Failure=Advancement

[C. Edwards]

There has been some discussion in the Level concerns thread about character advancement through failure.  Learning from one's mistakes takes place in the real world and it seems only natural to me that it should translate well into the realm of RPGs.  I'm using a loose form of  failure=advancement in the game I'm currently working on and I'd like to know of experiences others have had with failure=advancement mechanics.  What, in your opinion, worked and what didn't?

Here's a basic breakdown of what I'm working with:

-Two 12-sided dice used for resolution.  One of these is the Fate Die and the other is the Fortune Die.  The Fate Die represents any obstacles that a character attempts to overcome as well as being representative of fates predetermined path and relative indifference to the character's success or failure.  The Fortune Die represents the character and how in-line his actions are with fates whims.  The player rolls both dice at the same time.  To be successful, the Fortune Die must be higher than the Fate Die.  The Moderator only rolls dice in PC vs. PC conflicts, and then only the Fate Die which takes the place of the players Fate Die.

-Desire and Backlash are the main currencies.  They represent the character's tug-of-war with fate.  Desire can be spent to increase the total on the Fortune Die.  The amount of Desire burned must be stated before the roll.  Desire is gained on a failed roll where no Desire is spent.  The amount gained is equal to the difference between the Fate Die and the Fortune Die.  A player can put aside some Desire (12 points?) to gain a "Fortuitous Turn of Events" which can be used to gain an advantage for the character.  Backlash is the result of "extra success" on a roll where Desire is burned.  The amount of Backlash accumulated is equal to the difference between the Fortune Die and the Fate Die.  When a certain amount of Backlash accumulates (12 points?) it becomes a "Catastrophic Turn of Events".  The Moderator can use a CTE to turn up the heat for a character.

-Characters are defined by Goals, Motivations, and Style.  Style is similar to a Cover in Sorcerer and loosely defines the skills and abilities characters use to overcome obstacles.  Only Goals and Motivations have any numerical value and this is the amount of Desire assigned to them at any particular time.

Those are the basics and they still need some work.  I left out quite a bit (narration, descriptions of Goals and Motivations, injury and death) to keep the post from being unbearably long.  If you have any feedback on the little I've posted please feel free to give it.


-Chris

[Ron Edwards]

Hi Chris,

I know this exact topic has been discussed before, somewhere on the Forge. Can anyone help me find the thread?

The fundamental issue to consider is the role of character improvement in the long-term course of play. Although most role-players see both improvement in the first place and its expected rate as givens, this is merely the historical impact of the original designs at work and not a constraint we need to work under.

Let's take it as a given, though, just so people don't blow a gasket, that improvement of characters' abilities will occur via play in your game.

So now let's look at rates. The typical rate from the original RPGs is exponential: you start with very poor abilities, you improve slowly, but the more you improve, the better you are at garnering the reward mechanic, hence you improve faster, hence you get better at garnering the reward, and now you're skyrocketing towards godhood and have totally broken the resolution mechanics of the game.

D&D until recently was especially unpleasant in that one survives the first phase only by GM sufferance and then after that one is practically unkillable or un-affectable except through arbitrary means. Arguably, all the substantive changes to D&D for the 3rd edition can be seen as directly changing this very phenomenon about the game.

The BRP system (old RuneQuest, Call of Cthulhu, others) has a double-dial: as you succeed more often, you get more chances to improvement, but the chance to improve itself dwindles. This tends to have a "sweet spot" effect as well, but the curve is S-shaped rather than exponential, so the enjoyable phase is prolonged and the final phases have a curiously unsatisfying feel during play. Most BRP games' editions after the early 80s made sure to start characters after that point (unlike the original editions which started them with skills of 15% and 20%), and slowed the process so that the final phase was usually never reached.

The most widespread system used today (begun in the Hero System, preserved in GURPS, and thus now spread throughout all of White Wolf games and their innumerable imitators) is less coherent than the preceding two. It basically treats anything as bought in "points," and is subject to many of the criticisms I raise in my big essay about Currency, as what a "point" is worth changes radically through the course of play.

Hero Wars solves this problem very, very well. It represents one of the few point-based games in which improvement does not throw play-function all higgledy-piggledy. You can think of this model as a straight line rather than as a curve (although it really represents a series of stepped lines, but never mind).

Very few RPGs have used a negative-exponential rate, in which although Swashbuckler deserves an honorable mention. You begin pretty good, but your improvement rate is quite rapid ... briefly, as it slows and will continue to slow through the course of play. Therefore the decisions you make about what to improve through the first few scenarios will pretty much "set" your character (in terms of what's better than what) for good.

