PDA

View Full Version : Degrees of Success and Failure



Mr. Mask
2015-02-10, 10:50 AM
Something I've noticed that's rather lacking in a number of RPGs, is degrees of success and failure.

In many RPGs, you can hit, or miss. Succeed, or Fail. For perfect or abysmal rolls, you get critical successes and failures as well, their effects varying from vague and powerful to specific and decent, depending on the game--but either way they tend to only add two more degrees, for extreme cases.

I think this emphasizes certain elements of RPGs. Minmaxing for one. If you need to roll a 17 to succeed, what do you get for rolling a 16? Nothing. So a 1 point bonus can make an extreme difference, the divide between failure and success. Of course, there are ways to minimize this impact, namely have the players roll many times to decide an event. This both helps and hinders the person who isn't minmaxing the given skill. It helps them, in that they're more likely to succeed at least once within the allotted rolls. It hinders them, in that when competing with a minmaxer, they're going to come out behind in the long run (more rolls means more statistical certainty), so it is negative within a competitive environment.

That's not to say this style of play is bad. Merely, it is specific, and not everyone would want this emphasis for their RPG or game. So, I thought it'd be interesting to discuss mechanics for degrees of success. In many systems, they can be a little cumbersome.

Any notable systems or mechanics you've seen, that made use of degrees of success/failure? Any thoughts on systems that make use of degrees of success/failure compared to ones that don't? Any ideas of your own on how you like to handle degrees of success?

Segev
2015-02-10, 11:00 AM
The oft-decried White Wolf system actually does something of this, though it does it almost backwards to what you're discussing: When you attack somebody, you roll your die pool to see if you hit. If you get enough successes to overcome their defense (in their more recent systems, this is a static number called "Defense Value (DV),"), you add each success over thier DV to your damage die pool (on top of whatever your weapon does).

So there's "degrees of success" built in, there.

Part of the difficulty of "degrees of success/failure" is that there is a point where you just fail, and a point where your success is complete. If you try to define a gradient, you wind up, still, with the "minimum needed to succeed" being that "one point difference" between success and failure.

Take the Indianna Jones example from another thread where this was brought up: He leaps across a chasm. If he rolls too low, he never reaches the other side and falls to his doom. If he rolls sufficeintly high, he lands with both feet planted and his arms up in an Olympian's "V" over his head. In between, a "degrees of success/failure" roll would have him hit the ledge and cling to it and try to pull himself the rest of the way up. But how granular do we want this middle ground? Do we have "landed on his face, but just has to stand up" to compare to "landed with both feet firmly planted?" Do we have "half on, half off, needs a climb roll" and "landed barely clinging to the ledge with his fingers, needs a more difficult climb roll?"

That's certainly interesting, but the more grades you give it, the more you have to either spell out a ton of special cases and examples for each skill, and/or the more you leave to the DM to try to ad hoc (which, in a system rules-heavy enough that degrees of success/failure are being explicitly defined, is probably not the right way to go).

So...it's tricky.

The best way to tackle it I've seen is to have an optional rule blurb of a rules-light sort of, "If they fail by less than X, give them a second check they can make to turn it into a success."

Arbane
2015-02-10, 11:03 AM
The most recent Warhammer RPG uses custom-made dice to add 'boons' and 'banes' to pass/fail rolls - you get some added complication (good or bad) in addition to straightforward success and failure.

Most Apocalypse World derived games (Monsterhearts, Dungeon World, etc), you succeed on a roll of 7+ on 2d6, but you only get to pick one good effect from a list (not having things get more complicated is usually one of them) and on a 12, you get to pick two of them.

The comedy game Teenagers from Outer Space has the idea of succeeding TOO well. Like asking a girl out and having her become a crazed stalker, or trying to grow a mustache and having it become an international celebrity. (That apparently happened in an actual game.)

Synovia
2015-02-10, 11:10 AM
This seems like asking for rules for the sake of rules.

There's always going to be a point where if you rolled one higher you go from failure to success. I think degrees of success is something that should be in the realm of the dm/players. Indy needs a 15 to make that leap? He rolls a 16 - he barely gets his fingers on the edge, but manages to find a foothold and pushes himself up. He rolls a 14? He gets his fingers on the edge and slowly slips off.

He rolls a 3? He's 5 feet short and plummets into the chasm.

More tables to roll on here doesn't seem like it makes for a better game.

Morty
2015-02-10, 11:11 AM
There are, indeed, systems which use degrees of success and failure. Various iterations of the Storyteller systems make the number of successes you roll matter - the degree to which they do depends on the system in question and the situation. There's not much in terms of degrees of failure, though. oWoD's botch rules are infamous for screwing players over quite often. nWoD 2e actually gives players the option to turn a regular failure into a dramatic one for experience rewards. They happen rarely, otherwise. Other systems that make use of dice pools likewise tend to keep track of the exact numbers of successes you rolled.

