The Scarcity of DRPGs

[arminius]

I think the points made so far can be summed up as: if you are tasked with performing at a level near your limit of ability, then success is a matter of luck from your perspective. Outside that range, it's more of a sure thing.

I've got no argument with that, except that it seems misdirected. The most interesting/important resolutions are generally ones where characters are performing near their limit. So eliminating randomness entirely seems to work against emulation.

A side point: games vary in the range of situations on which they impose random outcomes. E.g. CORPS lets you succeed automatically whenever your skill exceeds the difficulty factor, and makes you roll otherwise. Since IIRC there are no exploding dice in the game, there's effectively a "low range" of difficulty where you always succeed, a "mid range" where you can't fully control or predict success (as a tennis player trying to hit an exact spot), and a "high range" where you'll always fail. Actually almost all games have these three ranges, although some substitute certain failure/success with small to infinitesimal chances (critical rolls, or consecutive exploding rolls needed).

From a perspective of "realism" a greater range of uncertainty is probably more appropriate when there's less detail, since randomness substitutes for undefined variables. However as we've all pretty much agreed, randomness can also represent lack of conscious control (the tennis player, again).

Rock/scissors/paper-type games are a pretty good method of representing uncertainty against opposition. The old Magic Realm wargame used essentially this for combat, for example. I think it's a pretty valid approach...and not being anything near professional sporting ability, I can imagine that it might even be valid for sports. E.g. you could posit that a pitcher vs. batter is purely a guessing game of pitch selection and location, with the performance of a given batter vs. a given pitcher being the minimax value of mixed strategies on both sides. This would mean every time that Andy Pettitte pitches to Bobby Abreu, if Pettitte throws a cutter down and in and Abreu guesses same, the result will be a double down the left field line.

However, I don't things are really that deterministic. Basically, in sports as in combat, people are going to be operating near the limit of their ability unless there's a gross mismatch. Therefore a degree of randomness is warranted to represent lack of conscious control.

But I'll grant that this approach is an okay model and possibly fun from a game perspective. The real problems with diceless (which I think JibbaJabba has at partly acknowledged) are that if you make the model detailed enough to really emulate the decisions involved, it becomes unwieldy. If you abstract too much, then either things become too predictable, or the game devolves into an easily solved set of mixed strategies that might as well be handled through dice. (In other words, if we're really playing RPS, then if I pick my strategy at random with equal weights, we might as well just flip a coin to see who wins.) And finally, as an extension of the unwieldiness issue, if you have combat situations with a variety of multiple opponents with different characteristics, it gets even harder to design and understand.

[Kyle Aaron]

This covers most human activities - a given individual is usually either consistently capable of an action or incapable of it. In everyday life we don't expect the same individual to turn up for work only to discover that they're incapable of tasks they performed perfectly successfully the day before, yet random dice rolls suggest that this could be a reality.

Only if the GM is stupid in how they interpret dice rolls.

People have consistent performances. However, while people don't go from capable to incapable overnight, their performances do vary. The person who is good today might only be ordinary tomorrow, or might be excellent. They will not be crap or outstanding. They are usually good.

However, their coworker is usually ordinary, sometimes does poorly, and sometimes does well. They are never outstanding but sometimes crap.

Both have consistent performances, however they are not simply "capable" or "incapable." They don't either have the skill or not have the skill, they have at different levels, and vary around those levels in their performances.

This is the reasoning behind systems like Classic Traveller. A 2d6 roll with the skill added gives us a varied performance around a certain baseline, the baseline shifting with the skill level. Skill+1 hovers around 8, Skill+3 around 9, and so on.

Thus, dice results need not be interpreted as simple success/failure, but can be read as varying performances. In some task you may definitely succeed or definitely fail, but succeed or fail in varying degrees. If I (Kyle) bake a cake, I will definitely succeed - but it could be just nice or great. If I try to pilot a plane, I will definitely fail - but I could fail and just land really roughly, or fail and crash and explode.

Unimaginative GMs will interpret all dice rolls as simple success or failure. "You have Surgery 80%, therefore 20% of your patients die on the operating table." More imaginative GMs will come up with more than two possible results for every action, and the dice will help in this.

