PDA

View Full Version : Stupid Idea #3 - D8 Chance Mechanic

FreakyCheeseMan
2013-04-29, 11:37 AM
Working on building a game from the ground up, and, in addition to everything else, I've come up with a very strange chance mechanic that I wanted to see what people thought of.

The System
Players will use a d8 to determine the number of successes or failures they get with any action they take; the number of successes is not equal to the result of the die. Rather, it follows the following chart:

{table] Number | Result
1 | 3 Failures, and roll again.
2 | 2 Failures
3 | 1 Failure
4 | 1 Failure
5 | 1 Success
6 | 1 Success
7 | 2 Successes
8 | 3 Successes, and roll again. [/table]

EDIT FOR CLARITY: When you roll the first time, you take the result you get on that chart; if and only if that result is a 1 or an 8, you roll again and do one of three things:

If the result of your re-roll is the same as your initial result, you add another 3 successes or failures, and roll again. (So, if you roll an 8 the first time, and another 8 the second time, you roll again, until you get something that is not an 8.)

If the result of your re-roll is the same type (successes or failures) as your initial result (but not the exact same number), you add that many successes or failures and stop rolling. (So if you roll an 8 the first time, and an 8 the second time, but a 5 the third time, you get a total result of 3 + 3 + 1 = 7 Successes.)

If the result of your re-roll is the opposite type (successes or failures), then you just stop rolling - you do *not* subtract anything from the results you've got so far. (So, if you roll an 8, and then a 1, that's still just 3 successes.)

After rolling, bonuses or penalties are applied directly - you add or subtract that many successes or failures. 1 Success +1 = 2 Successes; 1 Success - 1 = 1 Failure (there is no "neutral" result.) In the case of attacks, your opponent's defences - the equivalents of AC, SR and Saves - are just treated as a penalty on your attack.

Specific actions will either have a threshold they have to hit (in the case of, say, a Trip attempt or a Blindness spell), or a varying range of results by number of successes (in the case of... well, anything that would normally have a damage roll.) So, some weapons (say a quarterstaff) might hit more reliably (do damage with a lower number of successes) but never inflict much damage (not scale well as you get more.) Other weapons - let's say a scythe - might do no damage at all until you get three or four successes, but then quickly pile on damage with additional successes - representing a weapon that's hard to hit with, but inflicts a lot of pain when you do.

Justification - Gameplay
First off - yes, I know the idea looks horribly complicated and unplayable. Here's why I think it isn't:

First, this is the only chance system in the game - players will never use anything else, so they should be able to get comfortable with this system quickly, just through repetition. Additionally, there is almost no math to be done - that vast majority of the time, even after bonuses, values will fall between +5/-5, so there's no arithmetic to slow things down.

My main goal with this is to minimize the amount of dice work required; keeping everything under the same broad hood; essentially, the chance mechanism is isolated to this single, consistent model (which hopefully does everything needed), so the rest of the focus can be on more interesting aspects of play.

The mechanism itself is intended to give a smooth, consistent probability curve to actions - if you look, every result is exactly half as likely as the result one below it. (You're half as likely to get 4 successes as 3, etc.) This results in a very consistent and easy-to-work with model; a +1 bonus always means the same thing, and concepts like damage reduction, miss chance, critical hits and the like all fold neatly together.

This system should also let things be kept small and significant - even a +1 bonus is a very real advantage, and a +3 bonus is huge (I don't imagine anyone, soley as a result of their character design, ever getting a single bonus above +5, and even that would be rare; most of the time, higher bonuses will come situationally in combat through things like flanking, stealth attacks, etc.)

Additionally, the "Successes" model makes it easier to include interesting effects with enough successes or failures - for example, hitting someone with a war hammer with enough successes might knock them down; another character might have a feat that lets them make a free counterattack, if someone misses them on a melee attack with some number of failures. Spells may have "Bonus Effects" with enough successes - limited AoE damage, stun effects, or even expending fewer resources.

Finally, this sort of graduated system - where things like attack rolls aren't just a yes-or-no proposition - means it's worthwhile investing in things like AC, even if you can't pump them all the way. In D&D, after a certain level, another point of AC or SR is either crucial or useless - by that point either it's one of the core focuses of your character, or so low that it's never going to make a difference anyway. In this system, AC and saves are useful even if they're low.