I'd use the example of Dungeon World, but someone just did. The mechanics of partial success is quite important there, and the probability of the 2d6 roll skews towards the partial success result.

I do agree that at least some mechanics for degrees of success or failure are handy, or even necessary. It's both more realistic and more dramatic.

neonchameleon
2015-02-10, 11:14 AM
The most recent Warhammer RPG uses custom-made dice to add 'boons' and 'banes' to pass/fail rolls - you get some added complication (good or bad) in addition to straightforward success and failure.

Most Apocalypse World derived games (Monsterhearts, Dungeon World, etc), you succeed on a roll of 7+ on 2d6, but you only get to pick one good effect from a list (not having things get more complicated is usually one of them) and on a 12, you get to pick two of them.

The comedy game Teenagers from Outer Space has the idea of succeeding TOO well. Like asking a girl out and having her become a crazed stalker, or trying to grow a mustache and having it become an international celebrity. (That apparently happened in an actual game.)

Fantasy Flight (both WFRP 3E and Star Wars: Edge of Empire) and the Apocalypse World Engine were two of the three systems I was going to mention (AW is two good results on a 10+ normally). The third being Cortex Plus (Marvel Heroic Roleplaying/Smallville/Leverage/Firefly). Where there's pass/fail based on the highest two dice on each side for your dice pool, and each 1 you get adds a complication to the situation. (Lucky results are when your opponent gets complications). Which means that an extra d4 in your dice pool is a liability because it won't be in the two best normally, and has a 1/4 chance of creating a Complication.

Also going waaaay back there's the old Rolemaster tables which also give degrees of success.

Beta Centauri
2015-02-10, 03:28 PM
Yes, this is a major problem, particularly with D&D. Even they sort of get it; notice that in 4th Edition, every daily power either does something on a miss (half-damage, an effect, etc) or isn't expended (which is still a flat, dull miss, but at least lets you try again).

And it does lead to the issue you point out: there is absolutely no upside or mitigating factor to most misses, so there's no game-based reason not to optimize. The reasons not to optimize have to come from the player deciding to do things for character reasons or to adjust the difficulty for themselves, or from the GM choosing to find a way to punish optimization.

Some games get it. In Dungeon world, a middling roll gives the desired result, usually with the player's choice of the resulting downsides. A great roll gives the desired result with fewer or no downsides. A complete failure moves the game forward on the GM's side, and lets the player mark experience. Something happens that advances the game and sometimes advances the character.

I like Fate's approach: Don't roll the dice unless two things are true:

1. Success would be interesting.
2. Failure would be interesting.

Yes, this means that rolling to hit shouldn't happen unless missing (or whatever the failure would be) would be interesting. Maybe someone likes to describe misses and that's interesting enough. If not, though, there's no point in rolling. If a player doesn't want to roll to hit, what then? Then it's a matter of finding rolls the player would find it interesting to fail and focusing on those for that player.

Frozen_Feet
2015-02-10, 04:34 PM
Yes, this means that rolling to hit shouldn't happen unless missing (or whatever the failure would be) would be interesting.

Uh huh.

In any turn-based tactical game, failure to remove enemy piece on your turn means the enemy might remove your piece, or do something else detrimental. If the chance of losing combat is not interesting enough for you, what is? Systems like Dungeon World don't actually change this truism, they simply consolidate player and enemy actions into a single roll to speed up play.

For the system I'm making, it's going to have fairly large number of varying degrees, and I'll likely add in a mechanic like Dungeon World's where the player gets to pick extra good stuff or extra bad stuff depending on it. The RNG is planned so that the extreme ends happen rarely, and the better your character gets, the less you will see of the bad end.

At the end of day, you shouldn't fret too much about it. Even if a check "just fails", it will actually have effects on the game, if you give yourself and players some time to think about it. For example, what if players fail to pick that door? They might try again, but it will take time. Are there people about who might stumble upon them? If they won't or can't try again, they'll have to pick another route. Quite often, you don't need specific mechanics for failure - all the consequences will logically follow from all the other rules if your player characters are not operating in a void. This loops back to the combat example - why on earth should a miss need any more effects than letting the other guy live for few turns more to cause harm?

Beta Centauri
2015-02-10, 04:53 PM
Uh huh.

In any turn-based tactical game, failure to remove enemy piece on your turn means the enemy might remove your piece, or do something else detrimental. If the chance of losing combat is not interesting enough for you, what is? Losing a piece of equipment, giving the opposing force (if not this immediate enemy) some larger advantage, suffering a lingering effect. Lots of stuff. That's just off the top of my head.

But in any case, there's generally no immediate effect from missing. As you say, it means the enemy "might" remove your piece, but the enemy might also miss. If everyone misses in a round, it's was a complete waste of time.