The most interesting/important resolutions are generally ones where characters are performing near their limit.

Yes.

[David R]

Most gamers want to use dice because its a game. This gets equated to dice are an excellent way to predict events. This is not so.

Not really, at least IME. Most gamers want to use dice because it's fun. The randomness working for and/or against them is what brings about said fun. It's not about predicting events but rather the exhilaration that comes from seeing
the result of rolls translated into a narrative.

Regards,
David R

[Trevelyan]

Real life has a great deal of variability. Even top tennis players fault on serves. People make mistakes at work. The best programmers still have bugs in their code.

There is slight variation, yes, but that's not totally random. The top tennis players generally fault in their games when facing other top tennis players and pushing at the very edge of what they can achieve. But the top tennis players rarely, if ever, lose to second rate players, and I'm never going to beat a Wimbledon champion, no matter how well I roll. Likewise, pretty much all programmers have bugs in their code somewhere, I'm not suggesting that success needs to be perfect, but they are still capable of producing a fairly consistent level of accuracy. Top programmers never turn up and produce something that absolutely doesn't work and is more bug than programme without knowing it, for example.

People have consistent performances. However, while people don't go from capable to incapable overnight, their performances do vary. The person who is good today might only be ordinary tomorrow, or might be excellent. They will not be crap or outstanding. They are usually good.

However, their coworker is usually ordinary, sometimes does poorly, and sometimes does well. They are never outstanding but sometimes crap.

I don't agree that people's general performances actually vary that much. Someone who is average is generally average >90% of the time. They might put in extra effort and do better than average on some occasions, and they might have an off day and do worse once in a while, but the reasons for that are usually external and not related to chance. If someone does worse than usual it's going to be the result of illness, or personal problems distracting them from work. It's not likely to be simply because they had an unlucky day and randomly did less well than they usually do. Likewise, if someone does better than usual it'll be because they put in more effort, didn't stop to chat around the watercooler in the morning and so on.

Both have consistent performances, however they are not simply "capable" or "incapable." They don't either have the skill or not have the skill, they have at different levels, and vary around those levels in their performances.

I didn't meanto imply that they were, you've taken me out of context. Clearly within a given skill some people can perform better than others, my point is that for a given task related to a skill, people have a tendancy to either do something or not, or alternatively they have a tendancy to do something to a given standard each time.

This is the reasoning behind systems like Classic Traveller. A 2d6 roll with the skill added gives us a varied performance around a certain baseline, the baseline shifting with the skill level. Skill+1 hovers around 8, Skill+3 around 9, and so on.

That's still a pretty significant range of results, though. a roll of 2d6+1 can give a result between 3 and 13. The GM might be able to interpret that result and supply reasons why someone who typically might get an 8 is now only getting a 3, but it's a pretty unlikely scenario in itself.

Thus, dice results need not be interpreted as simple success/failure, but can be read as varying performances. In some task you may definitely succeed or definitely fail, but succeed or fail in varying degrees. If I (Kyle) bake a cake, I will definitely succeed - but it could be just nice or great. If I try to pilot a plane, I will definitely fail - but I could fail and just land really roughly, or fail and crash and explode.

Again, you miss my point. I'm not suggesting that there are only two outcomes - land or crash - but that someone who is an experienced pilot and consistently lands well will not, unless subject to external influence, crash, nor will he likely even land badly. Likewise, your surgeon with an 80% surgery skill will tend to perform consistently on the operating table. Even rolling to see if he operates well this time or not is silly, because the factors which determine if he does better or worse than normal are external to the surgeon. All things being equal, a given individual will perform to pretty much the same standard each time. Better or worse performances tend to result from changes in circumstance which are better emulated through modifiers to fixed skill levels than by random rolls, even if those rolls are subject to modifiers.

Not really, at least IME. Most gamers want to use dice because it's fun. The randomness working for and/or against them is what brings about said fun.

I don't think that "people use dice because it's a game" and "people use dice because it's fun" are really that different. Part of the assumption is that a game should be fun, afterall. But essentially you are right, people don't want accuracy, they want a little randomness which feeds back into a somewhat unpredictable story.

