SR3R Project Forum

Do you think it would be too much of a problem to allow a person to get extra successes from a single die for, say, every six or (if you want to be more prohibitive) nine or twelve points they go over the TN? Partly this is in response to the concerns about Thresholds; it would eliminate the problem that sometimes the threshold or the other guy's number of successes in an opposed test are greater than the total dice you can throw down, making the task utterly impossible. The potential problem is, well, that nothing would be impossible anymore.

Thoughts?

It undermines the philosophical underpinning of the success-based system. I'm -1 on this, but would like other opinions.

~J

_________________
Failure: when your best just isn't good enough.

While I am against multiple successes from a single die, it does bring up a concern about threshold that may be overlooked. I am sure we have all heard of the urban legend of a cocky GM, and a player who had 3 dice to roll 4 successes (TN 18 ) to diffuse a bomb. As the story unfolds, the GM is handing out new character sheets, when the individual rolls 3 18's, and burns a point of karma pool to buy the 4th success.

Threshold or not, I think its fair to say you should roll it out anyway. You never know what might happen

I'm not a huge fan of thresholds, either, but they're already in use and aren't as totally inconsistent with everything else as, say, open tests are.

~J

_________________
Failure: when your best just isn't good enough.

Which philosophical underpinning if this idea undermining? That is should be impossible to achieve great success (or, in some cases, any success at all) without a minimum level of skill? That there should be some instances where it is better for you to temporarily forget your training in a discipline and default to an attribute, because that's the only way you have a prayer of success? That a mage should never have to case Invisibility at higher than Force 1, because he can easily get the 4-6 successes needed to make his spell invincible?

I've seen you make this claim before, but I'm still unsure exactly why.

That your overall level of success is dependent on your total number of successes, and that exceptionally high single rolls give nothing in particular special. It's a philosophy that provides some stratification by skill level that I like—it's only when thresholds get involved that it breaks down.

~J

_________________
Failure: when your best just isn't good enough.

Thresholds, opposed rolls, "held" opposed rolls like the ones made in ranged combat tests and sustained spells, and "soft" thresholds like damage resistance tests. The only time it doesn't break down is when only one success really matters, and the rest are merely fluff.

(Edit): Now, I'll fully admit that allowing high rolls to count as extra successes significantly increases the difficulty of calculating the exact probability of success for a given number of dice, TN, and threshold, but given that we're talking about only allowing additional successes when you beat the TN by a very significant amount (I'd say minimum 6, but 8-10 might be better, maybe even 12 though that would basically ensure the rule is never invoked).

I've got a few questions, then. To start off with: what about negative TNs? If you get an extra success for every six-above, does that mean that when your TN is -4, every 2 and higher (every success, in other words) counts as two successes?

~J

_________________
Failure: when your best just isn't good enough.

Sure, why not? If a test is so easy that it's TN -4 (maybe more like TN -7 if extra successes need to beat the TN by 9, which is where I'm currently leaning), shouldn't you be able to achieve greater success more easily than if it were TN 2?

(Edit) Keep in mind we're looking at the case where the only reason to even bother rolling is to see if the Rule of Ones comes up; I have a hard time seeing the case where getting twice or even three or four times as many successes is going to actually mean anything.

I don't like extra mechanics, especially in an area I don't think there's currently a problem. Consider this (currently, nothing is ever set in stone in my mind) another vote against the idea. I'm perfectly ok with the fact that someone with only 1 or 2 dice doesn't have any chance at all in a resisted test against someone with enough skill to get 3 successes.

Such cases exist. Vehicle Dodge is one of them—when healthy, my Rigger's TN for dodging with her van is -3. Whereas she's already reliably able to rape physics by accelerating or decelerating 10*Acceleration, this would enable her to do it somewhere in the vicinity of 18*Acceleration.

~J

_________________
Failure: when your best just isn't good enough.

Yes, well I consider that a problem with the vehicle rules moreso than the proposal.

Other opinions?