So. How stupid is this one, on a scale from 1 to 10?

00dlez
2013-04-29, 01:46 PM
I give it a 3.

There just isn't enough variability in the results - penalties and modifiers of just +1 or -1 can swing the odds of success by a great deal and it doesn't leave a room to differentiate between the low skilled actors in the world and the truly exceptional. Without a few specific examples to pick through and dicuss, it's hard to fuly explain my point (or understand yours).

FreakyCheeseMan
2013-04-29, 02:01 PM
*Nods* It does lack the sort of gradient that 3.5 has, but that's actually intentional - I'm trying to keep the math to a minimum and give players as little to keep track of as possible (none of the "You have to add six different things together to get the relevant number" from 3.5), and that means having small numbers, and that means having numbers stay relevant.

Oh, another aspect of this system is that I'm trying to avoid "Leveling Up" in the way it works in 3.5, where the main benefit of leveling up is that your numbers get bigger. Instead, I'm trying to focus on leveling up as giving you more options - so, in a fight between a level 1 character and a level 20 character, the level 1 might win with good luck and tactics, if they stick to doing the same actions - but the level 20 will have all kinds of other options to shift the fight to his advantage.

OzymandiasX
2013-04-29, 02:09 PM
Successive dice rolls (as opposed to concurrent) slow down a game quite a bit. I'd be wary about making it part of a core game mechanic. It can be fun on occasion to go on a big winning (or losing streak), but eventually it becomes tedious.

The mechanism itself is intended to give a smooth, consistent probability curve to actions - if you look, every result is exactly half as likely as the result one below it. (You're half as likely to get 4 successes as 3, etc.) This results in a very consistent and easy-to-work with modelYour math is off a tad. You're half as likely to have 4 positive dice rolls (not successes) as you are to get 3, etc. And since each of those rolls can be up to 3, I think you're going to have a lot more instances of 5 or more successes/failures than you're thinking...

Finally, this sort of graduated system - where things like attack rolls aren't just a yes-or-no propositionUnless there are different results for different levels of success (like of you do +1d6 for every extra success) this IS a yes or no system, it just takes multiple rolls.

In D&D, after a certain level, another point of AC or SR is either crucial or useless - by that point either it's one of the core focuses of your character, or so low that it's never going to make a difference anyway. In this system, AC and saves are useful even if they're low. Yup. d20 games scale poorly after about 10 and even worse after 20. (This is why E6 is awesome)

Overall it isn't stupid at all. I like where you're trying to go with it. First thing I'd recommend is to playtest a large combat or two. It will give you a feeling for how long each action will take to roll. Also, jot down all of your roll results into a spreadsheet. I think you'll be surprised by how many rolls of 5+ success/failures you'll see.

FreakyCheeseMan
2013-04-29, 02:59 PM
Successive dice rolls (as opposed to concurrent) slow down a game quite a bit. I'd be wary about making it part of a core game mechanic. It can be fun on occasion to go on a big winning (or losing streak), but eventually it becomes tedious.

I figure a few things on that. First, this incorporates both damage and attack rolls, so you should actually be rolling less in total (you have a 1/4 chance of having to roll twice; a 1/32 chance of having to roll more than twice.)

Your math is off a tad. You're half as likely to have 4 positive dice rolls (not successes) as you are to get 3, etc. And since each of those rolls can be up to 3, I think you're going to have a lot more instances of 5 or more successes/failures than you're thinking...

Hmm... I think my math was right. Checking...

{table] Result | Probability | Explanation
1 Success | 1/4 | 5 or 6 on your first roll
2 Successes | 1/8 | 7 exactly on your first roll
3 Successes | 1/16 | 8 on your first roll, followed by 1 - 4 on your second (1/8 * 4/8 = 1/16)
4 Successes | 1/32 | 8 on your first roll, followed by 5 or 6 on your second ( 1/8 * 2/8 = 1/32)
5 Successes | 1/64 | 8 on your first roll, followed by 7 on your second
6 Successes | 1/128 | 8 on your first roll, 8 on your second, 1-4 on your third.
[/table]
My original idea was to just use a coin, and keep flipping until you get a different result than your first, then take however long a chain you got as that many successes or failures - this setup is intended to produce the same curve, but with fewer rolls.