Systems like Dungeon World don't actually change this truism, they simply consolidate player and enemy actions into a single roll to speed up play. Dungeon World does. Unless otherwise stated, on a miss the GM makes something dangerous happen to the characters. Hack & Slash doesn't specify, so it it could be just about anything that follows from the fiction. A monster move probably follows, as does "Deal damage," but "Show a downside to their class, race or equipment" could also follow.


why on earth should a miss need any more effects than letting the other guy live for few turns more to cause harm? Because "letting the other guy live" isn't a thing that happens, its the default effect of nothing happening.

Geostationary
2015-02-10, 04:54 PM
If you want something really weird, mortal actions in Nobilis and Chuubo's Marvelous Wish-Granting Engine don't actually correlate to success/failure- rather, they chart to greater productivity and impressiveness! For instance, let's say you form an intention to kill the man who burned your village down. You may or may not succeed in doing this, but even if you fail so long as your intention is rated high enough it's guaranteed to be varying degrees of productive and/or impressive for you to have done so; conversely low intention levels mean that it's likely to be counterproductive for you. Further, the GM can always give you more than what your actions guarantee, but they have to give you the minimum.

Frozen_Feet
2015-02-10, 05:15 PM
Because "letting the other guy live" isn't a thing that happens, its the default effect of nothing happening.

That's not a valid reason to add anything mechanical to a miss, because the other guy still standing can directly lead to:


Losing a piece of equipment, giving the opposing force (if not this immediate enemy) some larger advantage, suffering a lingering effect. Lots of stuff. That's just off the top of my head.

Griping about how it won't immediately or necessarily do so is missing the point. Same goes for saying it's "wasting time". Last time I checked, time is a valuable resource in most any game. Stuff is happening, the game is progressing, you just have to stop and think to see how. You can consolidate and codify rolls to speed up play, but doing so rarely adds to possible events.

Beta Centauri
2015-02-10, 05:21 PM
That's not a valid reason to add anything mechanical to a miss, because the other guy still standing can directly lead to: Can. Might not.




Griping about how it won't immediately or necessarily do so is missing the point. Same goes for saying it's "wasting time". Last time I checked, time is a valuable resource in most any game. That's why it's entirely the point. The time we spend at the table should feel like it's spent moving things forward.

We've all seen it: someone whiffs a roll and says, with exasperation, "Well that was a waste of time." Sure, I could tell that player some other way they should look at it, but I'd rather just have the roll do something right now that the player finds interesting and feels is a direct contribution to to moving things forward.

If all that happens on a miss is that it's someone else's chance to move things forward, the player really hasn't done anything. And the next person might not move things forward either.


Stuff is happening, the game is progressing, you just have to stop and think to see how. You can consolidate and codify rolls to speed up play, but doing so rarely adds to possible events. It's not about speeding up play, it's about making the play worth playing. I don't care how fast a character can miss three times in a row or how fast everyone in the situation can fail their rolls if none of that actually does anything.

TheThan
2015-02-10, 05:41 PM
So I was just talking about this in another thread. So Iíll just repost myself instead of retyping it all.


This is another reason why fate is such a good system. It takes into account variable degrees of success and failure.

Letís borrow from Indiana Jones for a second. In the opening of Raiders of the lost ark, Indiana has to leap across a chasm. He runs, makes his jump and promptly slams his chest into the edge of the precipice on the other side.
Did he succeed his leap? Did he fail his leap?

Beating an arbitrary DC is binary, you either succeed or you donít. However things arenít always so well defined. In the above example Indiana could have very well failed his attempt to leap across the chasm, but he may not have failed it by such a large margin that he fell to his doom; Or he may have succeeded his attempt, but only just barely, escaping death by the skin of his teeth. How do we know? Fateís shift system accounts for this and the outcomes are determined by a system that takes that into account. I call it degrees of degrees of success or failure. This allows Dms and players to describe in detail the action of whatís going on, this leads to a lot more exciting encounters.

(as you can see, i really like the fate system).


Uh huh.

In any turn-based tactical game, failure to remove enemy piece on your turn means the enemy might remove your piece, or do something else detrimental. If the chance of losing combat is not interesting enough for you, what is? Systems like Dungeon World don't actually change this truism, they simply consolidate player and enemy actions into a single roll to speed up play.

Fate doesnít look at characters and enemies as pieces on a game board. Thatís a very miniatures game approach with itís own set of assumptions. Instead it takes a more cinematic approach to RPGs and therefore makes different assumptions about the game youíre playing.

Unless the Dm is nice to you and has enemies captures you, or you run away, failure always results in death. But in Fate, failure results in other consequences. For example, the above scenario could have cost Indiana Jones too much time climbing up the cliff and he gets stuck in the room with the doors sealed, forcing him to find another way out and leading to more interesting and entertaining encounters. Failure does not necessary mean death in Fate It means the heroes are hindered and must overcome increasingly bad odds, but itís still fully possible to win and have a grand epic adventure.