~J

_________________
Failure: when your best just isn't good enough.

Me too. If I have 1 or 2 dice in, say, chess; no way am I going to beat a grandmaster.

In my games, I generally do allow this (I use an online diceroller, however, so it automatically rerolls 6's even if the TN is 2). In the case of extra-ordinary successes, like 12 above the TN, I generally add it as 1 success on the roll. I like it because it adds a tiny bit more danger and can recognize a little more the dice smiling upon you. However I recognize it's a house rule, one I generally don't even mention to my players. If I did table top, I'd never use it because that would require I reroll all my 6's on every test, which is time consuming (alright, I take that back, I have done that once or twice on cool tests, but not generally).

I'm neutral leaning towards no, with the realization that I'll continue doing it anyway.

Heh, and with that I guess the motion is defeated. If the people who actually use it can't support it, then the idea is doomed. Guess we'll have to figure on a new way to fix Open Tests, if indeed we need to bother (Sphynx and SirBedevere's argument can easily be applied there as well.)

We've used precisely this idea to great success in our games. It also has the advantage of solving the "6=7" problem, as long as you set the threshhold for the second success so that it isn't a multiple of 6 above the TN for a single success (we used 10, so a die roll of 1 to TN-1 = 0 successes; TN to TN+9=1 success; TN+10 to TN+19=2 successes; TN+20 to TN+29=3 successes, and so on). It's a good, solid mechanic, for several reasons:

1. The base principle of dice pool systems is "more dice equals higher quality of success". The base principle should never be "if you don't have a high enough dice pool, you're screwed" - any more than the principle of a straight 2D6 or 1D20-with-modifiers system should be "if you can't get enough modifiers for the TN to be below the max dice roll, you're screwed". Success (and failure) should always be possible; in some circumstances it should merely be incredibly unlikely.

In SR3 (and the old World of Darkness), the inherent strengths of a dice pool system are fully utilized: A dice pool mechanic allows task resolution to be two-dimensional. In most RPG's, task resolution is a single dimension: Your base skill/aptitude/whatever (X) is added to any difficulty or modifiers to get a final value of (X+N), which is either the number of dice you roll, or the modifier to the 2D6 or 3D6 or 2D10 or 1D20 or whatever roll. In SR3, you have your base skill/aptitude/whatever (X), which is then added to any skill or equipment bonuses to get the final dice pool (X+N). And then, on a completely different axis, you have the target number (Y), which represents how difficult the task is to complete, and any modifiers to that target number (Y+M). These are two seperate axes, which allows for a much richer resolution system. For one thing, it means that the number of successes you get on your dice roll can always be interpreted in a context-free manner; a 6-success roll means the same quality of success whether the TN was 3 or 24. For another, it means that a TN of 8 always means the same thing, no matter how high or low your skill.

I'd like to re-open this idea, especially given the fact that implementing it may well solve our 6==7 problem if we do it right, as well as helping deal with "impossible" corner cases when eliminating open tests. We already know that at least one person uses it in games without any real problems.

The opposition to the idea has a good point, though, that it will increase dice-rolling. Thing is, though, it's not like you are forced to go for extra successes; the only times they're really going to matter are the cinematic, exciting times where high rolls usually draw a cheer from your fellow players and I usually end up re-rolling them anyway (as does nexumi, it would seem).

This also solves the current problem that an Int 3 guy can never, ever positively ID anyone he sees, because you need 4 successes on a Perception Test to know something's exact nature, which I think is... very odd.

It would explain many things about society!

I'm still iffy about the idea, but why not give it a shot. What do you think is a good value for success accumulation?

~J

_________________
Failure: when your best just isn't good enough.

Well, I'm tempted to say +5, because it probides the greatest difference between TN6 and TN7. That is, if your TN is 6 you only have to roll an 11 to get two successes, but if your TN is 7 you have to roll a 12, which is twice as hard, relatively speaking. +5 seems a little aggressive at first blush, though.

I suspect it is. I'll see if I can't get us some hard numbers, but in the meantime feel free to bounce ideas around.