Unless there are different results for different levels of success (like of you do +1d6 for every extra success) this IS a yes or no system, it just takes multiple rolls.
That's almost exactly what would happen, except for the 1d6 part - a standard longsword might look something like this:
{table] Chance Result | Result
3 or more failures | no damage, provoke AoO
1-2 Failures | no effect
1 Success | 1 Damage
2 Successes | 2 Damage
3 -4 Successes | 3 Damage
5+ Successes | 3 Damage, + 1 Armor-Piercing Damage [/table]

I think you'll be surprised by how many rolls of 5+ success/failures you'll see.

Eh, such should be easy to calculate - there's a 1/32 chance of getting (at least) 5 successes. So, they should be substantially rarer than critical hits are normally.

I am considering modifying it a bit, to include the possibility of neutral results.

OzymandiasX
2013-04-29, 03:44 PM
I figure a few things on that. First, this incorporates both damage and attack rolls, so you should actually be rolling less in total (you have a 1/4 chance of having to roll twice; a 1/32 chance of having to roll more than twice.)hehe. Unless you explained your system wrong earlier, you've got this way wrong.

First you make a roll. No matter what you roll, you will roll again.
Then you make a second roll. You will always make this roll to see if you 'keep going' or not.
Based on your 2nd roll, you have a 50% chance of needing to make a 3rd roll (if it is the same fail/success as the first roll was)

So...
You will have to roll a d8 one time: 100% of your attacks
You will have to roll a d8 a second time: 100% of your attacks
You will have to roll a d8 a third time: 50% of your attacks
You will have to roll a d8 a fourth time 25% of your attacks

Hmm... I think my math was right. Checking...

{table] Result | Probability | Explanation
1 Success | 1/4 | 5 or 6 on your first roll
2 Successes | 1/8 | 7 exactly on your first roll
3 Successes | 1/16 | 8 on your first roll, followed by 1 - 4 on your second (1/8 * 4/8 = 1/16)
4 Successes | 1/32 | 8 on your first roll, followed by 5 or 6 on your second ( 1/8 * 2/8 = 1/32)
5 Successes | 1/64 | 8 on your first roll, followed by 7 on your second
6 Successes | 1/128 | 8 on your first roll, 8 on your second, 1-4 on your third.
[/table]
heh. Your math is off here (possibly because of the math being off earlier). And, on top of that, you're leaving out all kinds of possibilities. For example you can get 4 successes total by rolling (5,5,5,5),(5,5,5,6),(5,5,6,5),(5,5,6,6),(5,6,5,5), (5,6,5,6),(5,6,6,5),(5,6,6,6),(6,5,5,5),(6,5,5,6), (6,5,6,5),(6,5,6,6),(6,6,5,5),(6,6,5,6),(6,6,6,6), (7,5,5),(7,5,6),(7,6,5),(7,6,6),(5,7,5),(5,7,6),(6 ,7,5),(6,7,6),(5,5,7),(5,6,7),(6,5,7),(6,6,7),(8,5 ),(8,6),(5,8),(6,8).

And that is just the permutations to get a total of 4 successes. :)

FreakyCheeseMan
2013-04-29, 03:53 PM
hehe. Unless you explained your system wrong earlier, you've got this way wrong.

First you make a roll. No matter what you roll, you will roll again.
Then you make a second roll. You will always make this roll to see if you 'keep going' or not.
Based on your 2nd roll, you have a 50% chance of needing to make a 3rd roll (if it is the same fail/success as the first roll was)

...where did I explain it that way? Look at the chart in the OP - only 1 or 8 have "Roll again" on them.

EDIT: I think I see where the confusion came from: when I said

When you roll again, you use the same chart, but if you get a different type of result than you started with (successes or failures), you just stop rolling and keep what you already have. So, if you roll an 8 followed by a 1, you just treat that as 3 successes; if you roll two eights in a row, and then a 6, you treat that as seven successes.

