Dice systems, Advantages and disadvantages Options

[Ellery]

I'd like to discuss what the various advantages and disadvantages of dice systems are, in order to understand what the current change (as currently understood) means for Shadowrun.

First, let's be clear that dice systems matter. They matter to how you play the game, and they matter to the story. If you're writing fiction loosely inspired by the Shadowrun setting, it doesn't matter one whit what the dice mechanics are. You design the story the way you want to, throw in "smartlink" and "astral" here and there, and generate a piece of fiction essentially independent of the rules. But that isn't what happens when you play the game. You're building a story there, too, except you don't always get to decide which direction the plot will turn, what will succeed and what will fail. The dice decide.

So it's important to get the dice system right. If it's done wrong, you don't just end up with a system that's a pain to play. You can end up with a system that destroys the ability of roleplayers to participate in an engaging story, because the dice are fighting them at every turn. (At that point, some just give up and go diceless.)

I've grumbled at the dice system in SR2 and SR3 often enough, but it has a subtle genius to it when it comes to storytelling. When you want your story to go a certain way, you throw your dice pool at it, and then your karma pool. It's a mechanism for the dice to listen to how much you care about an outcome. This has real power in a roleplaying setting, and opens up all sorts of sophisticated tactics and opportunities if you care to take them. You can ignore them, if you want, and not be bothered by the complexity. Likewise, if you're trying to do something difficult, it's never truly out of reach with the Rule of 6. You'll never perform a really competent job at the difficult task (many successes), but you have some hope of getting it done.
It nicely captures the one-in-a-million basket thrown at 00:01 from across the court, the crazy luck of shooting an assassin in a distant window in the dark, and other nearly but not completely impossible tasks.

My question about the dice system is whether the current mechanics can possibly afford these same advantages. If the answer is no, I then have to ask whether it would make more sense to streamline the old system than switch to a new one. I have other concerns about the mechanics (and other complaints about the existing mechanics), but let's look at these two questions first.

First, let's look at how a player's determination can affect an outcome. With the stated mechanic, taken at face value, someone who is pretty good with a sword is probably also pretty good with a mace and at gunnery and every other skill linked to that attribute. So there's less variability in outcomes to begin with. It doesn't matter if the books declare "Str 3 is average!". It still reduces the differences in people's ability, so if they really care about one thing (like being a good swordsman), it matters less that they're skilled in it (again, everything taken at face value). So in the long-term, the mechanic proposed seems to reduce the impact of determination on abilities. (I'm waiting to see why it isn't always smart to raise your attributes until they max out before paying any attention to skills at all. Maybe raising attributes will be ten times more expensive?)

Okay, so if your determination is lessened in the long term, what about the short term? It looks like it's going to be hard to win a sword duel because your character built up his sword skill. Now that there aren't dice pools, you don't get to decide as much how determined you are to not be hit, or to succeed at once swing. Stuff just happens to you--unless, presumably, you use karma pool.

But there's a problem with karma pool too. Our master swordsman probably doesn't have as much karma pool as a strong runner who has lots of karma, but has barely ever touched a sword. In fact, if karma pool provides much flexibility, the strong runner is probably going to wipe the floor with our master swordsman. Now, granted, with existing pool rules you have some of the same problems. But it seems to me that it will be a challenge to allow the player's determination to affect outcomes, and you could get even less desirable results than in SR3.

Now let's turn to the fixed-TN system and whether an average person will ever, ever be able to accomplish an extraordinary task. The key problem is that if you have fewer dice than you need successes, you're screwed. Without some extra mechanic (like the Rule of 6), if you need five successes and you have three dice, you're never going to make it. Beginners' luck? Forget about it--not with that dice system.

It doesn't really matter how many dice are rolled, or how many successes are needed. The flaw is inherent in the system. If you need 3 successes, someone with a stat of 2 and skill of 1 can do it almost 4% of the time. But someone with a skill of 2 and stat of 0, or a stat of 1 and skill of 1 can never, ever do it, not in a million years. The only real way to fix this is to go to very large numbers of dice. For example, if you need 10 successes, and you roll 10 dice, your chance of success is about one in 60000. The difference between that and "no chance at all" isn't likely to matter in more than one game you play in your lifetime, so there everything is cool. But rolling such massive handfuls of dice all the time is really awkward.

What do other people think? Can you see how to keep the dice from destroying the impact of a player's determination and desire with this new mechanic? Can you see how to give the unskilled some chance of success at difficult tasks? Do you not care to have these things in your games?

