Yep, and I think the certainty-at-cost concept is a good way to model this with the SR4 dice mechanic.

Another thing I'd also like to see the "trading dice for successes" category is a way to not only model "take 10" by spending five dice for a success, but a way to model "take 20". Maybe you can spend 10 times the required base time to get a better trade-in on your dice, say three dice per success? You'd need to look at the probabilities and figure out how many times you would have to roll to statistically get one-for-three successes and double that time, but it would be a good addition to the rules, IMO at least.

Oh, and personally I think it should be specified that all of these methods for buying success only work in non-stressful situations. When you're in combat you should have to roll, regardless.

Well, if you really want to do that, you ask someone to build you a probability table, and then figure out how long it will take you to roll enough successes, and use the average value. I'm not sure it needs to be in the core system, though; probability tables tend to scare people away. (See what happened when I posted them on Gunner J's "always a chance" thread!)

Well yeah that might be a little *too* in-depth, but having maybe just one extra level, to accomodate those who want to take 10 times as long and be fairly assured of a decent amount of success isn't too out of line?

Perhaps someone could figure out a way to do that that's fair at all levels. I probably could, but it would take quite a while to calculate all the relevant probability distributions.

But buying successes seems unecessarily complex, especially when you start factoring in situational modifiers to your dice pool. You'll wind up with players wasting time trying to set up their characters to have the most automatic successes prior to taking the slightest action.

Likewise, if you create a sliding 'trivial action' scale, every munchkin in Oz will make it their life's goal to abuse it. I say the definition of a trivial action should be a fixed number, e.g., 1 or 2 successes. If you then set some game mechanic that drops the task's successes below this threshold (using a Savior med kit to drop your First Aid success threshold down by 2, making M damage wounds trivial, for example), sobeit. In my opinion, this would be easier to manage than an open-ended 'take att + skill + modifier dice and divide by 5 to determine your trivial threshold for this roll' approach.

can someone explain something to me? So attributes add directly to skills if I'm understanding properly. So quick3 and pistols 3 = six dice to roll while firing.

So...how will they address such factors as unarmed combat? It's conceivable that you could be very skilled in unarmed combat with a low str and a high quickness. Do anyone see what I'm getting at? Skills linked to attributes could go more than one way. I'm not entirely sober atm so I'll try to clarify later.

probably by raising the minimum number of successes required to get a hit in.

What's munchable about a flat 6 dice per guaranteed success? If you're going to buy three successes, you'll be rolling 18 dice--and with those same 18 dice, you'd probably get 6 successes, and would get 9 about as often as you'd get 3. Unless you're doing something pretty easy and yet very important, why would you want to do that? Even with only 12 dice, you still have about a 95% chance of getting at least two successes. It's a really lousy deal to trade, unless failure is really bad news, or unless more successes are really unimportant.

I can imagine abuses where people are doing something millions of times without fail, but that's where advanced rules regarding fatigue could come in if necessary. For normal running this would never come up, and if you can munchkinize your dice to get a 6-for-1 trade, you'd still usually be better off actually rolling them all.

i think you guys are stuck on the horror that was auto-successes in SR1. in that system, you could be guaranteed a certain number of successes against TN 100. in this system, auto-successes are much less of a problem.

As long as it is only usable in non-stressful situations, I vote Yes to automatic successes.

I still say it's overly complex and provides little benefit while leaving room for abuse. If they're really looking at simplifying the rules, I can't see them going with something like that. I guess I'll just need to see it used in a real game before I'm convinced.

You can say whatever you want, of course, but can you actually provide an example of an abuse? Any rule might be abusable in some unknown situation, so let's try to actually think of abuses before we throw away rules with abuse as the justification!

Also, some mechanism to allow an average person to reliably tie their shoes is highly desirable. Few things are quite as jarring as a supposed world-expert failing to perform what ought to be a routine task, and if they're not rolling lots of dice to keep them from failing, and there isn't some special mechanic to do it, then fail they will.

