Proficiency and Pitfalls

[Windthin]

Well... I've been thinking about how to show mastery, proficiency, in skills, and these are some thoughts that occured, and a possible solution I'd like feedback on.

    First off, I've looked at set dice versus dice pools, and decided I prefer the former.  Dice pools by their very nature are limited, because you can have only so many steps of mastery before you're rolling a bloody great number of dice.  My wife also points out that more dice often seems to mean more chances to botch, and I've seen this tendency myself often.  None-the-less, there are some nice things about dice pools, and they should not be discarded entirely.

    My problem lies in the desire to reflect mastery as a character grows in their abilities without entirely leaving behind the infamous botch, the chance that something goes horribly wrong.  I don't feel that a master swordsman should possess the same chance of something going truly awry as a rank novice.. there's simply a question of skill and experience involved, plus the fact that people who gain mastery typically learn to avoid total blunders or at least morph terrible mistakes into minor ones.

    I tossed around many ideas.  Negating failures altogether doesn't suit me... I feel there is always the slim chance for something to go wrong.  I know some disagree, and I can understand the desire to remove that bombshell from play.  I thought of adding one or two additional dice to a roll for a master, lesser dice that would add into the roll and stave off distaster.  But that's more dice to deal with, and too little of a chance, perhaps.  Re-rolls carry their own problems, in that if a person does have a re-roll function, utter failure is unlikely to happen again.  I know that in a game I currently play in, the player can shout FATE, and if they have any Luck left in their pool (which they never know if they have or not), can re-roll a failure, rank or regular.  This, however, really does reflect luck and not skill, and while not a bad mechanic, does not, again, do what I seek.

    What I finally hit upon as a possibility is pools of mastery; that as a character develops skills or other abilities to certain levels, these abilities gain their own pools of mastery, expertise, proficiency, whatever-you-want-to-call-them points.  These points are spent then to turn a total failure into a mild one, and perhaps (an idea I yet toy with) raise the level of success in general, within reason.  I have not decided how such a pool replenishes yet, but it is quick and easy to keep track of, all in all, and if a character continues to fail abysmally, well, luck has turned against their skill, perhaps.  Like I said, it's a raw idea... and I'd appreciate feedback.

[iago]

Er, hm.  When you say "botch" are you talking about "no successes rolled" or are you talking about a subset of the zero-successes condition.  In my mind a botch would be that, e.g., "rolled a one on every single die".  If you were rolling d10's, then this kind of botch definition (or really any one that involves the number range being less than 50% of the die's range) does minimize the more dice you're rolling (1d10, 10% botch, 2d10, 1% botch, 3d10, .1% botch, etc).  Maybe I'm missing your point?

[Bill Cook]

The only botch I know of is hitting the lowest end of a fixed range (e.g. 1 on a d20.)  Not sure what you mean by set dice . . .

So you want to decrease the botch range as the success range for a skill increases?  (e.g.  At 1st level: 1-3 = botch, 4-10 = failure, 11-20 = success; at 2nd level: 1-2 = botch, 3-8 = failure, 9-20 = success.)  Sounds obsessive.  How prevalent is botch to your game?

[Ron Edwards]

Hey Windthin,

I can only recommend to you the game formerly called Hero Wars, now called Hero Quest. It treats critical success, success, failure, and critical failure in an entirely new fashion, relative to "master" level degrees of competence.

I also think you may be confounding "dice pool" with "White Wolf games." There are plenty of systems which use a varying number of dice without running into some kind of botch or critical failure in the way you're describing.

Best,
Ron

[M. J. Young]

Assuming by "botch" you mean,

that the character may not merely fail but actually and seriously hurt himself or cause some significant serious accident in the process. Chance to botch means not only did he not learn to operate the gun properly, he also shot himself; not only did he not fly, he also fell.

let me point you to Multiverser (from which that is quoted).

Multiverser uses what I take you to mean by a "set dice" system; skill checks are percentile based, and chance of success is the sum of a couple numbers. You're looking for a way to reduce botches without eliminating them entirely. This was our approach. Your chance to botch is one tenth of your chance of failure. As your chance of success goes up, your chance of failure drops, and your chance to botch decreases proportionately.

