessary falls within the central region.

Probability math is very tricky. And most folks don't intuitively understand it, and the type of randomizer can make it hard to intuit the probabilities, which contributes to how they feel. But in the end, every random rolls breaks down into a probability (or set of probabilities if there is anything more than success/failure based on the roll).

 

[RangerRupert]

How so?

In a Dad Joke kind of way.

 

[ffilz]

Also, probability math is very tricky.

Doubling a skill on d100 doubles your chance of success if the skill is 42% or less (assuming RQ always fail on 96-00). But the affect on chance of failure is not simple, it doesn't halve the chance of failure.

Adding or subtracting a fixed number to the skill or the roll has a different proportional effect depending on the initial probability. I.e. adding 20% to chance of success doubles chance for a skill of 20% but only increases the chance of success by 1/3 for a skill of 60%.

One place where probability math does have a consistent describable meaning... With Cold Iron's normal distribution chart, adding +7 to your rating improves your result spectrum by one standard deviation. In other words, if your average roll was going to be one standard deviation above average, and you get a +7, now your average roll is two standard deviations above average. But what that means as to the relative chances of success as a ratio is just as wonky as adding +1 to your d20 roll.

 

[Raleel]

Swinginess doesn't exactly have to do with the dice rolled, it has to do with the relative probabilities of various outcomes.

I think I disagree with that. I think the granularity of the roll plays a big factor in the perception of swinginess, even if mathematically it's on the same curve as a d20.

 

[ffilz]

I think I disagree with that. I think the granularity of the roll plays a big factor in the perception of swinginess, even if mathematically it's on the same curve as a d20.

Oh, all sorts of things fuck with perception... That's partly because many folks don't have a good intuitive sense of probability. And even if you do, things can still fuck with your perception.

Another interesting thing on probability. RuneQuest allows you to split your attack if it's 100% or higher. If you do so, your chance to fumble significantly increases. At 100% you go from 1% chance of fumble (only on an 00) to almost 6% (you fumble each 50% roll on a 98-00, .97*.97 = .9409). And you rate to have 6 times as many fumbles (the chance to fumble in one round is slightly less than 6%, the difference is the chance to have TWO fumbles in the round). Your chance to crit has also dropped from 5% to about 4% (each roll crits on 01-02, .98*.98 is .9604), though you have a slight chance of a double crit but you still rate to have 4/5 as many crits (the difference between .0396 is the chance of 2 crits). As your skill increases, the math does look better, but your chance of fumble will always be higher because having a 1% chance of a fumble twice per round produces more fumbles than a 1% chance of fumble each round.