my "house rules" die mechanic

[talysman]

as I've mentioned, I'm developing a set of core or "house" rules that I will adapt to specific games in which the setting is the primary focus, sort of the way Chaosium adapted BRP to several settings. each game will use the same basic rules plus specific rules integral to that setting. I will still have a couple games that don't use these house rules at all, like C09C and the game that Jonathan Walton asked me to write last year (which I suppose I should finish this month...)

anyways, the house rules at first used a die rolling mechanic patterned somewhat after The Fantasy Trip, which you can see in the first version of Empedocles: roll 3d6 under a TN for ordinary actions, roll more dice for more difficult actions, talents can reduce the number of dice rolled or add difficulty dice to an opponent's roll. the object of rolls is to earn steps of success, with the goal being 10 steps of success (basically, every goal has up to ten hitpoints.) there was a critical success/failure method that altered how many steps could be earned on each roll.

now, as I fiddled with this basic mechanic, I had two main goals: simplify it as much as possible to make the game quicker to learn and faster to play, and eliminate a couple bugs (like a problem with the critical success system.) my first change was to try it without a variable TN; instead, roll under 12 on 3 or more dice. this is sort of OK, but I would really like to lower time spent adding up dice totals, so I fiddled some more...

now, I have come up with the following idea: stop being slavish to TFT, switch to d10s (counting 0 as 0,) and turn the difficulties and target on their head. in the new method, you roll 1 die for a near impossible goal or 9 dice for an ordinary goal anyone can accomplish, with one in-between difficulty level of 5 dice for "out of class" goals (people without training attempting a specialized task.) instead of rolling below a TN, you roll 9 or better. I like this, because you don't actually have to add all the dice: stop when you hit 9. most people should be able to "eyeball" the die roll results and tell immediately whether they've earned a step of success.

for extra successes, anytime you get multiple matches for a number, that gives you another step of success. thus, if you roll 9 dice and get 0, 1, 1, 2, 3, 3, 3, 3, and 4, you earn one step for beating the TN of 9, plus you have two matches (two 1s, four 3s,) for a total of three steps of success. with this rule, you can earn up to five steps of success on a 9d10 roll.

talents, of course, can allow you to add dice before you take your roll or add another step of success if you beat the TN. an opponent can also use talents to subtract dice from your roll.

this technique looks good to me for a couple reasons:

(1) it's intuitive: more dice means you reach your goal faster;
(2) the "eyeballing" for 9 or better seems pretty quick;
(3) the opponent's use of talents can happen after a roll, which means that the player and the opponent can trade descriptive "flourishes" back and forth as they negotiate the final number of successes;
(4) IIEE looks nice: state general intent to figure number of dice to roll, action actually begins when dice are rolled, details and effects are processed during the haggling.

here's my question: does this seem quick and easy to everyone else? are there any potential snags I'm overlooking? does the die roll matching part seem too complicated?

[Emily Care]

now, I have come up with the following idea: stop being slavish to TFT, switch to d10s (counting 0 as 0,) and turn the difficulties and target on their head. in the new method, you roll 1 die for a near impossible goal or 9 dice for an ordinary goal anyone can accomplish, with one in-between difficulty level of 5 dice for "out of class" goals (people without training attempting a specialized task.)

I like it. Reminds me of something Nathan talked about for Draconum: rolling one die for the final fight. Makes more sense when you think about it.  

here's my question: does this seem quick and easy to everyone else? are there any potential snags I'm overlooking? does the die roll matching part seem too complicated?

Matching seems fine.  The doubled numbers are easy to pick out. My question would be about the probabilities. Will it work out the way you want?

--EC

[Mike Holmes]

So, worst case scenario, I have to roll one die to get a 9? So the lowest chance is 10%? And that's to do something that I'm really no good at?

Couldn't you find a way to just say that things are impossible (not worth a roll) after a certain point, and make 10% just "hard"?

Mike

[Ben O'Neal]

Actually Mike, if you have to roll "9 or higher", wouldn't that be a 20% chance of success (two numbers, 9 and 10, out of ten possible results)? Or maybe I misread and only '9' is a success, and '10' is nothing...?

I'm not clear on what "steps of success" mean. I'm assuming they are like degrees of success, but how are these quantified/qualified in-game? Does it matter that you can't achieve any steps of success when rolling 2 or less die (for really difficult tasks)?

Is 9 the highest number of die you can roll? Like is doing a specialized task that you are highly trained at the same as doing something that anyone could do? Or is there a mechanical difference? And if 5 die is for a specialized task that you don't have training in, does it matter that you will actually have a pretty good chance of success?