[jibbajibba]

There is slight variation, yes, but that's not totally random. The top tennis players generally fault in their games when facing other top tennis players and pushing at the very edge of what they can achieve. But the top tennis players rarely, if ever, lose to second rate players, and I'm never going to beat a Wimbledon champion, no matter how well I roll. Likewise, pretty much all programmers have bugs in their code somewhere, I'm not suggesting that success needs to be perfect, but they are still capable of producing a fairly consistent level of accuracy. Top programmers never turn up and produce something that absolutely doesn't work and is more bug than programme without knowing it, for example.


I don't agree that people's general performances actually vary that much. Someone who is average is generally average >90% of the time. They might put in extra effort and do better than average on some occasions, and they might have an off day and do worse once in a while, but the reasons for that are usually external and not related to chance. If someone does worse than usual it's going to be the result of illness, or personal problems distracting them from work. It's not likely to be simply because they had an unlucky day and randomly did less well than they usually do. Likewise, if someone does better than usual it'll be because they put in more effort, didn't stop to chat around the watercooler in the morning and so on.


I didn't meanto imply that they were, you've taken me out of context. Clearly within a given skill some people can perform better than others, my point is that for a given task related to a skill, people have a tendancy to either do something or not, or alternatively they have a tendancy to do something to a given standard each time.


That's still a pretty significant range of results, though. a roll of 2d6+1 can give a result between 3 and 13. The GM might be able to interpret that result and supply reasons why someone who typically might get an 8 is now only getting a 3, but it's a pretty unlikely scenario in itself.


Again, you miss my point. I'm not suggesting that there are only two outcomes - land or crash - but that someone who is an experienced pilot and consistently lands well will not, unless subject to external influence, crash, nor will he likely even land badly. Likewise, your surgeon with an 80% surgery skill will tend to perform consistently on the operating table. Even rolling to see if he operates well this time or not is silly, because the factors which determine if he does better or worse than normal are external to the surgeon. All things being equal, a given individual will perform to pretty much the same standard each time. Better or worse performances tend to result from changes in circumstance which are better emulated through modifiers to fixed skill levels than by random rolls, even if those rolls are subject to modifiers.


I don't think that "people use dice because it's a game" and "people use dice because it's fun" are really that different. Part of the assumption is that a game should be fun, afterall. But essentially you are right, people don't want accuracy, they want a little randomness which feeds back into a somewhat unpredictable story.

Another great post.

For the non-fans of Diceless games we aren't saying dice aren't great. They are, love 'em. But you shouldn't think they are an accurate representaion of the actual way personal skill and performance work. They might be more accurate if your skill was 75 +1d6 % as opposed to 2d6+1 but that would defeat their purpose as fun.

[Croaker]

There's also Alienoids, which, as an alternative, is entirely diceless

Far fewer things in this world are anywhere near as random as most RPGs make them out to be, so for any given contest between two individuals operating at anything other than a very similar level, the outcome is usually a forgone conclusion, and ADRPG model this quite well, albeit in a manner which many traditional gamers find disconcerting.

This covers most human activities - a given individual is usually either consistently capable of an action or incapable of it. In everyday life we don't expect the same individual to turn up for work only to discover that they're incapable of tasks they performed perfectly successfully the day before, yet random dice rolls suggest that this could be a reality.

This is what I HATE in Chaosium's system. 70% is a good skill, yet you're expected to fail 3 times out of ten. Most are at 40-50 (in average characters).
And yet, this managed to be a successfull and well-liked system o_O

Most games, IMO, are like this, in the fact that the random element is very important. If you're adding dice+skill, having 1d6 + average skill of 20 is very different from 1d10 + average skill of 5.
Yet, it is the later model that seem to predate. I don't understand this. While people make mistakes and all, usually, skill is more a factor that luck when doing something.

[flyingmice]

This*:

I don't agree that people's general performances actually vary that much. Someone who is average is generally average >90% of the time. They might put in extra effort and do better than average on some occasions, and they might have an off day and do worse once in a while, but the reasons for that are usually external and not related to chance. If someone does worse than usual it's going to be the result of illness, or personal problems distracting them from work. It's not likely to be simply because they had an unlucky day and randomly did less well than they usually do. Likewise, if someone does better than usual it'll be because they put in more effort, didn't stop to chat around the watercooler in the morning and so on.