It also removes the ďplayers are screwedĒ situation where they must succeed on a survival check, or a search check or some such and just canít seem to roll high enough. What happens? The game comes screeching to halt and the Dm has to throw the players a bone to get things moving again. the Dm has to actually bend the rules to keep the game flowing, whereas with fate, he doesnít have to, itís got built in failsafe for such situations.

Beta Centauri
2015-02-10, 05:47 PM
Unless the Dm is nice to you and has enemies captures you, or you run away, failure always results in death. But in Fate, failure results in other consequences. For example, the above scenario could have cost Indiana Jones too much time climbing up the cliff and he gets stuck in the room with the doors sealed, forcing him to find another way out and leading to more interesting and entertaining encounters. Failure does not necessary mean death in Fate It means the heroes are hindered and must overcome increasingly bad odds, but itís still fully possible to win and have a grand epic adventure.

It also removes the ďplayers are screwedĒ situation where they must succeed on a survival check, or a search check or some such and just canít seem to roll high enough. What happens? The game comes screeching to halt and the Dm has to throw the players a bone to get things moving again. the Dm has to actually bend the rules to keep the game flowing, whereas with fate, he doesnít have to, itís got built in failsafe for such situations. That's not so much Fate as an approach Fate advocates. I mostly play 4th Edition D&D, and I rarely have failure mean death. And the chapter on skill challenges in the 4th Edition DMG directly addresses the "players are screwed" situation and exhorts GMs to avoid making skill challenges that have a dead end in the event of failure.

I'll happily grant that I got much of how I look at RPGs now from Fate (Spirit of the Century, really), and I like Fate, but a lot of its ideas are very easily transferable. They mainly just tried to codify things that good GMs and good players do without rules.

Frozen_Feet
2015-02-10, 05:53 PM
I mentioned this in the other thread too, but most old school games already had a failsafe: it's called "PCs do something else". The problem isn't with all-or-nothing checks, it's with linear structure of a game. It's the idea that the PCs must do something, and mechanics which turn "Grand Epic Victory" to "Grand Epic Victory with slightly more struggle" don't actually fix it, nor do they increase amount of possible actions. A game where it's possible to get lost in the woods and end with honest-to-God game over is preferable to that.

Yes, this is much closer to miniature games than it is to cinema - and I find it perfectly justified and far superior, because TRPGs are descended from and much closer to miniature games, than they are to movies and film-making.

neonchameleon
2015-02-10, 06:03 PM
I mentioned this in the other thread too, but most old school games already had a failsafe: it's called "PCs do something else". The problem isn't with all-or-nothing checks, it's with linear structure of a game. It's the idea that the PCs must do something, and mechanics which turn "Grand Epic Victory" to "Grand Epic Victory with slightly more struggle" don't actually fix it, nor do they increase amount of possible actions. A game where it's possible to get lost in the woods and end with honest-to-God game over is preferable to that.

Yes, this is much closer to miniature games than it is to cinema - and I find it perfectly justified and far superior, because TRPGs are descended from and much closer to miniature games, than they are to movies and film-making.

Most old school games were also run as relatively static open sandboxes where the PCs are the ones making the play by trying to tame the wilderness. In such a situation "PCs do something else" works well (especially on a hex map). It isn't so good when the PCs are on a ticking clock because the enemy is active NPCs more than it is the untamed environment.

Beta Centauri
2015-02-10, 06:15 PM
I mentioned this in the other thread too, but most old school games already had a failsafe: it's called "PCs do something else". Sure, that's an option, but that's often related to having not made any progress for a while on the thing they had been doing. How long had they been doing that before they decided they weren't going to get anywhere? How long did their desire to figure the thing out struggle with their desire to actually get somewhere in the game? How frustrated are they when they finally just do the other thing?


The problem isn't with all-or-nothing checks, it's with linear structure of a game. It's the idea that the PCs must do something, and mechanics which turn "Grand Epic Victory" to "Grand Epic Victory with slightly more struggle" don't actually fix it, nor do they increase amount of possible actions. Straw man. None of the mechanics discussed "turn '+Grand Epic Victory' to 'Grand Epic Victory with slightly more struggle.'"Players can and should fail and fail irreparably. They didn't arrive in time to save the village, and now that village is gone, forever. They failed, and failed badly. And the game goes on.


A game where it's possible to get lost in the woods and end with honest-to-God game over is preferable to that. I'd be fine with that, as long as the game makes it clear that it is a game over, and not something the players are meant to keep rolling do-nothing failures against.

My interest is in the players, and making good use of their time. If we decided to play for two hours then we will have fun during those two hours, whatever fun means to us. I tend not to use situations in which unlucky rolls mean that that one or more characters dies or is otherwise neutralized, but if I did, I would be prepared with other characters the players could immediately play. I believe that was advised in earlier editions, by having henchmen that could either be the ones being killed, or could be used as replacement characters. There are issues with that execution, but the idea is sound.