~J

_________________
Failure: when your best just isn't good enough.

Hm. Well, +11 would accomplish the same thing, but tell me honestly how often you see someone roll a 15 in a typical campaign? That's how high you'd have to roll to get a second success on a TN 4 with that much of a markup.

Keep in mind also that these markups will never make it easier to succeed at the original task; all they'll do is give someone an extra success or two when they roll above the TN. In any event we should be careful with how many negative TNs we toss out; the one that worries me the most is the Smartlink, armor's effects on damage soaking TNs, as well as some of the hacking TNs more than even the rigger stuff.

I chose +10 because it's the easiest to read. You see a TN of 8 and your average 7th grader can intuitively break that out into "8, 18, 28, 38, 48..." not so much with not-round-in-base-10 numbers.

That's the other reason I chose +5 instead of +11 as a starting point; adding 5 is pretty easy too, and it nets you that extra differential between 6s and 7s that I'm looking for.

Unfortunately, +5 results in way too many extra successes.

Maybe, maybe not. It certainly will increase the overall volatility of many Threshold-based tests, but there are relatively few of those that are actually hard-coded into SR mechanics. The most prominent of those I can think of right now are Perception tests, and I for one have no problem with increasing the probability there with having a higher probability of multiple successes. Keep in mind that this rule will never increase or decrease the probability of actually failing the test for any given TN; all it does is increase the expected number of successes and variance in number of successes once the first success is already assured.

The area we'd most likely see this rule having an effect is in opposed tests and similar tests, such as combat/damage resistance, where thanks to 'ware and armor the TNs are relatively low to start. Here we have an inherent balance due to the fact that the rule will be benefiting both sides equally. On the one hand this rule increases the expected number of successes per die roll, which favors the guy with the larger number of dice to roll. At the same time, it also favors the guy with fewer dice to roll, because it increases the variance in the number of successes per die, meaning for instance an Int 3 guard gains the ability to theoretically see through a Stealth 6 superninja, even if the possibility is vanishingly remote.

My initial opinion is that +5 success would be very dangerous to include in a vanilla SR3 campaign, but as we have the time to fully vet the rule here with actual playtesting and computer modeling it's worth looking into. This could change though as the numbers get crunching...

Speaking of which, Kagetenshi, would your probability calculator be able to handle a rule like this one?

It can't handle it right now, but what I meant by my "I'll see if I can't get us some hard numbers" above was that I'm intending on making it handle this as soon as I get the time (probably over the weekend).

~J

_________________
Failure: when your best just isn't good enough.

If we MUST include this idea, it's probably worth making the number related to the dice we roll, not the number of fingers that roll them. We use 6 sided dice, so a +6 or +12 is pretty easy to remember; just roll 6 twice more with that die!

I prefer a multiple of 6 myself, but that would still require a separate fix for 6==7 (or leave it unfixed).

~J

_________________
Failure: when your best just isn't good enough.

We could combine +9 -> extra success with option 3 in the 6=7 debate and make everything fairly neat.

Well, the problem with +6 or +12 is that it doesn't fix the 6==7 problem. Anything that isn't a multiple of 6 does fix that problem, due to the wonders of modulo arithmatic.

10 is good not because of fingers, per se, but because our minds are trained from an early age to work with adding/subtracting by 10's - so it's easier to work out the extra successes "on the fly", as it were.

For what it's worth, the idea of counting by tens in SR gives me the heebie-jeebies. It's not a dealbreaker, but it feels deeply unnatural.

~J

_________________
Failure: when your best just isn't good enough.

Well, +10 does seem to be the best solution, the more I think on it, simply because the option increases the difficulty by about 20-fold, on average. That is, on average it is 20 times as difficult--or 5% as likely--to hit a TN that is 10 higher than the original.

I'm not entirely sure why, but the number pleases me; maybe a little too much time playing d20?

I wasn't going to say it ;)

Probability calculator extension hopefully coming later today or tomorrow.

~J

_________________
Failure: when your best just isn't good enough.