I think you took the "you just stop rolling" to mean "That's the only time you stop rolling" - it isn't. That whole paragraph was just clarifying that you can't go back down (or up) after getting a 1 or an 8 - so, if you get an 8 (3 successes) followed by a 2( normally, 2 failures) it doesn't cancel out to 1 success, it's just treated as 3 successes (and you stop rolling.)

The only times you roll again are when you get a 1 or an 8, and it's either your first roll, or the result is the same as your first roll. So, you have a 1/4 chance of re-rolling after your first roll, and a 1/8 chance after that.

00dlez
2013-04-29, 05:01 PM
*Nods* It does lack the sort of gradient that 3.5 has, but that's actually intentional - I'm trying to keep the math to a minimum and give players as little to keep track of as possible (none of the "You have to add six different things together to get the relevant number" from 3.5), and that means having small numbers, and that means having numbers stay relevant.
You might look at WW systems which just use more dice and operate on the success/failure mechanic with very little modifiers/math

Oh, another aspect of this system is that I'm trying to avoid "Leveling Up" in the way it works in 3.5, where the main benefit of leveling up is that your numbers get bigger. Instead, I'm trying to focus on leveling up as giving you more options - so, in a fight between a level 1 character and a level 20 character, the level 1 might win with good luck and tactics, if they stick to doing the same actions - but the level 20 will have all kinds of other options to shift the fight to his advantage.
That... I just don't like, but thats a personal thing rather than mechanics.

FreakyCheeseMan
2013-04-29, 05:07 PM
That... I just don't like, but thats a personal thing rather than mechanics.

Huh - what don't you like about it?

OzymandiasX
2013-04-29, 05:30 PM
I think you took the "you just stop rolling" to mean "That's the only time you stop rolling" - it isn't. That whole paragraph was just clarifying that you can't go back down (or up) after getting a 1 or an 8 - so, if you get an 8 (3 successes) followed by a 2( normally, 2 failures) it doesn't cancel out to 1 success, it's just treated as 3 successes (and you stop rolling.)
Ah. That IS what I read it as. Oops. Scratch that stuff about rolling consecutive times. :)

00dlez
2013-04-30, 02:00 PM
Oh, another aspect of this system is that I'm trying to avoid "Leveling Up" in the way it works in 3.5, where the main benefit of leveling up is that your numbers get bigger. Instead, I'm trying to focus on leveling up as giving you more options - so, in a fight between a level 1 character and a level 20 character, the level 1 might win with good luck and tactics, if they stick to doing the same actions - but the level 20 will have all kinds of other options to shift the fight to his advantage.

Huh - what don't you like about it?

Well, when I think of the time and effort it takes to get from level 1 to 20, not only would I expect that the level 20 character has far more options at his disposal, but that he would be so much better than the level 1 character at even basic tasks that there is less than a remote chance that they could actually best them at a given activity they both specialize in. IMHO, the level 20 shouldnt be "shifting the fight to his advantage", it should be a true miracle that the level 1 even comes close to posing a challenge.

FreakyCheeseMan
2013-04-30, 02:38 PM
Well, when I think of the time and effort it takes to get from level 1 to 20, not only would I expect that the level 20 character has far more options at his disposal, but that he would be so much better than the level 1 character at even basic tasks that there is less than a remote chance that they could actually best them at a given activity they both specialize in. IMHO, the level 20 shouldnt be "shifting the fight to his advantage", it should be a true miracle that the level 1 even comes close to posing a challenge.

So, the issues I have with that are as follows:

First, it means that things don't really get more interesting as you level - if at level 1 you've got a +5 attack vs. things with an AC of 15, and at level 20 you've got a +40 attack vs. things with an AC of 50... nothing's really changed.

Second, it vastly limits the kinds of fights you can have at any level - only about 1/4 of the monsters designed are even an option for you to fight, as the rest are either too easy to be interesting or too difficult to be possible. With this system, fights can still be harder or easier, but, hopefully, every monster and enemy will at least have some potential to be fun at every level.

Third, there are some things that you want to be challenges regardless of level, but that simply wouldn't make sense in the context of a game - like, say, city guards. You don't want to guards to be so much more powerful than the players at level 1 that it's ridiculous that the players are the ones saving the town; you also may not want those same guards to be so weak that the players could conquer said towns single-handedly at level 10.