So far it looks like the mechanic would work and would be fairly fast to use, and the die matching part is hardly more complicated than many other quirks out there (I think preschoolers could do it pretty easily), but I'm just wondering if the maths of it are what you are aimig to achieve.

[clehrich]

If 0 is worth 0, then it's a 10% chance.

But what's not clear to me here is the issue of steps, as someone else mentioned.  If I roll one die, and have a 10% chance, that's a chance of getting one step.  Does that mean a bare-minimum success?  Do multiple steps make me succeed better, or might there be multiple steps required?  You said something about the original FT having 10 hitpoints in effect for each task; would this system here mean that if I'm rotten at something, I end up having to make 10 separate rolls, at a minimum, to get my 10 steps?  Does that factor as time, so that I can succeed but it takes a while, or what?

Could you construct a little example of how this would work in practice?  Without understanding what "steps" mean, it's difficult to assess the mechanic on its own.

[talysman]

Emily: I already tested all the probabilities in a quicky script I wrote and it seems in keeping with what I want. whether other people will like it will remain up in the air for a while, of course...

Mike: a 1-die roll is for a "nearly impossible" goal -- which shouldn't happen too often, but it's there to be used anyways, if needed. impossible goals -- and impossible-to-fail goals -- do not get a die roll, nor do unchallenged goals for which speed and quality do not matter.

I guess a better label for the 1-die category would be "impossible to learn". like magic in a rare-magic setting that doesn't include magical professions.

Chris and Ravien: some of what I've said about the whole step system is strewn about in various disconnected places, so I should really re-explain it all in a more concise manner. the in-a-nutshell concept behind "steps of success" is that it works like "successes" or "victories" in other dice pool systems, but I like saying "steps" because it's shorter. also, I tend to think of the steps for each goal being tracked with counters on some kind of "board" or playsheet; I mention this approach in the Court of 9 Chambers, for example.

here's a nearly-complete explanation:

first, my general philosophy on goals is that if a goal is (1) realistically possible, and (2) unchallenged, then you will always succeed. think of an experienced, well-equipped camper setting up camp and preparing a meal in the woods under good conditions: sure, when looking for firewood, the first branch or twig found may be unsuitable, but the camper will eventually find good wood... the first few sparks may fail to light the tinder, but eventually it will be lit... and the camper might spill the first batch of stew or burn the first batch of panbread, but eventually, the meal will be cooked.

so this means I don't feel like testing success/failure in some of my games, but instead testing who succeeds first or who does a better job. hence, the steps of success idea: both sides in a conflict have a goal, maybe even several goals. the object is to reach your goal first.

since this is a core or house mechanic, it's kind of vague on some things, like how many steps you need for a goal. things like that are going to be different in each game. for example, I'm thinking pretty much every goal in the teen romantic comedy game will have 10 steps of success; in a rotating GM Rune-like dungeoncrawl game using this system, each player may have a total number of steps to assign to elements in their encounter when it's their turn to design an encounter. if an easy encounter has a total of ten steps for the GM to "spend", that might be one full-power monster (10 steps to kill), or a 6-step monster and a 4-steps-to-disarm trap. most of the games, though, will just have 10 steps for major goals (cross the river) and 1 step minimum for player-generated goals (build a raft,) with the ability to "rollover" steps from player-generated goals like the raft into related major goals.

so my design goal with a step system like that would be to have ways of generating steps faster. under the dice system above, 1-die rolls have very slow accumulation of steps (and no chance of "bonus steps" from the die-matching aspect). 5-dice rolls will earn about 2 steps per roll up to a maximum of 3 steps, while 9-dice rolls will earn about 3 steps per roll with a maximum of 5 steps. this means that in a melee, it's better to be the trained warrior and roll 9 dice than it is to be the scholar and roll 5 dice.

the other part of this core mechanic is the talent system. you can use talents to add dice (making that 5-dice roll into a 9-dice roll.) if your roll is a success, you can use talents to add more steps; with enough talents, you can get all ten steps on one roll. and, when it's your opponent's turn, you can use your talents defensively to cancel individual dice.

that's the real trick here. if you play mechanically using this system, it works ok, but players will get better results when they describe how each of their talents come into play to afftect their goal. it's all a gimmick to get players to narrate their successes, combined with the innovation of tracking "hit points" for damage and other things in the play area using tokens, instead of on character sheets.

[Mike Holmes]

Mike: a 1-die roll is for a "nearly impossible" goal -- which shouldn't happen too often, but it's there to be used anyways, if needed. impossible goals -- and impossible-to-fail goals -- do not get a die roll, nor do unchallenged goals for which speed and quality do not matter.