(Hint to playtesters and those in the know: have you played these rules with badass munchkins who roll fists full of dice? With middle school kids? With a spindly geeky hacker in a fistfight? Do they still work there, or are the rules for typical runners only and broken for everyone else? This is where the problems are more likely to show up and terrorize the intended storyline.)

[Kagetenshi]

[Raskolnikov]

I echo the concern about the border cases. Additionally I'd like to say that threshold for difficult tasks reduces the degree of sucess significantly. Even skilled characters will find many, many all-or-nothing situations. While this may be applicable in some situations (defusing a bomb) is it really a desired mechanic to build into the system for all situations (running a race, tracking down data, tailing the badguys, or sneaking through the forest)?

I was quite happy when threshold was removed in the decking rules the first time around. It reduced the problems that existed in the border-cases. While I feel it aggravates them, I don't think just threshold causes these problems.

Without compensating mechanics, natural talent will quickly backseat training. This posses significant concerns in many cases. Every decker needs to be a genius. Every gunfighter needs to be a wiry, lithe athlete in prime condition. Every martial artist needs to be a hulking brute. If they are not, they will find themselves little better than the genius winging it, track star picking up a gun for the first time, or brick of a man swinging his fists like hammers.

Granted the system hasn't been fully disclosed but the jack of all trades seems to not be a character buying many skills but instead spending most points at creation in attributes.

[kevyn668]

What will be, will be. To struggle, however, gives meaning to life.

[ Spoiler ]

A big fat case of NERPS (will we need to rename NERPS in SR4?) if you can tell me where that's from.

[Spookymonster]

With a Rule of Six-ish ruling (e.g., every 6 entitles you to add 1 more die to your pool) and a cap on pool size for unskilled tests (e.g., no more than 3 dice allowed in your pool unless you have an applicable skill) you'll make the impossible possible (you rolled 7 successes with a pool of only 2 dice thanks to the Rule of 6) while keeping the implausible unlikely (a Night One with Qui-9 and no Pistols skill has the same chance as an unskilled human with Qui-3).

Of course, this is pure conjecture on my part.

[Vuron]

Grr I posted some possible solutions to the tendency of attributes to overweigh skills in this style system in the faq thread. There are so many damned threads these days it's hard to argue in all of them ;)

[Wireknight]

I agree with Ellery; this system has a potential to eliminate some of the good things that were possible in Shadowrun, i.e. control on the part of the player from moment to moment how much effort they expend toward defense or offense in a fight, down to a single die, and the fact that most any task one can undertake in Shadowrun can at most be very unlikely, rather provably impossible.

However, we know the dice system, but we have little idea how that dice system will actually be played out as far as rules go for combat, magic, hacking, and the like. Ideally, they will complement the system and allow greater modelling of outcome without sacrificing potential in exchange for simplicity. I also hope that they will ultimately work in similar fashions. The major problem I always saw, in Shadowrun, was not the way dice were rolled, but the fact that most systems were totally different in number of actions, target numbers, etc, etc, etc...

[Ellery]

Kage, I always keep the rules in mind when writing fiction myself. I'm just contrasting the different constraints faced by people who are working artistically to develop the feel of a setting through fiction (where rules-breaking or rules-bending happenings can nonetheless be thrilling and motivating), and what happens when you're actually playing the game.

But just imagine if authors always kept in mind the presence of a threshold system, and their hero could never perform the miraculous save at the last minute because he needed too many successes. It'd take out some of the thrill of the writing, wouldn't it? And of playing, too, in my opinion.

Rask makes a good point about the experts-suck phenomenon. I don't fail to tie my shoes 10% of the time, and I don't think my attribute plus skill in shoe-tying is as much as six (which has a 9% failure rate on a single success). I'd wager that I can tie my shoes properly 1000 times in a row without fail, and I really doubt that I have a skill plus attribute plus modifiers of 17 (0.1% failure rate). But maybe it's really, really inexpensive to buy a very high level of shoe-tying skill.

Spooky suggests a way to generate the possibility of success for an opponent of lesser skill. That's rather weird mechanically, since you have to keep grabbing more dice as you roll 6s, or reroll ones you've already rolled which makes the final count harder. But it seems to solve the problem. For the hypothetical example before, with three successes needed and only two dice, you jump from a 12% success rate with 3 dice to get 3 successes down to 4.5% with only 2 dice (and 1% with one die). That works pretty well. I like it. I hope they do it. If they weren't planning on doing it, they should.

Vuron refers me to his FAQ thread post. I'll go read his stuff there, then come back and comment here if appropriate.

