How odds make you feel

[Thaumaturge]

IME, the chance of success in 4e is more typically in the 60-70% range.

There are studies (which, IIRC influenced the design of 4e) that indicate that the "ideal" rate of success for games of chance is around 66%. Raise the odds significantly above that and people get bored; drop it significantly below and folks get frustrated.

I think that the designers of 4e took that to mean that on every roll, or almost every roll, there should be a ~66% chance of success. Hopefully, with their attention turning to the adventure, they will see we want to succeed at ~66% of our tasks, but some tasks can have a 33% and some can have a 99%, if it averages out to about 66%.

I, personally, want various challenges to challenge me at various difficulty levels.


Thaumaturge.

 

[Frostmarrow]

I begin a lot of posts on these boards with "it depends". This one is no different.

So, the feeling I get from varying odds depends on several factors. The probability of a successful roll is nearly meaningless without defining two crucial things: the consequence of success and of failure.

If I make twenty attack rolls during a fight, with a success taking some HPs from the enemy and a failure meaning I wasted some time, there is not much difference between 30% and 70%; I care more about average DPR, if anything.

It changes if the roll decides something important by itself. The more I can gain, the more even a small chance of success means. The more I can lose, the more significant is any chance of failure. If I have a chance of gaining a powerful ally, 30% chance is a lot; if the roll decides between living and dying, 90% still feels risky.

A secondary, but important, factor, is how the successes and failures are presented in game. If I'm playing an experienced fencer, I won't accept a situation where "I miss" 30% of time. But I have nothing against having 30% or less successful attacks against a similarly experienced opponent - if it is clear that the failed rolls represent opponent's tricks and parries, not my failures.

The terminology used in the rules and the GM narration style are very important here, and it is impossible to comment on % chances without this kind of context.

That DPR-comment is interesting. You could do 9 damage over three rounds in several ways. If you auto-hit your output per attack should be 1d5. If you hit 2 out of 3 your damage should be 1d8. If you hit 1 out of 3 your damage dice should be 3d5.

But I think we've tried this already. Looks a lot like the odds for wizards, fighters and sneak attacking rogues respectively.

 

[Frostmarrow]

I think that the designers of 4e took that to mean that on every roll, or almost every roll, there should be a ~66% chance of success. Hopefully, with their attention turning to the adventure, they will see we want to succeed at ~66% of our tasks, but some tasks can have a 33% and some can have a 99%, if it averages out to about 66%.

I, personally, want various challenges to challenge me at various difficulty levels.


Thaumaturge.

100/1.62=61.7

The golden ratio seems to be 62%. On 1d6 3+ comes pretty close. I.e 66%.

On a 1d20 we get 9+ (or 60%). This cannot be a coincidence.

***

Be that as it may. What I'm trying to explain is maybe the classes should have a baseline success rate that is different from other classes. Compare if you will how the different hit dice for different classes. 1d4 for wizards, 1d6 for rogues, 1d8 for clerics and 1d10 for fighters. They all have hitpoints but the allotment is extremely unfair.

There are signs that the game is pulling in this direction anyway. Rogues get flank bonus, wizards get True Strike and so on. Wizards expect to be successful. Rogues rack up the circumstance bonuses. Fighters on the other hand rely on innate bonuses. And clerics are left behind with their middling BAB and completely random healing.

 

[Thaumaturge]

I think it would be interesting to pick a class based on how risky you want to be. But maybe it could be expanded out so that each archetype had a 'risky' setting, a 'risk averse' setting, and a 'middling' setting. We've, in essence, had this before with wild mages being risky, while my brother built a 3e wizard who only cast spells that had no attack roll and few had saves. He was 'risk averse'.

We can imagine the cautious fighter, and the fighter who tears into combat without a thought. Similarly, in fiction we have risk taking thieves and risk averse thieves.

I really like the idea of being able to set your own risk tolerance, though, by picking class features or whatnot.

Thaumaturge.

 

[Fanaelialae]

I've been nagging people about that 66% chance ideal for years. In fact I've written an RPG called 66 Demoner (in Swedish mind you) with that success rate as a base.

But now I looking at other odds. I think you'll agee with me when I say that 5% chance will frustrate players and that 95% will make them nonchalant. My point is what if I want players to be frustrated? What if I want players to be nonchalant? So if my players are fighting hordes of zombies I might want to frustrate them or if they are fighting mooks I might want them to behave like rockstars.

I learned this playing Warhammer. A lot of the difference in feel between playing Space Marines and Imperial Guard is that marines succeed 66% of the time whilst guardsmen succeed 50% of the time. You feel confindent with marines and you feel exposed while playing Imperial Guard.

This could be used in D&D. We might want wizards to be perfectionists in a few areas and we might want clerics to be thoughtful (yet useful in more fields). Or even we might want evil clerics to be full of themselves and good clerics to be self-doubting.

Traditionally high level characters are sure of success (80%) and low level characters rely on sheer luck (50%). It might be interesting to mix this up.

"The whole problem with the world is that fools and fanatics are always so certain of themselves, and wiser people so full of doubts."
(Bertrand Russell)

***

I really like the idea of built in auto-success with high ability scores.

I see what you're getting at but I think the end result would more likely be that some classes are favored more heavily than others (because those appeal to the majority of players psychology of success). I could see perhaps variants built around those guidelines, but I don't think most players would be satisfied with a character whose primary shtick is a coin toss (like your proposed cleric). IMO, the core classes should be based on mechanics that most people will enjoy trying.

 

[Frostmarrow]

I think it would be interesting to pick a class based on how risky you want to be. But maybe it could be expanded out so that each archetype had a 'risky' setting, a 'risk averse' setting, and a 'middling' setting. We've, in essence, had this before with wild mages being risky, while my brother built a 3e wizard who only cast spells that had no attack roll and few had saves. He was 'risk averse'.