For example, how often do you fail to make a post to this forum and lose the post into the ether instead? I haven't failed yet....

yeah. i have a hard time seeing how guaranteeing a single success, in a situation where the law of averages say you should be getting two successes, is horribly broken.

edit: ell, that happens a lot more often than you'd think. you may want to pick a different example. though, to be fair, i've never lost a post because i hit the wrong button.

Twice, and both because of power outages, but then I'm horrifically unlucky.

Additionally, I seriously doubt that any situational modifiers will affect the number of dice you roll at a given test. Those will more than likely affect the Threshold, with the mods to dice pool being related to available tools and such.

Assuming your dice pool will remain largely under your own control, it makes feasable the ability to withhold dice for various reasons, only one of which would be gauronteed success. For instance, a spellcaster could withhold dice to make his casting less noticable, or maybe to increase/decrease the area of an area effect spell. Melee attackers might withhold dice to increase the Threshold to hit them in melee (fighting defensively). Ranged combatants might withhold dice to make a Called Shot. Maybe someone doing B/R can withhold dice from their test to reduce the base construction time, essentially hurrying the job along deliberately rather than working faster as a byproduct of doing the job well, as is currently the case in SR3?

The ability to withhold dice from a skill test in order to achieve optional effects seems a very versatile and powerful mechanic, one that could be much greater in scope than merely withholding dice for gauronteed successes (though that is one good optional effect to include).

i don't see that they've actually simplified anything. It seems like the exact same thing to me except it's a TN 5 base instead of variable. Now you have to add/subtract 500 dice for the different modifiers instead of just modifying the TN in your head.

Well, right. That's what I was getting at. It has to be more dice or auto successes (aka "negative threshold"), and auto successes would be completely insane.

I was just trying to make that obvious to Eyeless, who for some reason is "seriously doubting" anything will result in you modifying the number of dice you roll (despite there being no reasonable alternative, with TN modifiers a thing of the past, and auto successes a thing of insanity).

I think there is a difference between "situational modifiers" and cyberware, magic, and the like even to karma and such. With aim actions it almost makes sense to have a negative thresehold. The sacrifice of an action for one less hit needed is rather straight forwards.

Right, which was my point. The environment modifies the threshold, which I suppose can get as low as 1, much like TNs under the current system can get as low as 2. Modifiers from gear and such, things the player has control over, add directly to the number of dice you roll for a test, which may or may not be the same thing as your "dice pool".

So, you have an Quickness att of 3 and a Pistols skill of 3, making a dice pool of 6 for a given shooting test. You can withhold, say, four of those dice for a Called Shot, or five of those dice for an automatic success, or something along those lines. When you actually make an attack, you use whatever dice you have left over from your dice pool (let's say for instance I'll throw all six). Then you add in other modifiers from magic, aiming actions, cyberware, etc. Say I have a Smartlink, which adds four extra dice to my roll, and I took one aim action, which adds another two, for a total number of dice on this test of 12. The Threshold starts out at 1, plus 1 for fog, plus two for soft cover, etc, for a final Threshold of 5. Extra successes stage damage upward at 2-for-1.

Leaving the number of dice completely in the hands of the player and the Threshold Number (TN ) completely in the hands of the GM will definately stremline play, as the GM doesn't have to consult the player's character sheet as much to determine the TN, and the player doesn't have to consult with the GM as much to find out how many dice to roll.

Eyeless : Word.

yes, that would be a clean mechanic, but the fact remains that the difference between threshold 1 and threshold 2 is very, very wide. there would be no minor bonuses or penalties. this, to me, makes for bad play. it would be comparable, in SR3, to having TN modifiers whose lowest increment is 3. that is, you can't add or subtract 1 or 2 to or from the TN; the least amount you could modify the TN by would be 3.

Agreed. It's roughly comparable to subtracting six dice (five exploding).