Point being, it sounds like we agree that failure is fine, it's just the failure that wastes players' time in a non-progressing situation that is problematic. Yes, the players can do something else, but the GM can simply avoid non-progressing situations. It doesn't even really require new mechanics.


Yes, this is much closer to miniature games than it is to cinema - and I find it perfectly justified and far superior, because TRPGs are descended from and much closer to miniature games, than they are to movies and film-making. Things have changed. Lots of people don't come to TRPGs after having played miniature games, and they see TRPGs as distinct from board games that have a clear loss condition that tells them the game is over, who won and who lost, and that it's time to set the game back up. The games themselves even draw from movies, (or books that have been turned into movies), giving people the (perfectly valid) expectation that these games will let them pretend to be characters like the ones in movies. It's been that way for a while.

Just because a thing relates back to the way things were doesn't make it "justified" and "superior" now, unless one's intention is to ignore the present and try to undo change.

Deophaun
2015-02-10, 06:52 PM
This seems like asking for rules for the sake of rules.

There's always going to be a point where if you rolled one higher you go from failure to success. I think degrees of success is something that should be in the realm of the dm/players. Indy needs a 15 to make that leap? He rolls a 16 - he barely gets his fingers on the edge, but manages to find a foothold and pushes himself up. He rolls a 14? He gets his fingers on the edge and slowly slips off.

He rolls a 3? He's 5 feet short and plummets into the chasm.

More tables to roll on here doesn't seem like it makes for a better game.
The thing about this example is that it's already in 3.5:


If you attempt a Jump check untrained, you land prone unless you beat the DC by 5 or more.

If you fail the check by less than 5, you donít clear the distance, but you can make a DC 15 Reflex save to grab the far edge of the gap. You end your movement grasping the far edge. If that leaves you dangling over a chasm or gap, getting up requires a move action and a DC 15 Climb check.

So that shows how important degrees of success or failure is: people ignore it even if it's in the official rules.

TheThan
2015-02-10, 09:39 PM
The thing about this example is that it's already in 3.5:




So that shows how important degrees of success or failure is: people ignore it even if it's in the official rules.

No, its not that people ignore it (I knew those rules going into this thread). Itís that the game ignores it, itís an afterthought. Itís in the rules but itís not the fundamental rule of the game.

In D&D, the universal rule, the core of the system, is to roll a D20 and add a modifier to it, hoping to beat a static preset DC for any and all situations. It doesnít innately take into account the possibility of near failure or near success in its core mechanic. They had to create more rules detailing specific scenarios in order to give their game flexibility, this makes the game much more clunky and slow as people have to stop and refresh themselves on what exactly happens when you fail a grapple check.

A game like Fate goes in with that already being considered, itís part and partial of its core mechanic, roll 4 fudge dice, add a modifier and try to beat a DC, the greater the success the better the effect, the greater the failure, the worse the consequences. There is no need for the system to have extra rules stuffed into it to make the game flexible, it already is.

Mr. Mask
2015-02-10, 09:50 PM
Plus: What are the degrees of success rules for attacks in 3.5? Swimming? Crafting? Saves? Diplomacy? Sneaking? Riding? I recalled certain rolls have an additional degree of success, but have difficulty remembering them for some others.

Segev
2015-02-11, 10:26 AM
It's hard to have degrees of success for something like attacks. Beta Centauri seems to be arguing that "if you don't hurt them, then they hurt you" is enough to make it "degrees of success," but that's just pushing the goal posts around. Under his apparent preferences, it goes from "well, that attack was a waste of time" to "well, I'm actively worse off for having tried anything at all."

Note that this still can happen even with "whiff" style misses.

But more to the point, there is always a point at which what you're trying to do doesn't lead to the results you were seeking. If not, there's little point in rolling; just declare your action successful. (Everybody does this in games from time to time; the more pedantic and trivial you want to get, the more apparent it becomes: "I get out of bed, get dressed, and go down to get some breakfast" could, in theory, involve 3 or more rolls to see if you mess anything up or even manage to wake up at all when you say you do. But it's usually not interesting enough to bother with, and the chances of failure should be so miniscule that any mechanical rolling to determine success that was in any way meaningful would perforce increase everybody's clumsiness to comedic levels.)

So, any time you care enough to roll dice, that implies there's a chance of not achieving the goal the player had in mind. Anything that comes up with those results counts as a "failure" to the player's mind, even if the GM and the conceits of the game try to claim otherwise. Similarly, if the mechanics of the system require multiple rolls for resolution (as most systems' combat does: attack then damage), the second roll is, itself, a degree-of-success measure.

If I attack and hit, I then roll damage, and that damage roll says how "successful" my hit was. (Systems with "critical hits" further complicate this, making degree of success minima and maxima increase as a general rule when crits are scored).