We can imagine the cautious fighter, and the fighter who tears into combat without a thought. Similarly, in fiction we have risk taking thieves and risk averse thieves.

I really like the idea of being able to set your own risk tolerance, though, by picking class features or whatnot.

Thaumaturge.

So if fighters are confident at 70% chance and this is what players can expect from fighters, but of course there are variants. Barbarians could be fighters with a 60% base chance and rangers could have 80%. That way you could play a wreckless fighter or a fighter with increased precision.

Now, I don't think its absolutely necessary to balance damage (or what have you) to make up for the differing success rates. Instead weapon selection could be more restricted. Greatswords might not be available to fighters (being the domain of barbarians). Or other class features, such as rage, could make up for the loss of accuracy. Fighters and barbarians need not be balanced but instead offer different play experiences.
Playing barbarian means failing more often than playing fighter but it will still be fun in it's own way. (As in failing more often gives you more opportunities to growl, spit, and curse).

 

[Frostmarrow]

I see what you're getting at but I think the end result would more likely be that some classes are favored more heavily than others (because those appeal to the majority of players psychology of success). I could see perhaps variants built around those guidelines, but I don't think most players would be satisfied with a character whose primary shtick is a coin toss (like your proposed cleric). IMO, the core classes should be based on mechanics that most people will enjoy trying.

Some classes are more popular. There is no need to strive for an equal opportunity role-playing game.

I admit that playing a coin-toss class would make me sick to my stomach. I want to play something I can rely on to get the job done. But that's me (and hordes of gamers everywhere). Still, Imperial Guard remains a popular army in 40k. Some people do enjoy playing chaotic (as in unreliable) avatars in games. I, for one, is perfectly content playing the oh so fragile eldar just so that I can have the speed and accuracy to make a plan and see it through. But I have friends who wouldn't play any army that has no need for the scatter die.

Last night we made characters for Pathfinder in a new game. My friends tried to persuade me into play something out of the ordinary. I was adamant, however, and went for a human cavalier. The rest of the group is a side show.

 

[Fanaelialae]

I think that the designers of 4e took that to mean that on every roll, or almost every roll, there should be a ~66% chance of success. Hopefully, with their attention turning to the adventure, they will see we want to succeed at ~66% of our tasks, but some tasks can have a 33% and some can have a 99%, if it averages out to about 66%.

I, personally, want various challenges to challenge me at various difficulty levels.


Thaumaturge.

I don't think so. A character with an ability score of 10 and no skill training will not have a ~66% chance of success in 4e. Bonuses in 4e range from -1 (8 ability score without skill training) to in excess of +12 (20 ability score with skill training and racial bonus). There's no way to group such a wide range (-1 to +12) into such an narrow range (60-70%). What they did do is make the typical skilled bonus of +6 to +8 fall within that range. There's still significant variation outside that range for characters with greater or lesser talent, as well as for checks that are easy or hard.

 

[Fanaelialae]

Some classes are more popular. There is no need to strive for an equal opportunity role-playing game.

I admit that playing a coin-toss class would make me sick to my stomach. I want to play something I can rely on to get the job done. But that's me (and hordes of gamers everywhere). Still, Imperial Guard remains a popular army in 40k. Some people do enjoy playing chaotic (as in unreliable) avatars in games. I, for one, is perfectly content playing the oh so fragile eldar just so that I can have the speed and accuracy to make a plan and see it through. But I have friends who wouldn't play any army that has no need for the scatter die.

Last night we made characters for Pathfinder in a new game. My friends tried to persuade me into play something out of the ordinary. I was adamant, however, and went for a human cavalier. The rest of the group is a side show.

I have no issues with nonstandard character types. However, they should not be the default.

You, yourself, admit that most people aren't likely to want to play a coin toss class. If the cleric is a coin toss class, then players are potentially burdened with undesired mechanics either for the sake of story (they want to play a priest archetype) or for the sake of other mechanics (the party wants a healer). That's just going back to the "no one wants to play the cleric but every group needs one" issue that was fairly prevalent for the first three editions.

Now, if you want to design a non-standard class around non-standard mechanics, such as a Gambler class, I'm all for it. Even a non-standard variant of a standard class, such as a Priest of the god of gambling, is fine. Or having optional abilities that leverage a greater amount of luck (such as a more powerful spell you can memorize that only works 50% of the time) might be okay, though there's a risk that it will lead to a rocket tag style of play that I dislike.

I just don't think it's a great idea for the default kernel. Some variation is to be admittedly desired. There's an argument to be made to the effect that 4e made classes too alike. However, that doesn't mean that one needs to leap over the other side of the boat and design a primary class whose core abilities only have a 30% chance of success.

 

[Frostmarrow]

I don't think so. A character with an ability score of 10 and no skill training will not have a ~66% chance of success in 4e. Bonuses in 4e range from -1 (8 ability score without skill training) to in excess of +12 (20 ability score with skill training and racial bonus). There's no way to group such a wide range (-1 to +12) into such an narrow range (60-70%). What they did do is make the typical skilled bonus of +6 to +8 fall within that range. There's still significant variation outside that range for characters with greater or lesser talent, as well as for checks that are easy or hard.

Indeed. The game is rather swingy. Some players put all their eggs in the same basket and become a one trick pony. This is not something they do out of design. They just like to be the best they can be in a specialized field and can live with being next to useless in any other field. However, given the choice i.e. more points, they branch out.

I see all these one trick ponies to be a symptom rather than elegant design on part of the game. It's like players try to break out of the class mold. Perhaps the class system is really quite uncomfortable given the effort spent by players to not conform.