... and this*:

Again, you miss my point. I'm not suggesting that there are only two outcomes - land or crash - but that someone who is an experienced pilot and consistently lands well will not, unless subject to external influence, crash, nor will he likely even land badly. Likewise, your surgeon with an 80% surgery skill will tend to perform consistently on the operating table. Even rolling to see if he operates well this time or not is silly, because the factors which determine if he does better or worse than normal are external to the surgeon. All things being equal, a given individual will perform to pretty much the same standard each time. Better or worse performances tend to result from changes in circumstance which are better emulated through modifiers to fixed skill levels than by random rolls, even if those rolls are subject to modifiers.

... makes me feel you don't understand what that roll is supposed to represent in a simulation. That randomness *is* all those external factors, all rolled up into one and abstracted. All things are *not* equal, and the random factor is there to tip the equation. That doctor with a 99.99% success rate going into surgery may be coming down with a cold, the scrub nurse perhaps is upset, having just broke up with his girlfriend, and the anesthesiologist is possibly worried about an IRS audit coming up. Maybe there's an equipment fault, or an electronic glitch. That's what the roll represents. You can argue the amount of external random factors in the roll, but not the existence of external random factors.

Don't get me wrong - I like diceless too, it may not be less accurate, but it is not a more accurate simulation of reality. It's fun on it's own terms.

-clash

* Emphasis mine

[flyingmice]

Here's another way of looking at the surgeon with a 70% rating. It doesn't mean the surgeon fails 30% of the time, it means that surgeon has a 30% chance of allowing outside factors to interfere in what would normally be a successful operation, if such factors are present. The vast majority of the time, they are insignificant, but for this operation, they are significant. We normally don't play out the stuff with no chance of failure. For those things every qualified surgeon is at or close to 100%. It's the tricky ones that get played out. If there is no pressure from external factors, there should be no roll - and there isn't, unless you have one of those ass GMs who make PCs do a dex check every time they walk.

-clash

[Trevelyan]

This ... and this ... makes me feel you don't understand what that roll is supposed to represent in a simulation. That randomness *is* all those external factors, all rolled up into one and abstracted. All things are *not* equal, and the random factor is there to tip the equation. That doctor with a 99.99% success rate going into surgery may be coming down with a cold, the scrub nurse perhaps is upset, having just broke up with his girlfriend, and the anesthesiologist is possibly worried about an IRS audit coming up.

I addressed the applicability of external modifiers a little further down. in the second passage you quoted (empahsis mine):

All things being equal, a given individual will perform to pretty much the same standard each time. Better or worse performances tend to result from changes in circumstance which are better emulated through modifiers to fixed skill levels than by random rolls, even if those rolls are subject to modifiers.

In essence, if we know that the surgeon is coming down with a cold, that the nurse has split up with his girlfriend and the anesthesiologist is distracted by the thought of ending up like Capone then we can apply a penalty to the skill of the surgeon.

In practice, of course, we probably don't much care about the nurse or the anesthesiologist, who are likely nameless NPCs only vagely referenced by the GM on the understanding that such individual must be present while the PC surgeon undertakes the surgery. In that context it is therefore possible to retrospectively assume that a poor roll could be down to such factors, but that's really making a virtue of a necessity. Conversely, a good roll might be down to the fact that the nurse just got back together with his ex, and the anesthesiologist recently managed to bluff the IRS.

But consider the opposite situation, where all three participants in the operation are PCs in a game of Doctor: the Operating. Let's assume that these various personal traumas have been played out earlier in the session and the GM applies a -3 penalty to the roll. Using the 2d6+1 basic skill that someone suggested upthread, this would give a net roll of 2d6-2. Now lets assume that the surgeon player lucks out and gets a 12 on the 2d6. For shits and giggles, and to illustrate the point, let's also assume that in this hypothetical system, getting the maximum roll on the dice grants a further d6, and the player rolls a 5. He's now managed to pull off a net skill roll of 15, close to twice his average result of 8, and under pretty trying conditions. How does that makes sense?

Yes you can retcon it, maybe everyone was conscious of the problems they were facing and put in more than the usual effort so as not to let the patient die on the table as a result of their personal issues. But by that stage we're already accepting that the result doesn't match expectations and we're trying to justify it.