WK's point about the dice system not really being the source of complexity in SR3 sounds right to me. It's the million different types of rules for every different situation. I suppose with a simpler dice system, it's harder for people to make rules seem different since there just isn't that much choice in what you can say for a rule. I'd prefer that editors keep things consistent, but hey, if the dice system needs to do it, maybe that works too.

[Demonseed Elite]

Yeah, I wouldn't say it was necessarily the dice system that was the source of complexity in SR3. The actual mechanic of rolling dice in SR3 was pretty easy. The more complicated part was there really was very little in the way of a universal standard for determining the nature of a test, there were so many potential modifiers to target numbers, and you had weird quirky situations where you had a flat change with TNs of 6 and 7, for instance, which ended up with the same modifier being less meaningful to someone in one situation than it would be to someone else in a slightly different situation.

The NDA forces me to be vague, but here's about what I can say:

Did similar issues like these come up when I looked at the playtester rules? Yes.
Did they occur to the people actually doing the playtesting? Yes.
Have ideas been presented to address them? Yes.
Are those ideas being tried? Yes.
Am I, personally, 100% satisfied? Not yet.
Is the work done yet? Nope and nothing is written in stone.

I realize that's not all that helpful, but my hands are pretty tied.

[Ellery]

DE, I like that this system seems less prone to weird threshold effects (TN5 to TN6 vs TN6 to TN7 is pretty bad) but only if it is not prone to the much more drastic threshold effect of dropping below the minimum required number of dice. Adding big new problems in order to fix minor old problems is usually a bad design decision.

I think I'm starting to be able to see a way out of the potential problems that have been created. With a Rule of Six (which is hardly a "complicated" rule if it always applies), the success issue goes away. With very cheap skills, the attribute-focus goes away. With some sort of effort-based pool, the impact of the player's attention is retained. And then the only problem remaining is that I can't tie my shoes reliably, which is probably tolerable (if caring about success has a big enough impact).

Those changes (if they are changes) still leave us with a simple ruleset. Not quite as simple as before, but pretty simple and working is better than really simple and broken. And it is definitely simpler than the existing ruleset.

[mfb]

i think the 'exploding' dice idea is the best one, for allowing one-in-a-million situations.

[Demonseed Elite]

You should be able to tie your shoes reliably. I've seen a simple mechanic that would address that. I don't think I've gone over the bounds of the NDA by saying you should be able to tie your shoes reliably in SR4. :S

[Ellery]

You could have automatic successes, of course. But even I have failed to tie my shoes at least once in the past decade.

[Sandoval Smith]

[mfb]

it's a metaphor, meant to convey the concept of easy tasks.

[Critias]

[vapor]

Critias, that had to be the most eloquent post i've ever had the pleasure of reading.

[Shockwave_IIc]

[Wireknight]

[mfb]

like when aircraft and anti-aircraft meet, creating a massive explosion and the annihilation of both.

[Austere Emancipa...]

I can't find it right now, but I'm sure Patrick Goodman implied quite strongly in one thread that they won't let impossible dice rolls to come up in SR4, ie. that they will put in some system of allowing you to score however many successes even if you've got a limited base dice pool.

Like mfb, I think exploding dice are the best for this.

[Patrick Goodman]

[Critias]

Which, unfortunately, still doesn't help with the loss of player-controlled pools. I'm more upset about the loss of those, and the tactical thinking they brought with them, than I am the basics of a fixed-TN system.

[Synner]

Let me submit - for the sake of argument - that the " tactical thinking" that "player-controlled pools" supposedly brought was actually quite limited, essentially boiling down to a boost to the first attacker's strike and to the defender's first defense - to ensure the first strike takedown which is what's kept most SR3 firefights to less than a Combat Turn. In the years I've run SR3 that's been the case for all but a handful of ocassions, and the number of characters with consistently no pool left whatsoever for second or third actions (and hence truely tactical usage) is minimal (except in the case of particularly high-skilled or augmented skilled attackers) - and even then 1-2 dice left over do little for the probability curve.

Yes, there is the illusion of tactical use. Yes, there is the perception that they allow more leeway in Combat. And yes, I quite liked them. However, in very few cases did Pools (Combat and Sorcery) really work that way in practice.

They did somewhat increase survivability in first action combat, but even that was balanced by combat types heaping on modifiers to either increase their skill pool or lower their target numbers - which in turn led to people using up all their combat pool on first strike defense and mostly cancelling out any true tactical application.

[Backgammon]