Death to the Expert: Probability and the Romance of the Primary Die

[David C]

I was running some sessions with "high level" characters.  These characters were legends in their own right.  But I had this nagging sensation that something was wrong. Then it hit me.

D&D uses d20s
GURPs uses 3d6
White Wolf uses xd10
Fate uses 4dF

In my game, you roll 1d10+skill to see your chances of success.  So each point of character advancement increases your chance of success by 10%. Right?  Well, actually almost all contests are against an opponent... who rolls 1d10+skill. I never had a linear success rate in my game. In fact, I had a 'triangle curve' from effectively rolling 2d10. This presented a problem because a player who had only put mild emphasis on "min/max" would succeed nearly every time, whereas the balanced character would almost never succeed.  All a character had to do was put 6 more skill points in a skill to have a 95% chance of winning. That's not a very big margin.

I have several options to fix this.
1) Instead of 15 levels, there would be 5.  Each level would mean becoming a lot more powerful. 
2) Instead of using a d10, I use a d20
3) Instead of the defender rolling their skill, they just add "5."

If I went with #1, it might have interesting results.  It might make "leveling up" feel more like a profound advancement.  Also, characters DO have other ways of becoming more powerful than leveling up. On the other hand, this just doesn't feel right to me.

#2 jumps out at me as being the best thing to do. It adds more discrete steps into advancement. This lets the GM throw around bonuses without worrying about them trivializing the game. On the other hand, there are special abilities that activate when a player rolls a 1 or a 10.  I'd either need to expand this to 1-2 & 19-20, or these abilities wouldn't come up in play very often.  I'm also kind of biased against using a d20 because that's so old hat.

#3 This could work, but I feel it takes some excitement out of playing. When the attacker rolls, you know you have a chance of rolling better than him...

Now here's the other thing. I've discussed this with all my play testers (there's about 20 of them).  Almost all of them don't like the idea of switching to a d20. Just like D&D has romanticized the d20, my play testers feel the d10 is right.  Either that, or they also feel like d20 is old hat?  I'm not really sure. 

I've been agonizing over this for days.

[Callan S.]

Hi,

Ron would probably say this far better than me, but the whole 'don't like D20's' doesn't seem to be rational at all. There is no practical reason for your or their feelings, it's just a hang up they have and you as well.

Yeah, I know were supposed to take into account what people don't like - but sometimes it's not about any physically disliked thing and it's just a hang up. Of course roleplaying is an area that kind of blurs what is physically being enjoyed (until people talk solely in terms of what they can do 'in the fiction'...)

[Eero Tuovinen]

Well, mechanical aesthetics are a real thing, they do really exist. So even if somebody disagrees or even thinks that the whole issue is stupid, that doesn't change the feelings of the guy doing the designing, nor that of his target audience (although it's unlikely that David's target audience is primarily defined by a disdain of the d20).

One solution here is to look for alternate dicing mechanics that preserve the d10 while decreasing the bell curve, hoping to hit on something that matches your mechanical aesthetics. The current system is effectively d10-d10, so how would you remove that likelihood for results to approximate zero? One method would be to have the opposing players roll d10s as before, but also have the GM roll a "consequence die". Aside from hooking whatever special effects you want on the consequence die ("on a consequence result of 5+ the whole hallway collapses" or whatever), you could state that whoever got the higher roll from their die gets to add the consequence die value to their roll as well. This might seem a bit strange if you're wedded to simple die+constant approach, but mathematically it fits your dilemma: the high roller will be rolling an average result of ~12.5 while the low roller will be rolling ~4, which increases the average difference between the rolls (and thus the average skill discrepancy sufficient for victory) by over four points. Should be enough, I imagine, and if not, roll more or larger consequence dice.

I do think that you would benefit from getting over parochial aesthetics in this matter, David. For one, your design work might be faster, more exact and more expressive if you learned to work with a wider range of expression. I don't think that it is possible to get away from mechanical aesthetics, but in my experience it's eminently possible to widen the range of your tastes, seeing more acceptable solutions for individual design cases and developing ability to match different mechanical aesthetics with different design cases. I don't know if this could be the case for you, but I myself have found that increased variety has helped me look beyond the details of dice math to what the players are actually doing in the game. This is helpful because it ultimately isn't very interesting and important whether you roll a d20 or d10 if that roll result is used in the same way by the game anyway. There are so many deeper issues in mechanical aesthetics that getting stuck on the size of a die seems like missing the forest for the trees.