Finally, there's the issue of things that quite simply should be threats at any level - falling off a cliff, getting trapped in a burning building, etc. With the 3.5 style of progression, either these things are an utter death sentence at level 1, or they're completely meaningless at level 20, or both.

So, really what I want is for higher-level characters to be objectively more badass, but at the same time to remain essentially human.

00dlez
2013-04-30, 05:03 PM
I just feel like we disagree on how we like our games to "feel", theres nothing mechanically wrong with either of our lines of thought. Just assume I preface/end my responses below with "IMHO"

First, it means that things don't really get more interesting as you level - if at level 1 you've got a +5 attack vs. things with an AC of 15, and at level 20 you've got a +40 attack vs. things with an AC of 50... nothing's really changed.

Except that the +40 to attack will always hit AC 15, and a +5 will never hit AC 50. The "at-level" challenges don't really change, but I don't think they should change either.

Second, it vastly limits the kinds of fights you can have at any level - only about 1/4 of the monsters designed are even an option for you to fight, as the rest are either too easy to be interesting or too difficult to be possible. With this system, fights can still be harder or easier, but, hopefully, every monster and enemy will at least have some potential to be fun at every level.

That's just fluff - you can bolster or weaken creatures to pose a threat at most any level, just a matter of DMing the way you like, not a matter best handled by mechanics.

Third, there are some things that you want to be challenges regardless of level, but that simply wouldn't make sense in the context of a game - like, say, city guards. You don't want to guards to be so much more powerful than the players at level 1 that it's ridiculous that the players are the ones saving the town; you also may not want those same guards to be so weak that the players could conquer said towns single-handedly at level 10.

Fluff again - if you want to scale city guards, thats fine... but I never will

Finally, there's the issue of things that quite simply should be threats at any level - falling off a cliff, getting trapped in a burning building, etc. With the 3.5 style of progression, either these things are an utter death sentence at level 1, or they're completely meaningless at level 20, or both.

I agree, but feel that these issues can often be delt with through house rules (falling over 100ft. = death), DM fluff, etc - and don't have much to do with the core dice rolling mechanic

Foxwarrior
2013-04-30, 06:56 PM
Have you considered the part where uncapped rolls mean that anyone can succeed at anything with enough tries, FreakyCheeseMan?

FreakyCheeseMan
2013-04-30, 07:34 PM
Have you considered the part where uncapped rolls mean that anyone can succeed at anything with enough tries, FreakyCheeseMan?

Eh... with attack rolls, that actually makes sense to me - sooner or later, even a beginner will get lucky.

With casting checks, that's actually intentional - it makes sense to me that people would be able to work some spells outside of combat, with enough time to focus, that they couldn't really pull off in a fight.

With skill checks... I'm still deciding how skill checks work, but they may not include as much rolling. ("Climb" skill may simply give you a climb speed, for instance.)

Other checks (that you would roll for) may not be re-rollable; in the same way that you can't re-try a failed Knowledge roll, a failed, say, Open Lock roll may indicate not that you were unlucky this one time, but that you're not familiar with this type of lock, etc.

So, I really don't see any instances where that would be a problem, unless you have some you'd like to point out.

OzymandiasX
2013-05-01, 09:06 AM
First, it means that things don't really get more interesting as you level - if at level 1 you've got a +5 attack vs. things with an AC of 15, and at level 20 you've got a +40 attack vs. things with an AC of 50... nothing's really changed.

Second, it vastly limits the kinds of fights you can have at any level[...]

Third, there are some things that you want to be challenges regardless of level, but that simply wouldn't make sense in the context of a game - like, say, city guards.[...]

Finally, there's the issue of things that quite simply should be threats at any level - falling off a cliff, getting trapped in a burning building, etc.[...]

So, really what I want is for higher-level characters to be objectively more badass, but at the same time to remain essentially human.
You've just described the things I like least about the way d20 (D&D 3.5 especially) scales! If you haven't already, check out the E6 variant. It fixes all of these problems very well, imo. It allows characters to become much more badass than a typical guard, for example, but it keeps them from becoming superhuman... which allows for a much more realistic 'heroic' feel to games.