I am currently reading Sorcery & Super Science! by Expeditious Retreat Press. It uses a mechanic called the Floating Dice System. In short it does the following:


    "Conflicts in Sorcery & Super Science are resolved using the floating dice
system. In this system there are no defined difficulties - the success or
failure of any action is based upon the relationship between the PC and the
task. Sorcery & Super Science uses a multitude of dice when rolling these
conflicts- 4, 6, 8, 10, 12, 16 and 20-sided. If you don't have a d16 handy, 2d8 will work in pinch.
    Most conflicts are ability conflicts: tasks that are successfully performed
based upon a primary ability of a PC measured against the primary ability
of an NPC or against the difficulty of an action. There are seven primary
abilities ranging in value from -4 to 20. A rank of 0 is considered the lowest
human rank and a rank of 8 is considered human maximum. The majority of
people fall within the 1 to 3 range. Such is considered average, with 2 being the most typical rank.
    In a primary ability conflict, the PC's ability rank is added to the NPC's
ability rank. This addition determines what die is rolled to resolve the conflict and determines the target number for success. For example a PC with a rank 4 ability is in conflict with an NPC who also has a rank 4 ability. Adding the rank values (4+4) results in an 8. Thus, 8 is the die rolled (a d8) and the target number for success is also 8. The player would then roll a d8 and add the PC's rank 4 ability to the roll and any result of 8 or better is a success or even a greater success.
    The terms greater success and success are used throughout Sorcery &
Super Science. When rolling conflicts, a result that is equal to the number
needed or 1 point greater is termed a success while a result that is 2 or more points higher than the highest number on the die is termed a greater success. A 10-11 is a success on a d10, while a 12 or above is a greater success. On a d8, a success is 8-9, and a greater success is a 10 or higher and so on. Dice determination is always rounded down if the sum of both
numbers is not equal to 4, 6, 8, 10, 12, 16, or 20. For example, a d12 is rolled for a conflict between a rank 8 and a rank 5, a d6 is rolled for a conflict between a rank 4 and a rank 3, a d16 is rolled for a conflict between a rank 9 and a rank 10 and so forth. The d4 is the lowest die that can be rolled and any totals resulting in less than 4 result in rolling a d4."

    Now, I have just read this, I have not played it. But only after a few lines I began to ask questions and make observations. In the primary example, "a PC with a rank 4 ability is in conflict with an NPC who also has a rank 4 ability. Adding the rank values (4+4) results in an 8. Thus, 8 is the die rolled (a d8) and the target number for success is also 8. The player would then roll a d8 and add the PC's rank 4 ability to the roll and any result of 8 or better is a success or even a greater success" the GM *has* to reveal to the player what level of skill his opponent has. By saying, 'use a d8' the GM is telling the player that his enemy is a skill rank 4 challenge. I don't like that. I don't want my players to have hard data on what skill level their opponents have. That should be revealed through actual play.

    My second observation is: Why? Why use this mechanic? What does it bring to the table? How is it superior to a static target number that must be bested by a die roll? One that would not require the GM to divulge information that they might not wish to reveal. Is it just to be different? My math skills are abysmal, does this mechanic bring anything different to how the statistic pan out? Like the difference between a d20 roll and a 3d6 roll for example.

    The point spread between each die type is 2, 2, 2, 2, 4 & 4. That seems odd to me.

    This just strikes me as being different for the sake of being different. I don't really see the point. And I really don't like that the GM has to tell the players a whole lot about the world around them whether they want to or not.

    Can anyone shed some light on this system for me? Am I just missing some key feature? Is there actually some elegant and beautiful bit of mechanical subtly that I am not grasping?

    Thanks in advance.

Sounds fiddly and confusing.

Also, 1d16 would be an average value of 8.5, 2d8 would be an average value of 9. Doesn't leave me hopeful for the rest of the system.

Quote from: Pseudoephedrine;450990Also, 1d16 would be an average value of 8.5, 2d8 would be an average value of 9. Doesn't leave me hopeful for the rest of the system.

I noticed that too. I think this is especially poor advice as a D16 can be simulated (without changing the odds) by rolling a D8 with any other die, if number the other die is even, you add 8 to the D8 result, otherwise you just take the D8 result. Suggesting rolling 2d8 instead just strikes me as weird.

I'm guessing that the objective is to increase the amount of randomness in contests where the ability scores are very high.