Combining the above two paragraphs, my message is this: whether you use a "consequence die" like I suggest should ideally depend on whether you want to have a consequence die as a tool of play, and not on whether your players have an irrational bias against the d20. The former is a core issue for the usability of your system, while the latter is a pretty fleeting and not nearly universal aesthetic judgment. It's not an insignificant judgment, and if you think that it's powerful and lasting, then you don't really have a choice - but I think that you should at least try to consider the situation from a more general vantage point.

For what it's worth, my pick between your three solutions would be to reduce the number of experience levels. This is because of a bias against long campaigns on my part, it's much more interesting if you get radical changes in your character through a shorter time-frame. That is, of course, rather contrary to your own goals, so perhaps not very useful.

[NN]

how about, option #4: make min-maxing more difficult? 

[Callan S.]

Well, mechanical aesthetics are a real thing, they do really exist. So even if somebody disagrees or even thinks that the whole issue is stupid, that doesn't change the feelings of the guy doing the designing, nor that of his target audience

They aren't real - you can't change a rock by thinking about it because it's real. But this you can change by thinking about it. The synapses are not made in stone. And they could change if the hangup wasn't given protection and validation as a 'real' thing. Am I trying to force that change? No. I'm just, as far as my part of the conversation goes, not allowing it to exist as if it's 'real'.

Fie on validation through supposed rigidity.

[David C]

Using a d20 wouldn't be so bad, but sometimes I feel like the game isn't different enough from D&D.  I've already made several concessions to make the game more accessible. One of those was to have character classes (originally, the game was classless, you just chose skills that made you better at what you wanted to do). If players sit down and start rolling d20s+skill and pick out a class, I feel my game will have totally evolved my game into a heartbreaker.  Instead of the game's focus (telling a story in the classic "Hero's Journey')  People are going to be bombarded with D&D tropes and think it's a D&D wannabe. 

It might be completely illogical to feel this way, particularly about the d20, but it's out there. 

For what it's worth, my pick between your three solutions would be to reduce the number of experience levels. This is because of a bias against long campaigns on my part, it's much more interesting if you get radical changes in your character through a shorter time-frame. That is, of course, rather contrary to your own goals, so perhaps not very useful.

Honestly, I could see this as being really cool. I know exactly how I'd do it (especially since a lot of things are built around multiples of 5).  I can't really put it on the table though, it just requires me to revise too much.

Also, I personally prefer campaigns to last about 3 months at most. Generally, the way I've been treating levels is to have characters start at a higher level depending on what they're supposed to accomplish. A trio of dragon slayers would probably start as level 10 characters.

how about, option #4: make min-maxing more difficult? 

I've done a fair amount of this already. The problem is, if the players are level 15 and if you've got your skill at 9 (6 less than 15), your skill only has a 5% chance of succeeding.  It makes the guy with 10 skills at 9 useless, while the guy with 3 skills at 15 is perfect... at 3 things. 

[Eero Tuovinen]

Saying that people are not influenced by their aesthetic preferences on matters of game mechanics is absurd. You might not appreciate nor care, but these preferences are real. Some games cruise along merrily almost solely on the strength of their mechanical aesthetics - take BRP, for instance: it seems to me that people who like the system usually cite the intuitive percentage skills and percentage rolls as a major reason. Of course they've learned this aesthetic by playing BRP games, but that doesn't change the fact that they have these preferences.

David, though: your game is not going to become a fantasy heartbreaker by using a specific dicing mechanic. Also, I can't see your audience really caring about some historical analysis of rpg design, which is what the whole concept of "heartbreaker" is in the first place. So if your only reason for making bad design choices is that you're afraid of disaproving shadows, I suggest that you cast aside your fears and make the game you want, not the one you imagine others wanting.

[Callan S.]

Instead of the game's focus (telling a story in the classic "Hero's Journey')  People are going to be bombarded with D&D tropes and think it's a D&D wannabe.