~J

Er, technically it's closer to subtracting about 2.5 dice, actually. Which is better than a +1 TN under the current system, which is more like cutting the number of dice you roll in half.

(Edit): Looking over the probabilities though it may actually be better to lower the static TN to 4 or maybe even 3, though that would have odd effects in itself. That would give you a much finer grain on the threshold number, which really does need widening, now that I take a look. These kinds of things are the reason game creators really need at least one mathematician on staff, to analyze the probabilities and crap.

Two successes is roughly comparable to six dice (five exploding); the difference between a threshold of one and two is roughly comparable to three dice (two and a half exploding).

This is only "roughly comparable", though. In the dice analysis paper, it looks like you need more like four or five dice (three or four exploding) per threshold increase of one if you want to maintain what should be a reliable action. And if you're already grasping at straws, you'll be grasping to the same degree if you get an extra two dice (maybe a little less, exploding).

Also, it looks like a threshold increase of two is worth about one reroll for somewhat skilled people with typical numbers of dice (10ish). (As you'd expect, since they'll be rerolling about six dice....) The same caveat with wanting to maintain a reliable outcome and just wanting a chance applies here, too.

Has anyone brought up that the impact of a smartlink is massive? Consider a short-range pistol shot between two unmoving principals on a clear day with no distractions. Pistols 6, no combat pool. TN4 normally, you can expect, what, half successes (no good at probability math)? So basic 3 successes on 6 dice. Add a smartlink and the target goes down to 2, making it likely to get 6, or at least 5 successes. There is now a 17% chance of failure on a single die.

Fixed TN5, ignoring threshholds, how can the same affect be acheived here? Can it? Is 4 extra dice enough? Did I just stick my foot into something that'll blend me up into chunky Sharaloth bits (metaphorically speaking)?

Four extra dice doesn't really do the same thing as TN-2. Four dice is huge if you are only rolling four dice to start out with. But it's nothing if you start off rolling 16 dice. However, the -2 to TN in SR3 always helps you out (making each die between 17% and 300% more likely to come up as a success, depending on where the TN would have been otherwise).

This does nicely illustrate the quirk in SR3's system, though: 6 to 4 is an amazing bonus (3x as many successes), while 8 to 6 does almost nothing (17% more).

I'm not sure which type of problem is worse. Neither is ideal. I haven't yet thought of a good way to get around both problems without needing to roll a huge number of dice. The best one I've heard of is adding "five splitting" to SR3; you reroll 5s separately from 6s and add one if it comes up 5 or 6 again. (So 2/3 of 5s count as 5 and 1/3 count as 6.) But everyone who thinks SR3 is too complicated would probably riot if that were published.

Wouldn't a better solution be for a smartlink to reduce the number of successes you need on a test by a fixed amount?

Not that I'm a fan of the fixed tn system, but if we have to go with it...

Well I'm not sure that we can fundamentally assume that augmentations (cyber, magical, etc) are going to function in the exact same manner given the new system. Currently we are making alot of suppositions based upon a paucity of information.

Consider that currently we don't know the answers to the following questions.

1) Is there a mechanic for botches (all 1s meaning a catestrophic effect)?
2) Is there a mechanic where sixes count for 2 successes like nWoD?
3) What are the mechanics for unskilled skill tests?
4) For skill tests is there a limit to how many attribute dice are rolled?
5) Is the system designed to require 1 success period or numbers of success?
6) Are attributes purchased the same way or are higher than average attributes much more expensive than higher than average skills etc?

Plus a multitude of other questions.

For example they could model the Smartlink as affecting the number of dice rolled, the number of successes needed or perhaps giving a reroll. There are a variety of ways to simulate the increased accuracy of the concept that can be done. Just because the smartlink has worked in a particular way (pretty much insanely powerful for the given cost) doesn't mean it will continue to work that way.

Further I would suggest that stuff like muscle augmentation (+1 to Quickness and Strength per Level) might be significantly changed in SR4 as while increased quickness and strength were important under SR1-3 rules with attribute + skill rolls in SR4 increased attributes are at least doubly important.