I think you miss my point. I may have been unclear. What I'm wondering is if a 10% chance is dramatically commensurate with "nearly impossible". I don't want to go too sim on ya, but even in Sorcerer you can get less than 10% chance to win in some cases.

Mike

[anonymouse]

Pokethulhu uses the "number of dice = difficulty" method. Roll 3 dice for just kinda-tricky stuff, 2 dice for challenging, 1 die for stupid-hard. Roll under a TN of the appropriate stat (Grade Level for smarts-stuff, Phys Ed for athletics, et cetera).

It's super-simple and I love it, so if your mechanic is based around the same kind of thing, I'd say go for it.

[talysman]

oh, and although it's not important, here are some sample results from my perl script I wrote to test the probabilities using this system. the script rolls Xd10 10,000 times and counts up the number of times the total is below 9 (failure,) 9 or above with no matches (1 success,) and 9 or above with matches (2+ successes). it then calculates the % probability from these figures. the probabilities naturally drift a bit on each run of 10,000.

Code Select Expand


probabilities when rolling 5 d10, based on 10000  rolls:

      116 failures                (1.16 %)
     3108 rolls with 1 successes  (31.0 %)
     5688 rolls with 2 successes  (56.8 %)
     1088 rolls with 3 successes  (10.8 %)


Code Select Expand


probabilities when rolling 9 d10, based on 10000  rolls:

        2 failures                (0.02 %)
       29 rolls with 1 successes  (0.29 %)
     1445 rolls with 2 successes  (14.4 %)
     4942 rolls with 3 successes  (49.4 %)
     3248 rolls with 4 successes  (32.4 %)
      334 rolls with 5 successes  (3.34 %)


(not typical; most of the trials show 0 failures in 10,000 rolls of 9 dice... but I did it several times untill I got a failure in the batch, just to show that failure is possible on 9-dice rolls.)

one thing I am considering is changing the middle difficulty from a 5-dice roll to a 4-dice roll; it has the same maximum number of successes (3) but the probabilities are shifted more towards the lower number of successes, and failure is a weensy bit higher:

Code Select Expand


probabilities when rolling 4 d10, based on 10000  rolls:

      479 failures                (4.79 %)
     4806 rolls with 1 successes  (48.0 %)
     4471 rolls with 2 successes  (44.7 %)
      244 rolls with 3 successes  (2.44 %)


I'm not sure if it's worth it, though.

[talysman]

I think you miss my point. I may have been unclear. What I'm wondering is if a 10% chance is dramatically commensurate with "nearly impossible". I don't want to go too sim on ya, but even in Sorcerer you can get less than 10% chance to win in some cases.

nah, I caught your point: that the phrase "nearly impossible" seems to have too high a chance of success. like I said, I may change the wording to make it clear I'm being slightly Simmy, but from a functional angle instead of a realism angle. certain goals should be harder to achieve, but i'm being very generous with the chances of getting *something* accomplished.

anonymouse: yeah, Pokethulhu is a nice little system; I'm certainly aiming for something almost as simple, but with a little more depth of detail and a slightly more serious focus.

[Mike Holmes]

OK, cool, you're going for very dramatic, then; that's fine. But on the other end of the spectrum, now you're trying to say that a .02% chance of a failure means that it isn't "impossible". Well, the number is actually probably less than .01% overall, but let's say that it was. Let's say we played a campaign in which we rolled 1000 times using 9 dice. I think that's probably pretty generous seeing as only some small fraction of rolls will be 9 dice. In this scenario, you'd have less than a 10% chance of seeing even one failure. That's only a 60% chance or so of seeing a failure in ten full games played.

For practical purposes, that's "impossible". For a game that wants to make the "near impossible" roll mean 10%, it seems odd to have the "impossiblility" of failure on the other end be so different. Shouldn't there be something more like a 10% chance of failure?

Actually, this sounds like the Sorcerer odds.

Mike

[talysman]

yep, it's mainly for a very dramatic in a game; I assume in the games that use this system that characters are actually competent at what they do, so they are mainly rolling to see how quickly they can do it.

as for the difference between "impossible to fail" (no roll allowed) and "ordinary difficulty" (9 dice,) keep in mind the opponent will be using talents to affect the roll. I could roll 9 dice only to see an opponent eliminate my 6 highest die rolls with defensive use of talents. the focus on success and failure, then, really switches to how talents are used.

still, I can consider whether to up the "roll over" target to adjust the odds. I mainly selected 9 dice maximum, roll 9 or more because, on a 9 die roll, all 1s is a success, whereas several 1s and one zero is a failure. I'm trying to keep it easy to read. I'm perfectly willing to allow failure on 9 dice rolls to be very rare, with the main point of the roll being to determine whether you earned 1, 2, 3, or 4 steps of success.