The reason I went into all this hairy detail is because your improve-through-failure mechanic isn't going to be "good" or "bad" based on any considerations of realism - if we try to discuss that, it'll go 'round in circles for days and we will have departed from practical issues of RPG design in the first two minutes. It will succeed (relative to your goals, specifically that of enjoyable play) only if its rate, of the three I described above, is suitable to the other aspects of the game.

So which of the three is what you'd like to see? Why? And then when you've answered that, make sure the numbers fly that way.

Best,
Ron

[damion]

Q:If Fate==Fortune, who wins?

Actually, I'm confused by ron's post, as charachters don't seem to advance at all in this, at least not in the traditional sense of advancement. Of course this seems to be more of a narrative game of struggle and an traditionl RPG, so advancement isn't really  apropriate.


Other points: As I understand it, your better off not spending Desire on a roll unless you spend 11 or 12 points, i.e. make it automatic. This is because if fail, but don't spend it, you get Desire, but if you spend it and fail, you get Backlash, so your better of saving up for Fortuitious turn of Events, which is automatic, or burning alot of Desire to auto-succeded if a good situation presents itself.
A possible solution would be to have ALL failures generate Desire, with 1 Backlash for every 2 desire gained or something.  (Mainly so it doesn't look like the two balance out. Since Backlash is controlled by the GM it is actually more powerful. )

I admit this could be modified by how Goals and Modivations work.

[C. Edwards]

Thank you for the replies, and for going into such detail Ron.

It will succeed (relative to your goals, specifically that of enjoyable play) only if its rate, of the three I described above, is suitable to the other aspects of the game.

In play I want the game to have a somewhat cinematic feel with a very bitter-sweet edge.  Regardless of whatever setting or other elements are layered on top of them, the mechanics are supposed to represent the characters struggling against whatever the whims of fate have in store for them.  When the characters start pushing fate starts pushing back.  Basically, failure can add to a character's potential to succeed but using that potential not only reduces that potential but can be the cause of future failures.  As Helen Keller put it, "Security is a superstition.  The fearful are caught as often as the bold."  With some fine tuning I think the mechanics will work and should promote dramatic and creative play.

damion wrote

Q:If Fate==Fortune, who wins?

Actually, I'm confused by ron's post, as charachters don't seem to advance at all in this, at least not in the traditional sense of advancement. Of course this seems to be more of a narrative game of struggle and an traditionl RPG, so advancement isn't really apropriate.

I'm still debating options for the results of a tie.

You're right, there isn't any conventional rpg advancement.  Mechanically a character's potential for success will ebb and flow with play.  The characters can create game-world success by accomplishing goals to fulfill their motivations.  They can slay the evil sorcerer, discover the truth behind the murder of the president, etc., and impress their friends, gain riches, respect, etc.  I want widely divergent character concepts to all have the same basic opportunity to succeed.  The Grizzled Knight, the Parapalegic Accountant, and the Martian Empath all have the same chance to achieve their goals.  What seperates them is how they go about it; their Style.

Other points: As I understand it, your better off not spending Desire on a roll unless you spend 11 or 12 points, i.e. make it automatic. This is because if fail, but don't spend it, you get Desire, but if you spend it and fail, you get Backlash, so your better of saving up for Fortuitious turn of Events, which is automatic, or burning alot of Desire to auto-succeded if a good situation presents itself.
A possible solution would be to have ALL failures generate Desire, with 1 Backlash for every 2 desire gained or something. (Mainly so it doesn't look like the two balance out. Since Backlash is controlled by the GM it is actually more powerful. )

You only get Backlash if you burn Desire and succeed.  That is fate pushing back.  So how much you burn is a real issue.  If you gain 12 points of Backlash on one roll ( which can happen if you only spend 1 point of Desire, let alone 12) you've automatically given the Moderator a Catastrophic Turn of Events that will be used against your character.  Turn of Events can only be played after the normal narration for a resolution roll, so instead of an automatic success a Fortuitious Turn of Events becomes a "but fortunately".  This is, IMO, less powerful than an automatic success.

Since this is a less than forgiving system and I tend to agree with you about the power of Catastrophic Turn of Events, I plan on giving players a Cut-Scene whenever a CTE is played on them.  Let's say, for example,  that a character is outside the door to what he believes is the mad scientist's laboratory.  The player can use a Cut-Scene to have the Moderator narrate a scene concerning what is currently happening on the other side of the door.  The character wouldn't actually see this but the player could use the information to help decide a course of action.  Failure would be used again to promote success.