I'm not sure we need (or can expect) that characters can routinely have mid to high double digits for certain attributes in the new game. In fact I would be sorely disappointed if that was the case.

Oops. Well, if you will have so many threads with so many pages to them.

Anyway, no, I don't think it is in all cases. A -1 threshold modifier doesn't seem ridiculous for a smartlink. Don't ask me what you'd do for weaker things like laser sights, or aiming once. I guess adding dice. Likewise it would be a problem when you only need 1 success - you wouldn't want the smartlink to be useless then, nor would you want it to guarantee you success.

Like I said, I'm not a huge fan of this mechanic as, among other things, it makes all this kind of stuff more difficult.

It already is a must have if you use firearms regularly, and it already does provide about that level of bonus.

Y'know, a core mechanic of fixed TN doesn't actually mean that nothing in the game modifies TNs. It means that such things are rare and unusual, as opposed to being the default scaling technique.

Consider: Exalted, a fixed TN system, has Sidereals, who are able to adjust the TN by +/- 1 using certain powerful magic.

Consider also: in many variants of d20, the core mechanic is d20 + fixed modifiers vs DC. And yet, sometimes in unusual cases, you get to add additional dice to the roll (examples being Force Points, from Star Wars; and Action Dice, from Spycraft/Stargate) or even get rerolls.

Core mechanics are just that: core. Doesn't mean they're the only mechanics.

If they add things as significant as TN changes and rerolls to non-core rules, SR4 will be an impenetrable tangled morass of rules and improper balance in no time.

Some things, like variable TN, are much better built in from the start so you can plan around them, and so you don't have to buy 15 books just to get the core mechanics used by other SR players (who gleefully use optional rules).

Honestly, I don't think you'll see anything in SR4 that will modify the TN. That would be introducing specific, limited rules exceptions, and would be counterproductive to the idea of streamlining the system.

Yes, only smartlinks. And then adepts. And then a magical spell. And then there will be a totem that provides TN4 for a special kind of test. And then (etc. etc. etc.).

This seems like a bad idea for a simplified system.

Well, I'm not a developer, so I'll just leave it at "wait and see". I wouldn't put my chances of being right above 50% in any case.

I'll just observe that almost all roleplaying games have character options that violate the core mechanics, because such abilities are generally interesting, novel, and intriguing and therefore appealing; and leave it at that.

I don´t care that there would be no small increments to the difficulty of tests, things that are not directly influenced by the character should be handled with the threshold. That is: Smartlink, damage modifiers, etc = more or less dice. Light, terrain, things the target does, etc = higher threshold. I don´t want to inspect every die roll no longer. If that means that tests will be either easy (threshold 1) moderate (threshold 2) hard (threshold 3) very hard (threshold 4) extreme (threshold 5+) or something, so be it.

The whole point of rolling dice is to, y'know, then inspect, sort, and count them, and then to eventually determine what that roll means in game terms. It's why dice are rolled, in any system. Maybe you shouldn't be playing RPGs if you don't like looking at dice.

Me being the GM, I have to inspect all the players dice in the current system. Unless I want to reveal the targetnumber. In a matrix run that is the same thing as revealing the subsystem ratings so in some cases I don´t. Sorry for the confusion.

The players should be able to calculate their dice pools by themselves, counting their own modifiers from damage and equipment. Then make the rolls and call out the number of successes, then I should be able to tell if they succeed or not.

As far as I can tell, if it works any other way it will be slower and harder to use. Having large increments in difficulty is a small problem compared. In fact it could be a positive thing even, as it would be easier to remember situational modifiers if there were fewer of them. There would be no need to consult the visability table, for example, because there might be only two levels of reduced visibility. The calculating of target numbers is not an area of Shadowrun that I´m particularily fond with. I think you could simulate the difficulty of any situation even with larger increments.