If one wished to complicate things, one could assign some scoring system that is akin to "hit points" to any obstacle being overcome. Upon rolling successfully to do anything meaningful, the player next rolls "degree of success," and sees if he's gotten sufficient degrees to overcome the obstacle or not. If not, he's still made progress; the obstacle has fewer "hit points" left.

goto124
2015-02-11, 10:58 AM
I"I get out of bed, get dressed, and go down to get some breakfast" could, in theory, involve 3 or more rolls to see if you mess anything up or even manage to wake up at all when you say you do. But it's usually not interesting enough to bother with, and the chances of failure should be so miniscule that any mechanical rolling to determine success that was in any way meaningful would perforce increase everybody's clumsiness to comedic levels.

Campaign idea: Party gets drunk and passes out. The next day, they wake up with a massive hangover and -7 to pretty much every check.

Use only in comedy campaigns.

Knaight
2015-02-11, 11:32 AM
It's hard to have degrees of success for something like attacks. Beta Centauri seems to be arguing that "if you don't hurt them, then they hurt you" is enough to make it "degrees of success," but that's just pushing the goal posts around. Under his apparent preferences, it goes from "well, that attack was a waste of time" to "well, I'm actively worse off for having tried anything at all."

Degrees of success are really easy for something like attacks - some benefit by exceeding defenses by a wide mark pretty much covers everything. Degrees of failure gets trickier, though from a simulation perspective missing horribly giving your opponent a bonus to strike you is completely fine (it's the whole "that's not going to hit me, I'm not going to block, and I might just chop at your arm as it goes by" defense).

Segev
2015-02-11, 11:50 AM
Sure, you can do that. But my point is, attack rules as a general thing already have "degrees of success."

The degree of success is how much damage you do.

I have also pointed out that White Wolf games have more attack successes lead directly to more damage dice.

"Degrees of failure" just means you're starting to invent critical fumble tables, at least in lite form. Miss by 5, and you just missed. Miss by 10, and you provoke an attack of opportunity. Miss by 20, and you drop your sword. Miss by 30, and you cut off your own foot.

These are exaggerated and not fine-grained, but you still get the idea. Even with "degrees of failure," the fact that you crossed a threshold is still there, and you failed completely (even if not catastrophically) if you miss by even 1.

Even if you had some rule that said that you do half damage if you hit by no more than 5, in order to introduce a "degree of success," that's mathematically no different than halving the damage of all attack calculations, and then saying that a hit by 5 or more gets double damage (is a critical hit).



I'm not saying the concept of varying degrees of success and failure isn't interesting. I am simply saying that there will always be a line between success and failure that is clear, unless the "partials" result in additional rolls/efforts.

In combat, you get additional rolls for a while, as you whittle through hp and try to keep yours from being whittled through.

If you assign obstacles some sort of "hit points," then you could achieve a similar result in non-combat checks. "Partial" success and failure would accumulate failure poitns or success points, and when you reach a threshold, you succeed or fail.

Another way to do it would be to create a system where you have a "complete success" and an "utter failure" value, and if you roll somewhere between them, you roll again with a bonus or penalty based on how close you were. Maybe set a difficulty, and success by 5 over is "complete" and failure by 5 under is "utter." Success or failure by less than that adds the amount by which you succeeded or failed to the next roll.

So if our Indianna Jones example involved an Athletics check which required a complete success rating of 25, and had a complete failure rating of 15, and Indy rolled a total of 23, the fluff is that he landed with his upper torso on the ledge, but his feet dangling, so he now has to roll another Athletics check with a +3 bonus to scramble up.

If he rolled less well, and came up with only a 19 on the next one, maybe he slipped further down, and his next Athletics check is at a net of +2 (suffering a -1 for the 19, but retaining the +3 from before).

Pushing it even further in depth would allow for some guidelines on choosing new skills to roll based on the changing situation, still with the bonus or penalty accumulated from before, until you finally completely succeed or utterly fail.

Jay R
2015-02-11, 08:34 PM
This is not really a problem, when you realize that how good a success it is can be the next roll. The first roll determines if the sword hit. The second roll determines how good a hit is was, when you roll for damage.

Mr. Mask
2015-02-11, 10:35 PM
That brings us back to my example in the OP. Minmaxing is the sensible option, as if you miss by 1, you get nothing.

Jay R
2015-02-12, 09:23 AM
... if you miss by 1, you get nothing.

Of course. If my arrow misses by two millimeters, I don't do a small amount of damage; I do none. If I miss moving silently by one little noise, then she knows I'm there. If I try to avoid a fireball, and almost make it, then I'm inside the fireball.

Some things really are binary. The problem in modern D&D (3E forward) is the silly notion that all actions can be simulated with the same mechanic.

Frozen_Feet
2015-02-12, 09:47 AM
Most old school games were also run as relatively static open sandboxes where the PCs are the ones making the play by trying to tame the wilderness. In such a situation "PCs do something else" works well (especially on a hex map). It isn't so good when the PCs are on a ticking clock because the enemy is active NPCs more than it is the untamed environment.