Hopefully that helps clear some things up.  I really appreciate the interest.

-Chris

[M. J. Young]

This has bothered me since I read this the other day. I don't know if it bothers you, but let me lay it out maybe you'll see what I'm after.

The odds of one die rolling a higher number than another of equal sides is always 50% minus the probability of a tie. In this case, that probability is 8.33%. Thus if (as stated) the character only wins if the fortune die is greater than the fate die, the character will win 41.67% of the time and lose 58.33%. I personally don't like those odds. You could turn it around if you stated that a tie goes to the player, making it 58 to 42, which is better. But it also might not be the point.

I suppose the point that comes in is at the level of when and why you use the resolution system at all. Personally, I don't take risks; I don't bet on 50/50 chances. I don't like to bet on 75% chances. But I don't  think most of life is a 50/50 chance. I think we overcome obstacles much more frequently than that. Someone (Lee Iacoca perhaps?) said that the problem with American education was that it taught people that a 25% failure rate was a passing grade; in the real world, a 2% failure rate is about the minimum most companies will accept. You've created a world in which our efforts have at least a 42% failure rate (depending on what you do with the ties).

I recognize that the desire points are supposed to mitigate this; but it is equally clear that they create greater risks. It is also a currency that demands waste. That is, I can't guarantee a roll will succeed unless I spend 12 points. If I spend less than that and the roll fails, I've thrown away the points. On the other hand, if I spend those twelve points and the roll would have succeeded anyway, I've created a catastrophe automatically.

I immediately note that it is not possible for any roll to automatically create a fortuitous turn of events; only a catastrophic turn of events can be automatically created. It could (1/144 chance) happen if I spend just one point (if I roll 12 against 1 and spend a point, I've got 12 fate points immediately). The risks of spending the points is far greater than it sounds; the benefits far less.

If I were trying to tweak the system, I'd probably do this:
--Make the expenditure of desire points a bid system that doesn't require full payment; that is, the player can say "I'll bid six desire points" on this roll, and then up to six points are taken from him if that would make the roll successful. He loses no desire points if the difference is more than six, and he does not lose all six points if less than six will make it a success. Then at least he is not throwing away his desire points.
--Backlash now only comes in when desire points are offered unnecessarily; that is, if I say I'm willing to spend one or more desire points to make the roll successful and it proves to be successful without that expenditure, the backlash points come in as the difference between the Fortune die and the Fate die, but is not inflated by my desire point bid.

As it stands, I'm penalized if I don't bid enough desire points (because I've spent the points and gotten nothing) and I'm twice penalized if I bid too many (because I've spent more than it cost and I'm repaid in backlash). Change that, and you're getting closer to something workable, I think.

I'd still want a way to create uneven odds. Two ways spring to mind:
--The referee could also have a pool of points which before the roll he announces that he is spending. I think it might make sense for this to be a self-balancing pool, that is, he starts with five fate and five fortune points, and every time he spends a fate point to make the fate die higher he earns a fortune point, and if he spends a fortune point to make the fate die lower he gets a fate point. But then, this might not work for every scenario.
--The referee could set difficulties by declaring unequal dice. There would be some situations in which the fate die was only d10 or even d8, others in which it was the fortune die that was lowered. This obviously skews the odds significantly, but it never gives a guaranteed success or failure unless fortune points are spent.

Those are my reactions. I hope they're helpful.

--M. J. Young

[C. Edwards]

Thanks for taking the time to post M.J.,

The dynamics don't bother me.  I'm not a big proponent of reality modeling and that's not my goal with this system.  I'm going towards more of an algorithm for dramatic (cinematic?) storytelling, and I'm not too concerned about happy endings.

Failure and success aren't necessarily black and white. The resolution rolls will be geared towards resolving a whole scene, not an action by action break down, so a great deal of narration should be taking place.  I find that narration tends to turn black and white into various shades of grey.  

As in many other RPG's, a roll is only required when a character is involved in something that is dramatic, a key part of the ongoing story, or when he is in way over his head.  So if you include all theoretical game-world success, not just the more intense situations that resolution rolls will be aimed towards, people generally succeed far more often than 50% of the time.

I've purposefully made it where a player, if he runs out of Desire, is forced to rely on 50/50 chances until he gains more Desire.  I don't think life is a 50/50 chance either but games are not life.  Games also are not movies, so indulge me for a second while I use a an example from film to give an idea of what I'm after.