And that's really what we were doing in the first case, taking results which inherently don't make sense (perfectly competent surgeon undertaking a routine operation that somehow goes very badly) and then altering the assumptions surrounding them until they do. The fact that this is required is evidence in itself that the result is not what we would expect. That's not to say that you can't play a perfectly good game like that, but it's a game where players are expecting results to fall outside the norm due to events that, being retcons, are outside their control. It's fun to have that element of randomness, and most games do just that, but it's not a method which actually produces realistic results unless you narrow the range of dice results, or the circumstances in which you require a roll, so much that you might as well not bother.

Essentially, if you need to interpret the roll and create previously unmentioned external factors to justify it then you're already accepting that the raw roll doesn't make sense.

[Nicephorus]

They might be more accurate if your skill was 75 +1d6 % as opposed to 2d6+1 but that would defeat their purpose as fun.

You're forgetting about the greater central tendency of multiple rolls.  In an event is complex and interesting, it's rare to resolve it in a single action/roll.  This causes a small advantage to translate into a larger advantage over the course of the rolls. For example, suppose you were using D20 system (single die with a wide range of results=fairly random)  to model tennis.  Someone a skill of 5 could beat someone with a skill of 15 a quarter of the time.  But if you roll for each game of a match, they're almost never going to win the match.  If you roll for each point, it's even less likely.  If you use 3d6 instead, rare events are truly rare.

[arminius]

I covered all this in my previous post.

You want to emulate without dice? Then you need to track a lot more variables, many of which are hidden from the players. The fewer you track, the greater the variance of predicted performance so randomness should play a larger part.

In opposed contests it's also question of the range of possible ability compared to the significance of differences in ability. Where the significance is relatively high you should have less randomness; nevertheless two opponents who are closely matched will be effectively emulated by a random model. (Also to consider: the precision of the stats used; if you have low granularity then two people with the same numeric rating won't necessarily have identical ability, so a random variable again would substitute for the missing details.)

[Ian Absentia]

This is what I HATE in Chaosium's system. 70% is a good skill, yet you're expected to fail 3 times out of ten.

Under tense, trying circumstances, yes.  Under average* circumstances, the GM shouldn't be asking for a roll.

!i!

[*Edit: And perhaps average and environmentally controlled circumstances.]

[jibbajibba]

Wow Trevelyan is right again  We should form The Amber Consensus...

Because the vast majority of games have too much randomness they are not accurate simulations. They might well be great games but ..

The typical game situation is this ... Surgeon has 80% they can get modifiers from various elements the nurse performing the operation on a ship under enemey fire. We give a toal modifier in this case of -10% so 70% he will fail 3 times in 10.

Now a far more accurate system, that uses dice, would be the Surgeon has 80% the operation is rated on a scale of and each rank on the scale has a value. The surgeon get a modifier for the difficult situation say -10 then they add a smaller randomiser say a d10 and they have to hit the value for the Operation so mose operations those with a difficulty of 70 or less our surgeon will always complete under these difficult conditions for operations above 70% which will be stuff on the edge of their competance. So here we reduce randomness becoming more but not totally diceless. We could go further and itemise each variable and assign a penalty and then just compare the adjusted score to the target. Now the second one is maybe not as much fun you could make it more fun by having the GM not reveal the penalties and just play through the procedure.

So the problem vis a vis realism from dice based systems is that, for skill use certainly, the random factor is much too large, but this makes for a more exciting game. In diceless games they have been designed to draw their drama from other elements. A dice based system that adopts narrow ranges of randomness looses the drama the dice provide unless its been added in from elsewhere.

[flyingmice]

I addressed the applicability of external modifiers a little further down. in the second passage you quoted (empahsis mine):

In essence, if we know that the surgeon is coming down with a cold, that the nurse has split up with his girlfriend and the anesthesiologist is distracted by the thought of ending up like Capone then we can apply a penalty to the skill of the surgeon.

In practice, of course, we probably don't much care about the nurse or the anesthesiologist, who are likely nameless NPCs only vagely referenced by the GM on the understanding that such individual must be present while the PC surgeon undertakes the surgery. In that context it is therefore possible to retrospectively assume that a poor roll could be down to such factors, but that's really making a virtue of a necessity. Conversely, a good roll might be down to the fact that the nurse just got back together with his ex, and the anesthesiologist recently managed to bluff the IRS.