This is getting onto a broader design question, but: Are you putting anything into mechanical design that will facilitate what you call a classic hero's journey?

Or are you trying to make a new game simply so the other players wont do the same old D&D thing they always do, because it's a new game? And thus you can do something different, something perhaps closer to your classic hero's journey? I'm asking to get a grasp of what your trying to do, in case I can offer something that maybe you'd think would help with that.


Eero, I'm pretty sure I've refered to the 'agonising for days' as demonstrating that this particular 'aesthetic preferences' is something that's fucking David over. Or atleast fucking him over in my opinion. If you have 'aesthetic preferences' that adds fun to your life, cool. But I'm not okay with 'aesthetic preferences' which seem to fuck people over - you seem to want to call them all real and thus preserve all of them, both the beutiful ones and the sickness ones.

They aren't real - they are desires we can humour and nuture in ourselves to benefit our lives or cull if they somehow prove unhealthy. They aren't real as in they exist whether we like it or not. If that were the case and you have an unhealthy one, you'd be damned to that sickness for the rest of your life. There is an alternative to 'they're real!'.

[David C]

You know what, I'm going to change the die to a d20.  It's the least drastic thing I can do. 

This is getting onto a broader design question, but: Are you putting anything into mechanical design that will facilitate what you call a classic hero's journey?

Players write a short background on their character as to why they are adventuring. It's essentially a Kicker.  They also write out their "Destiny."  It's essentially how they view their "Bang!" happening.  During play, players can declare or make milestones in their story. As a player I can say this town we're visiting isn't just any town, it happens to be where "The Shadow" has been operating.  Also, children have been kidnapped from there. 

[Callan S.]

Hi David,

Thanks for your answer. In playtest, how much did the players pick up on writing their short background, destiny, milestones and detail injection into towns?

If they are and it is affecting play, you don't have to worry about being like D&D so much. Is it affecting play enough?

Also when you say 'their bang' it sounds like your saying they have one, single bang. That they write. This is a multi session sort of game, isn't it? Or is it one session?

[JoyWriter]

Honestly, I could see this as being really cool. I know exactly how I'd do it (especially since a lot of things are built around multiples of 5).  I can't really put it on the table though, it just requires me to revise too much.

How about you make a brand new game that does what you really want to do?

That might sound factitious, but you can create a game with a set of playtesters that they really love, but doesn't do what you want. It forms a local maxima of familiarity and newness that achieves a lot of what they want and looks pretty good to you.

No prob, just do a new one, and push the old game more and more in the directions they are interested in. That way you don't have to worry about that game and the new one being confused, they just become heroquest and runequest to each other. Of course, if you start designing specifically for your playtesters, I'd probably ask them to start paying for ashcans/preorders or something, just to insure that this market you are now chasing is a real one.

[Locke]

it sounds like your rolling scheme is flawed if you have to knock levels out from 5 to 15.  That seems somewhat drastic to me.  Why do you want to make a separate rollign system?

I designed something called "Elegant10"  which is used in my game Age Past.  I designed the rolling to do something that a d20, or a success systems can't do. 

D20's and successes are based off straight probability.  This is why dnd 4th moved to a +5 "trained" system as the 1's and 2's given as bonus didn't amount to a hill of beans for gaming purposes.  YOU are great at hearing that's why you have a +10% chance at it; not really a big bonus at all.

I wanted people to be able to choose how to roll with every roll and with every roll there is a consequence.  Also I wanted the roll its self to matter, so with each roll there is a chance to fail, fumble, have a guaranteed result, or if the skill is high enough critically succeed.

Is skill granted based on level?  Is the skill guaranteed?  Or is the skill purchased?  Because you could raise the price of the skill so that a level 5 might have a value of 3 and a level 15 might have a value of 5.  High level character should always succeed at moderate difficulty tasks.  So if its a balanced system charaters w/o the skill should fail often at moderate tasks and those with the skill should succeed.  It might just be some simple re balancing.  It took me a while to get it right, but make 3 characters at a low level, middle, level and high level.

Say what the character is good at and say absolutely when it should succeed.  Then make the system to succeed.  When I make a system I make a character the way I would be happy playing ti at 1st level, then give it that points to design it to build that way at first level.  Do the same thing with the skills.