In a lot of RPGs at the higher end of the scale you get a very small proportional attribute difference giving guaranteed success to one side. For example,  you could have a wrestling contest between giants with Strengths of 100 and 120, that the 120 will virtually always win even though the relative difference in the system might not be that that much - no more than between two humans with Strengths of 10 and 12.  (Unless the Strength table is logarithmic like 3E D&D's, of course; but this is a problem in GURPS that actually has a specific patch rule, and also is applicable to say Tunnels & Trolls, or 4E)

Here the two scores being higher -both your own and and the opponents - means a larger dice is rolled so the bonuses are becoming less significant.

Its quite interesting since I hadn't seen the problem tackled this way before - there are a couple of systems that use tables or whatnot to do this (GURPS has a specific rule for superstrong giants...Forgotten Futures or JAGS damage rolls makes opponents divide the relative scores until both their values fit on the same table)...but I haven't seen it built into the core mechanic before.

Looks like for no other reason than to be gimmicky.

RPGPundit

Quote from: Pseudoephedrine;450990Sounds fiddly and confusing.(...)

To say the least. When I made an attempt at reading the explanation I was like: what the fuck! SKIP

Such gimmickry for its own sake has been rather a fad for a while. It probably goes back further, but I recall finding it a turnoff in the DC Universe game.

Risus, by S. John Ross, is hardly anything but dice gimmickry! It's as if someone had published the attack, saving throw and undead-turning tables from D&D Volume 1 as a complete RPG -- except that would have taken only a single sheet of paper.

I've seen such a fetish made of needlessly complicated "universal resolution systems" over at RPGNet that people routinely post such contraptions as apparently the most interesting feature of a proposed (or published) new game.

Ken St. Andre, from what I gather, did T&T with "dice pool" combat and "doubles add and roll over" saving rolls because 20-sided dice were uncommon.

I gather that he also likes to "fudge" things a lot, so maybe the difficulty of the maths to figure out what probabilities he was setting up was even a positive advantage (since the players would typically be even more in the dark, if that were possible).

Wow, seems weird. Did I miss something - or what happens when the rank+rank is an odd number, say 5? Do you roll a d6 or a d4?

This system produces strange scaling, it seems to me at first glance. Just thinking out loud, but, assuming I'm reading the system right: say your opponent is 4 times better than you by a rank of 8 to a rank of 2. Then you roll a d10 for resolution, and you succeed 30% of the time (8-10). But then let's say your opponent is 5 times better than you by a rank of 5 to a rank of 1. Then you roll a d6 and you succeed 33% of the time (5-6).

How does that make sense? In case B, you're less skilled, and your opponent outmatches you by a 20% greater margin, and yet your chances for success are increased.

Doesn't pass the smell test.

This awkward "scaling by running up the dice" thing is wierd. I don't guess I mind much when there's a single fixed target number. You have that with the Window, or Savage Worlds, or some others. But there's a nice little generic game called HDL--Half Dice Level--that works like that. It's ok, but it suffers from the statistical anamolies already mentioned, and it seems too fiddly to get many players interested.

It really kills me in the new editions of Metascape. When the game started, you scaled all your rolls by using 6s, 8s, or 10s with multipliers. That was fine. You roled a multiplyind die along with your regular die, so it all seemed nice and coherent. But now they do this dice scaler, using every kind of die I've heard of, in weird combinations, and keeping the multipliers, anyway. That's just uncalled for. I can't get anyone to play the free download editions. The cleaner first edition still has some traction.

I don' mind gimmicky dice, but the gimmick has to be pretty simple and straightforward. Stacking gimmicks drives people nuts.

The math on scaling dice has to be done carefully, or yeah, you get some weirdness.  One of the ones that bugged me was Earthdawn, which used single dice, but to get over the d12-d20 game they used combinations of dice instead, which gave certain parts of the scale a curve and others flat probability.  It's easy to fix too, you just use two dice for the whole scale, I did the math on the conversion once.  

It's always weird to me when supposedly professional game designers pump out games where it's clear they didn't think about the math for two seconds to even make sure the results were consistent.  

Game design isn't that hard, but RPGs in particular seem to lull people into thinking you can just slap any old crap together without sanity checking it and it'll be A-OK.  From a random nerd selling his homebrew on Lulu for $2 above cost, that's one thing, but you see this shit from major publishers and guys charging full cover price, and it's just pathetic.

Yep. Another thing that's somewhat along the lines of obvious math-scaling errors is obvious geographical-scaling errors.

Now, truth be told, I'm pretty much an easy touch when it comes to RPGs. Bad rpgs are like bad pizza or bad sex for me; even when they're bad they're pretty good. Math stuff along the lines of Savage Worlds's hiccups are no problem to me, really, and d12 to d20 actually makes as much sense as d4 to d6.