I'm not rerospectively justifying the roll, I'm talking about the reason the roll is there in the simulation. In practice we don't much care about the nameless NPCs, so we abstract their effect to a random roll. If you are going to work solely with modifiers, then you have to itemize every influence. It isn't worth the bother, so we abstract it to a roll.

But consider the opposite situation, where all three participants in the operation are PCs in a game of Doctor: the Operating. Let's assume that these various personal traumas have been played out earlier in the session and the GM applies a -3 penalty to the roll. Using the 2d6+1 basic skill that someone suggested upthread, this would give a net roll of 2d6-2. Now lets assume that the surgeon player lucks out and gets a 12 on the 2d6. For shits and giggles, and to illustrate the point, let's also assume that in this hypothetical system, getting the maximum roll on the dice grants a further d6, and the player rolls a 5. He's now managed to pull off a net skill roll of 15, close to twice his average result of 8, and under pretty trying conditions. How does that makes sense?

How the hell should I know? I didn't design this game, I don't even know if it's supposed to be a simulation of surgery at all. Maybe it's emulating a medical soap opera. Maybe during the operation the surgeon recommends a good tax lawyer to the anesthesiologist, the anesthesiologist hooks up with the nurse, and the nurse gives the doctor some vitamin C which heads off his developing cold. Not only is the operation a success, but everyone concerned benefits.

If I were designing this game as a simulation, then I wouldn't be using exploding die results in the first place. That would be just royal stupidity. If I were designing this game as a medical soap opera emulation, then they might be perfect for the task.

And that's really what we were doing in the first case, taking results which inherently don't make sense (perfectly competent surgeon undertaking a routine operation that somehow goes very badly) and then altering the assumptions surrounding them until they do. The fact that this is required is evidence in itself that the result is not what we would expect. That's not to say that you can't play a perfectly good game like that, but it's a game where players are expecting results to fall outside the norm due to events that, being retcons, are outside their control. It's fun to have that element of randomness, and most games do just that, but it's not a method which actually produces realistic results unless you narrow the range of dice results, or the circumstances in which you require a roll, so much that you might as well not bother.

That is not what I was doing. Maybe that's what you thought I was doing, but you are wrong. Here is what I am saying, in a nutshell: If the GM calls for a roll, it means unquantifiable external factors are potentially interfering with the usual, expected result. The factors are balled up and abstracted into the surgeon's roll. The roll is not "Is something going to go wrong" but "When the thing which is going to go wrong goes wrong, how well does the surgeon handle it?" If there are known factors which are easily quantified, the GM is free to apply situational modifiers as required.

Essentially, if you need to interpret the roll and create previously unmentioned external factors to justify it then you're already accepting that the raw roll doesn't make sense.

Essentially, you are hitting a strawman. You are missing the purpose of putting random rolls into a simulation entirely. They represent unquantifiable factors - unquantifiable either because they are so by nature or because they individually are below the resolution of the simulation. Quantifiable factors are always best expressed as modifiers. By taking out any random component, you are essentially stating that either all factors are quantifiable or that all unquantifiable factors are collectively below the resolution of the simulation.

Now whether anyone designs games as pure simulation is another point entirely. I certainly don't.

-clash

[flyingmice]

So the problem vis a vis realism from dice based systems is that, for skill use certainly, the random factor is much too large, but this makes for a more exciting game. In diceless games they have been designed to draw their drama from other elements. A dice based system that adopts narrow ranges of randomness looses the drama the dice provide unless its been added in from elsewhere.

Hi jibbajabba!

The problem comes from the first assumption, in which the dice are assumed to be rolled every time. If this first assumption is not true, then the random system is a much closer approximation to reality. If instead the assumption is that the 999 iterations that went without a hitch are not played out because they have no interest for the participants, then when in the 1 in 1000 occasion which is rolled for there is a 70% chance of success, then the actual odds of success are 99.97%. Adjust the particulars to suit your game.

Other than that, your analysis is spot on. Diceless games get their appeal from other factors.

-clash