Old school games had active enemies specifically to force PCs to move on instead of repeating failed attempts. They were called "random encounters" or "wandering monsters". This is why they didn't need to key anything special to a failed check - it was already assumed failed attempts take some time, and after some time more things will happen. Degrees of success and failure already existed when you took into account the larger context of the rules. This is the same thing I tried to get at by using combat as an example. "Waste of time" =/= "no progress".


That brings us back to my example in the OP. Minmaxing is the sensible option, as if you miss by 1, you get nothing.

All games have a winning strategy. All games. Even ones where there's no official winning condition, as a player can set one for themself. You can not fight off optimization, only change what parameters count as optimized. This will be true even if a game has no obvious numerical parameters.

Mr. Mask
2015-02-12, 10:05 AM
Of course. If my arrow misses by two millimeters, I don't do a small amount of damage; I do none. If I miss moving silently by one little noise, then she knows I'm there. If I try to avoid a fireball, and almost make it, then I'm inside the fireball.

Some things really are binary. The problem in modern D&D (3E forward) is the silly notion that all actions can be simulated with the same mechanic. Well, two millimetres could mean hitting them in the eye, taking off their ear, or grazing their cheek, depending on your target relative to the head (centre head point or general head). It could mean a angled shot ineffective against their armour, it could mean the arrow hits something on its way over. The concept of "hitting," it abstract enough that you really can do as you please with it. If realism of missing by two millimetres is the argument, well, why doesn't how well you hit them effect damage? You make a little noise while sneaking up on someone, that doesn't always mean getting spotted. They might not hear it, they may see nothing and decide it was nothing, they may be fully alert, they may sound an alarm, or you may be able to attack as you make the noise so that you can still surprise them. And obviously, if you almost make dodging a fireball, tjhen you have reduced the damage significantly (unless getting out means still being badly burnt and stunned).


All games have a winning strategy. All games. Even ones where there's no official winning condition, as a player can set one for themself. You can not fight off optimization, only change what parameters count as optimized. This will be true even if a game has no obvious numerical parameters. Examining mechanics does have point, as does trying to change or refine things.

neonchameleon
2015-02-12, 10:40 AM
Old school games had active enemies specifically to force PCs to move on instead of repeating failed attempts. They were called "random encounters" or "wandering monsters". This is why they didn't need to key anything special to a failed check - it was already assumed failed attempts take some time, and after some time more things will happen. Degrees of success and failure already existed when you took into account the larger context of the rules. This is the same thing I tried to get at by using combat as an example. "Waste of time" =/= "no progress".

Point. Each failure has a consequence (Wandering Monster checks). It is however a very artificial situation and when you aren't playing in a specific playstyle, wandering monsters mess with versimilitude. So people dropped them.


All games have a winning strategy. All games. Even ones where there's no official winning condition, as a player can set one for themself. You can not fight off optimization, only change what parameters count as optimized. This will be true even if a game has no obvious numerical parameters.

You can also narrow the gap significantly.

Jay R
2015-02-12, 10:42 AM
If realism of missing by two millimetres is the argument, well, why doesn't how well you hit them effect damage?

It does, of course. But that's measured by the second die. There aren't degrees of missing. If you missed, you don't do damage. There are degrees of hitting, that affect how much damage you do.


You make a little noise while sneaking up on someone, that doesn't always mean getting spotted. They might not hear it, they may see nothing and decide it was nothing, they may be fully alert, they may sound an alarm, or you may be able to attack as you make the noise so that you can still surprise them.

All of not making noise, them not hearing it, them deciding is was nothing are all exactly equal results - successful hiding. The range of unsuccessful hiding, form being almost unheard but still getting their attention to standing out in the open yelling, "I got a 4!" all represent the bad guys reacting to you. A good DM will describe them differently. If you miss a roll by 1, he'll say, "They look a little distracted, and start to move away, when one of them says, "I think I heard something by that bush," and starts waking towards you. If you roll a 1, he might say, "You accidentally belch. All the guards look in your direction." But in either case, you've been spotted.


And obviously, if you almost make dodging a fireball, tjhen you have reduced the damage significantly (unless getting out means still being badly burnt and stunned).

If you are inside the fireball, you take the full damage, whether you fumbled and are in the center, or almost made your reflex save and are near the rim.


Examining mechanics does have point, as does trying to change or refine things.

Of course. I prefer that gradations of effect be handled by the DM, so we don't need a 10,000 page book detailing all possibilities for every kind of roll in every situation.

Knaight
2015-02-12, 10:50 AM
If I miss moving silently by one little noise, then she knows I'm there. If I try to avoid a fireball, and almost make it, then I'm inside the fireball.

Some things really are binary. The problem in modern D&D (3E forward) is the silly notion that all actions can be simulated with the same mechanic.