But geography-scaling nonsense really irks me. Like when a game comes out with a map where the continent is supposedly settled and civilized at a medieval level, but the average population density turns out to be 1*10^-18 people/square km or something. Or when the largest city and supposed giant center of civilization to rival Rome/Constantinople/Paris/etc. has a population of 8,000.

Quote from: YcoreRixle;451152Yep. Another thing that's somewhat along the lines of obvious math-scaling errors is obvious geographical-scaling errors.
...

But geography-scaling nonsense really irks me. Like when a game comes out with a map where the continent is supposedly settled and civilized at a medieval level, but the average population density turns out to be 1*10^-18 people/square km or something. Or when the largest city and supposed giant center of civilization to rival Rome/Constantinople/Paris/etc. has a population of 8,000.

Demography, not geography, I think you mean. And yes, it bothers me as well.

This was a minor but annoying oversight in Eberron, where Khorvaire's population should have been increased by a factor of 10 (to approximately 150 million humanoids), especially since they count goblinoids, orcs and lizardmen amongst those numbers.

Quote from: YcoreRixle;451134Wow, seems weird. Did I miss something - or what happens when the rank+rank is an odd number, say 5? Do you roll a d6 or a d4?

This system produces strange scaling, it seems to me at first glance. Just thinking out loud, but, assuming I'm reading the system right: say your opponent is 4 times better than you by a rank of 8 to a rank of 2. Then you roll a d10 for resolution, and you succeed 30% of the time (8-10). But then let's say your opponent is 5 times better than you by a rank of 5 to a rank of 1. Then you roll a d6 and you succeed 33% of the time (5-6).

How does that make sense? In case B, you're less skilled, and your opponent outmatches you by a 20% greater margin, and yet your chances for success are increased.

Doesn't pass the smell test.

Yeah, that's odd.  If they want to go that way, should make it opposed.  That way Rank 2 vs. Rank 8 means they roll d10+2 and d10+8 respectively (possible but unlikely for Rank 2 to win), where 1 vs. 5 would be d6+1 vs. d6+5 (absolute best Rank 1 could get would be a tie).  That at least appears to be scaling properly, but there might be holes.

Quote from: YcoreRixle;451134Wow, seems weird. Did I miss something - or what happens when the rank+rank is an odd number, say 5? Do you roll a d6 or a d4?

This system produces strange scaling, it seems to me at first glance. Just thinking out loud, but, assuming I'm reading the system right: say your opponent is 4 times better than you by a rank of 8 to a rank of 2. Then you roll a d10 for resolution, and you succeed 30% of the time (8-10). But then let's say your opponent is 5 times better than you by a rank of 5 to a rank of 1. Then you roll a d6 and you succeed 33% of the time (5-6).

How does that make sense? In case B, you're less skilled, and your opponent outmatches you by a 20% greater margin, and yet your chances for success are increased.

Doesn't pass the smell test.

I think this is a misreading of what they're doing.
You don't need to beat the opponent's rating; the target is just equal to the highest number you need on the dice i.e.
d8...you need an 8+
d10....you need a 10+.

The dice mechanic (aside from rounding error) would be a proportional system.
If you have stat X and your opponent has stat Y, your success chance is X out of [X+Y].  
(actually X+1 out of [X+Y] since you could roll the target even with a +0).

Quote from: YcoreRixle;451134Wow, seems weird. Did I miss something - or what happens when the rank+rank is an odd number, say 5? Do you roll a d6 or a d4?

This system produces strange scaling, it seems to me at first glance. Just thinking out loud, but, assuming I'm reading the system right: say your opponent is 4 times better than you by a rank of 8 to a rank of 2. Then you roll a d10 for resolution, and you succeed 30% of the time (8-10). But then let's say your opponent is 5 times better than you by a rank of 5 to a rank of 1. Then you roll a d6 and you succeed 33% of the time (5-6).

How does that make sense? In case B, you're less skilled, and your opponent outmatches you by a 20% greater margin, and yet your chances for success are increased.

Doesn't pass the smell test.

Yeah there's a slight problem with their math.
I think they're aiming to give character with stat X vs. guy with stat Y odds of X out of [X+Y], but because you just need to roll the dice's maximum to hit the target number, the odds are actually [X+1] out of [X+Y].

They need to bump all their target numbers up by 1 to get it to work, I think, though there's still possibly a problem with rounding error.