In Die Hard Bruce Willis is faced with a situation where he must get past a couple terrorists armed with heavy weapons.  His only path is through a room with glass walls (or big windows).  He has a machine gun, but no shoes.  Bruce knows that if he doesn't succeed in getting past these guys he may very likely end up dead.  Bruce only has 12 Desire points available to spend.  Bruce figures that whatever unknown future catastrophe he may generate can't be worse than being dead now, so he spends all 12 Desire on this roll.  Two dead terrorists later Bruce is alive.  The Moderator took advantage of the Catastrophic Turn of Events that was generated so Bruce is going to be busy the next few minutes pulling large shards of glass out of his feet.  His life is going to get even more hairy for a bit because he's out of Desire and will only have 50/50 odds on his next roll.  This example excludes some things, like injury mechanics, but is a fairly good example of how play should go.  And I expect the system to be good for more than gun-toting action.

In the movies I'm sure Mr. Willis has more than a 50/50 chance for whatever happens next, but for the purposes of this system that suites me fine.  Spending enough Desire for an automatic success should be a fairly rare occurence.  I expect normal Desire expenditures to be around  4 points, enough to increase the odds of success to over 80%, for as many rolls as possible before having to rely on 50/50 odds.  Ofcourse, if a player wants to save Desire for a guaranteed success thats fine too.

You're right about Fortuitous Turn of Events.  The points required to gain one should be lowered.  I'll probably start playtesting with the requirement at 6 Desire and see how that works.  I'm also going to try your proposed bidding system or some variations on it.

I don't agree with you about bidding Desire and still failing being a penalty.  That seems to me like having a fencing skill at 85% but everytime you fail a fencing roll you declare that the points you spent to get your fencing to 85% were wasted.  As far as bidding too much goes, those are the chances you take.  It's all about maximizing potential for success while minimizing the possible repercussions.  Basically, if you want it bad enough you can have it, but you'll probably get a little burned yourself in the process.  The main thing to keep in mind is that Turn of Events aren't actually successes or failures, and they can't overturn the results of a resolution roll.  They're more of a way to create tension and drama.  Sort of a micro-bang, if you're familiar with Sorcerer.

I'm still undecided about ways for the Moderator to create uneven odds, or even if there should be.  As it stands the odds are really in the players favor.  Without Desire they have a 50/50 chance, and with Desire they can increase their chances, plus Fortuitous Turn of Events and Cut-Scenes.  The only tool, besides narration (the players also narrate), at the Moderator's disposal is the Catastrophic Turn of Events.  Right now I like it this way, but that may very well change.  I'm sure that during playtesting I shall be gifted with an epiphany and all will be made clear to me in neon Sanskrit.

Thanks again for the feedback M.J., I really appreciate it.

 -Chris

[Robert K Beckett]

The fundamental issue to consider is the role of character improvement in the long-term course of play.

AND

your improve-through-failure mechanic isn't going to be "good" or "bad" based on any considerations of realism

Ron's points about concentrating on the long-term course of play and not getting into the whole "realism" debate is spot-on. I think these two statements go hand-in-hand in explaining some of the resistance I have seen to the Failure = Advancement mechanic.

On the one hand, my goal is to automatically reward PCs of any "ability level" for attempting challenging tasks. This would take place over many play sessions, so that the net result over time is that those PCs who attempt tasks often and under challenging circumstances become more proficient at those tasks.

On the other other hand, people freak because the actual instance of advancement is the exact opposite of what they are used to. It flies in the face of convention - perhaps one of the most time-honored conventions in RPGdom: You reward success with experience points. That's just The Way It's Done. This, in spite of the exponential, system-breaking nature of this convention. And the fact that success has its own rewards (ie you did what you were trying to do).

Plus people say it's not "realistic" because people don't learn only through failing. For instance, they want to give advancement points for crit Success as well, which completely short-circuits the mechanic. They just keep trying to interpret the mechanic as modeling the particular learning event itself, rather than contributing incrementally to the long-term growth in capability.

Very few RPGs have used a negative-exponential rate, in which although Swashbuckler deserves an honorable mention.

I would be interested to know how that worked exactly. I wonder what kind of curve the Failure = Advancement system would have.

In such a system, once you earn a high-enough skill level such that you rarely fail when using a skill, you stop advancing in that skill. Or, another way of looking at it: If you use a skill to accomplish only easy tasks, you will not advance in that skill.

Thus advancement is on a sort of "needs only" basis, where the system/GM detects the PC's "need" to advance a particular skill by the frequency of failure, then ameliorates the "need" by rewarding Advancement Points for that skill.