You can't get rid of the chance to botch until your chance of success exceeds 90%. It can do that; it can exceed 100%. However, because of unseen adjustments for bias and situation modifiers, neither the player nor the character can know with certainty whether his chance of success on this roll is that high.

Because it's a percentile system (low for success) everything can be incorporated into a single roll. Target number or lower succeeds, above that fails; both success and failure are relative, so the roll represents degree (such as weapon damage) as well as success; top few percent are botch indicators.

There are other ways to do it. There is a percentile based system out there in which all numbers which end in a specific digit are critical successes (and thus, the greater your chance of success the more chances you have to roll a crit), and you could do the same thing on the failure end--if your percentile system says any failed roll that ends in 0 is a botch, then you get something very close to the same odds for botch as Multiverser (slightly higher, as your break points are shifted--there would still be a 1% chance to botch at a 99% chance of success) without the mental number crunching (what is 10% of the chance to fail given a 64% chance of success? it's 3, 98 or higher, but I've been doing it for a while).

This might point you in the right direction.

--M. J. Young

[Andrew Martin]

I don't feel that a master swordsman should possess the same chance of something going truly awry as a rank novice.

... and I'd appreciate feedback.

Miyamoto Mushashi was definitely a master swordsman who fought in wars and fought many duels without loosing. This seems to me to indicate that he made no "botches" at all. Could a player with a skilled PC do the same in your perfected system?

Closer to home, chess grandmasters (almost certainly people who have mastered the skill of playing Chess) have on occassion made mistakes. Are these mistakes "botches"? After all, it really depends upon how skilled the opponent is to turn a chess grandmaster's mistake into a "botch". Wouldn't the results of the mistake by the Chess Grandmaster be different if the opponent was a novice, another Chess Grandmaster, or chess program running on a supercomputer?

Strangely enough, Idiot-Savants regularly produce spectacular works of art, yet regularly "botch" all other areas of skill. Is an Idiot-Savant PC possible in your game system?

Are you sure that your goal of producing botches for skill resolution is realistic and reflects real world skill use?

Or are "botches" really just a tool to smack down cocky players with?

[Windthin]

A few clarifications... yes, by "set" dice I referred to any system where you have a certain range, and specific dice for it.  3d6, 1d100 (don't laugh, I come from years of on-line gaming as well as off, and still play in a game where d100 is the die used), 2d10, so on.  And by "botch" I did indeed mean catastrophic failure.

To be honest, I think the botch has always been a means of saying "you're never wholly perfect", just as some games won't quite let you succeed ALL the time (case in point would be Rifts' 99% cap).  I don't know that I agree with either.  The problem I see with botches is that they don't feel realistic, even as I admit that Murphy's Law is always hovering like a specter.  There is the pure possibility of simply removing the chance for catastrophic failure after a certain level... but I am not certain whether or not I want it gone wholly.  There's an appeal to the chaos factor, even as it disturbs me.  I get to wondering if, at higher levels of mastery, what qualifies for a terrible mistake takes on lesser value... the mistakes of a Chess Grandmaster as opposed to a rank novice, or a moderately skilled player.  If it is all relative... perhaps that is my solution, and I can see the removal of this phenomenon altogether.


M.J: Your system sounds like one I am somewhat familiar with, play and did minor design work on now, though more advanced in its handling of that particular facet.  I will look into it, and perhaps suggest it to the bear (the main creator of said game).


Ron: Thanks, I'll take a look at that.  I am familiar with White Wolf, yes, but it is not the only game by far I have come across with dice pools... I've explored a lot of systems, and never turn down the chance to try a new one, to find what portions of it I like and don't.  I suppose I might change my previous statements to "dice pool" games that also include the chance of botching.  In general, however, I've still been a little unhappy with dice pools, since they seem limited to me by the number of dice one can reasonably expect players to be rolling (and counting up) at high levels.


Bcook1971: A botch is basically a critical failure, however a system gauges it... in one I knew, d100s were used and rolling a 100 was a botch (there were some fairly seriously design flaws at times in that system, but it makes a good example for this).  As for prevalence... not nearly as much as the situation you outlined there, but enough so that I don't feel a master and a novice should face the same odds.

Iago: I've not really seen the botch situation you've mentioned here, where every single die most lowend... where that the case in a dice pool situation, yes, it would most definitely minimize the more dice were added, and it is something to consider.