These have the mechanics be binary, but there's no particular reason they have to be. There are obvious degrees of failure that could be employed. For instance, a noise could reveal that you're out there somewhere, but it was quiet and brief enough that only a vague direction is known. Alternately, it could reveal your exact location. As for the fireball, while D&D currently uses uniform damage for the entire area, it decreasing with distance is entirely reasonable.

Segev
2015-02-12, 11:35 AM
Technically, the "they didn't hear it" is covered by the fact that your Stealth check is opposed by their Perception check. It all ultimately breaks down to whether they know you're there or not.

The "they know you're out there, but not where" is handled by other mechanics: If they'd seen you before you hid, they know you're out there somewhere; if you have Concealment, they may know you're present by your failed Stealth check, but they cannot pinpoint your position.

Other systems may have other mechanisms; the "concealment" one is D&D's.

Degrees of success need not be part of the one roll you make. In fact, it's usually better if it isn't. Games which have rules governing situations tend to be most rich when a lot of mechanics and factors work together to determine the simulated reality. Rules-light games are almost always binary; they assume the GM and players will use that as a guide for whatever degrees of success or failure they want.

Mr. Mask
2015-02-12, 10:20 PM
These have the mechanics be binary, but there's no particular reason they have to be. There are obvious degrees of failure that could be employed. For instance, a noise could reveal that you're out there somewhere, but it was quiet and brief enough that only a vague direction is known. Alternately, it could reveal your exact location. As for the fireball, while D&D currently uses uniform damage for the entire area, it decreasing with distance is entirely reasonable. Precisely.

Segev
2015-02-13, 09:07 AM
Notably, D&D 3e did have some places where degree of success came into play. The one that occurs to me right now is Knowledge and Bardic Knowledge checks. Modules and the like which had background details which various Knowledges and the ever-covering Bardic Knowledge could let PCs just happen to know often had the player make a single roll, and then compare that to a chart of DCs. The highest DC they exceeded and all DCs below that were information the character was considered to know.

This is a very straight-forward sort of "degree of success" mechanic. It does, obviously, still have a binary threshold below which you know nothing of relevance to the topic, though.

NichG
2015-02-14, 09:19 AM
The system I'm currently running uses two design conceits:

- Proactive action is always favored over mechanics that help one retain status quo. So attack is prioritized over defense, things that cause wounds are prioritized over things that heal them, etc.
- Characters have a resource pool they can expend to shift that balance.

The result is that most rolls are not about 'do you succeed or do you fail?', its about 'how much does it cost you to secure the outcome you want?'. So that's a very simple way of adding in degrees of success.

That said, I think a better starting point would be to actually start from a system with no concept of chance associated with success or failure outcomes in the first place. So the kinds of questions you ask with dice are just different. You don't ask 'did I jump the gap?' with dice, because its a situation which is defined too strongly by success or failure - so instead, you have an ability on your sheet that says 'can jump gaps of 10ft or less', and so there's no randomness in your answer - you either jump the gap or you don't. On the other hand, things like how the world interprets your actions, whether situations become simpler or become more complex, etc would be the sorts of places that die rolls would be used.

The idea would be that you have certain parameters of the scenario that shift as a result of taking 'risky action'. Those parameters would act as prerequisites for using particular abilities, or having certain kinds of events, and so the meta-conflict would be to try to guide the scenario in a direction that allows you to use big abilities. On top of that, you'd have a timer that increases every time someone takes any action, which pushes the scenario towards conclusion. So for example, maybe there's three state variables: Hostility, which determines whether or not actions in the scenario are primarily cooperative or antagonistic; Intensity, which determines how much people are risking in the conflict; Complexity, which determines whether the situation is straight-forward or has many factors that can intervene. Characters could have one or more 'hitpoint' style resources that are depleted through interactions, such as e.g. 'Stamina', 'Leverage', etc.

In such a system, if someone uses something that would be considered an attack (Sunder; Hostility > 3, Intensity > 3: Destroy one item belonging to the target unless they pay Leverage based on the result of a die roll), then the dice pool could be interpreted in terms of how it modifies the state variables in addition to the cost to resist. So e.g. maybe someone has an ability that lets them risk +1 Intensity in exchange for an increased chance of success or increased cost to resist in the form of adding an 'Intensity die' to the die pool.

At lower Intensity, the fight would tend to be more about dominance: does one party step down and let the other have what they want? So a fight between parties at low Intensity means that everyone is nursing wounds, but no one gets killed - people flee before it gets to that point. On the other hand, a fight between parties at high Intensity may mean that people are killed outright by attacks, or at very high Intensity that the actual goal of the fight is destroyed by the battle (e.g. if the goal is something like 'defend this town' or 'prevent X from being kidnapped', then a high Intensity fight means that even if you succeed the town may be damaged, or X is killed rather than captured).