Clash, that meant "yes, go for it!"

Xanther, those are all possibilities, but I was thinking I'd combine it with the Action Point system we're discussing in another thread. I just gave them separate threads so we could focus on each part.

The various odds aren't too difficult to calculate. The average is just 3.5 per d6 used, so if the average is below the skill total, then you've more than 50-50 chance of success.

Quote from: Kyle AaronClash, that meant "yes, go for it!"

Awesome! Thanks, Kyle!

-clash

Quote from: Kyle Aaron...
Xanther, those are all possibilities, but I was thinking I'd combine it with the Action Point system we're discussing in another thread. I just gave them separate threads so we could focus on each part.

Will take a look.

QuoteThe various odds aren't too difficult to calculate. The average is just 3.5 per d6 used, so if the average is below the skill total, then you've more than 50-50 chance of success.

Yep, I was thinking of a more detailed analysis, especially on the higher numbers of risk dice as the change in odds from rolling under 13 to a 10, is very different on 3D6 than 5D6.  It also means a +2 to doing something has a very different meaning between 2D6 say and 4D6.

The detailed analysis wouldn't be something the players would need but could help you determine if the system is giving the results you want.  How impossible is that roll on 4D6?  You can say it is very hard but a 20% chance is different than a 10% chance.  If I use a situational modifer range of -4 to +4 what does that mean?  You can also get an idea on the overall effectiveness, normalizing/weighting the result (e.g., damage) by the probability of obtaining it to avoid odd "sweet spots".  That is you are probably going to want some monotonic relationship of effectiveness across numbers of risk dice and you can better design how much you want the all-or-nothing 5D6 roll to be all or nothing for various levels of skill.

In talking about this with my players, the point came up, "okay for combat you want to roll lots of Risk Dice because you do more damage, but what about other stuff like diplomacy and lockpicking? Why wouldn't you always roll just 1d6? What does "number of successes" really mean?"

When I noted that rolling all "6" is a Balls-Up in the system, I suddenly realised that I actually have a cinematic game here

The idea is that "number of successes" influence-

• How well you perform the task, and
  
• How quickly you perform the task.

Lockpicking
If  you roll 1d6 to do your lockpicking, one success, okay you pick the lock but it takes half an hour. That's okay if you're at home messing about, but not so good if you want to slip quickly into someone's house, and you're picking the lock on the front door in broad daylight; to do it in just moments - twist, click, snick, twist, you're in - might require five successes.

Diplomacy
Or suppose you're using diplomacy to try to get a discount on a $10,000 car. Roll 1d6, get one success and he throws in a free pair of driving glasses and a windscreen shade. Roll 5d6, get five successes and you get the whole thing for half-price.

Interview
You're trying to get information from someone, say you're a cop interviewing a suspect who's actually innocent. Let's call him, "Momo Haneef." Roll 1d6, get one success, you find out that he's done nothing - but it takes you two weeks of having him locked up. But roll 3d6, get three successes - so you interview him for a day, realise he's done nothing and let him go.

Think of the Risk Dice as quite literally how "risky" you're trying things. Bigger risk leads to bigger payoff.

The other point is that any time you roll all sixes, you get a Balls-Up. So with 1d6, while you'll succeed 5 out of 6 times (assuming any ability at all), you'll have a total Balls-Up 1 in 6 times. That is, when you approach a task with minimum risk, very cautiously, if you fail at all you fail spectacularly! Whereas with 2d6 it's a 1 in 36 chance, with 3d6* it's 1 in 216, and so on. As you take more risk, the chance of simple failure increases, but the chance of a total balls-up drops. No, this isn't supposed to be realistic - it's cinematic reality. In movies, the greater the risk someone takes, either they succeed brilliantly or they simply fail.

Consider driving:- Jason Bourne drives his car off the back of a building, either he lands on all four wheels and drives away, or else he's stunned for a moment, gets out and walks away. He does not end up with whiplash, a broken leg, trapped in his crushed car and have to be prised out of it four hours later by paramedics. That would be a Balls-Up, but Jason Bourne's player never rolls less than 3d6 Risk Dice, he doesn't do anything by halves, he goes in there balls-to-the-wall like a maniac. If he went cautiously - 1d6 Risk Dice - then in a fair number of scenes he'd have a Balls-Up. That's what the cops and his pursuers do, which is why their vehicles end up scattered across the landscape like betting papers after a horse race, while Bourne zooms on happily, or else crashes out but just lies there stunned a moment, climbs out of the wreckage, shakes his head and runs off.

We've playtested this for one session so far. The players seem to be taking a while to accept that how many dice they roll is up to them. There's that moment of hesitation when they look to the GM for guidance or ruling. It gives me great pleasure to say, "how many do you want to roll? The more you roll, the better your result if you roll under!" This really shows each player's personality, one is quite conservative, another eggs everyone on to roll lots of dice, one sits there thinking for a moment calculating the exact probabilities, etc.

We've not had a combat yet, so I can't comment on the Action Points in the related thread, or on how Risk Dice work in combat. But in the non-combat stuff we've had, the Risk Dice are working well. Players are getting a real sense of control over how hard they're pushing their characters. They're getting choices.

Incidentally, this is a newspaper report from the game. The PCs are pursuing an escaped murderer and drug dealer, Bill Cluny. They got him, but it wasn't very neat and nice. That's what using Risk Dice does


Arson, Gunshots at Munnamurra



October 12th, 2007 - CFA and Victoria Police on Wednesday were called to a scene of violence at a homestead in Munnamurra near the CCT border, with reports of fires and gunshots. On arrival shortly after 10pm in response to several emergency calls, they found a burning unregistered vehicle, and three men and one woman bound with No.8 wire. The woman had been shot in the stomach with a .308 calibre rifle. The bound people were released, and the woman, Sheryl Johnson, 38, an unemployed paralegal, was flown by helicopter to Monash Hospital, Melbourne. Shortly afterwards, one of the sheds of the homstead went up in flames, the CFA having to call in a chemical unit. The fire spread to nearby bushland, and it took until 6am to control the fires.

Two of the men immediately absconded, but the property owner Bill Martin, 39, a dog-trainer, remained to assist police with enquiries. Martin claims that he and his guests were victims of an armed robbery which went wrong, and reports several thousand dollars stolen. Police are investigating connections between Martin, and Bill Cluny, the so-called "Hedgetrimmer Killer" and his associates Dan Dinke and Sol Klinsky, who last week escaped from Remand Prison, Bidawal AFP HQ, and are implicated in the armed robbery of Coochimurra Credit Union. Police believe that two men seen on the property that evening, a well-dressed Chinese and a tall man of European appearance with close-cropped hair and a military demeanour may be able to assist them with their enquiries.

I haven't done exactly this system, but it does remind me of some trade-offs of hit and damage.  It works mechanically, but for some situations it seems like a strangely arbitrary out-of-character choice.  i.e. You decide at the beginning that you could fail, but you can't get a marginal success.  

As a parallel within traditional systems, something that's bothered me about many hit location systems is that trying to hit the center of the target reduces the chance of hitting the target at all.  (i.e. trying to hit the heart makes you more likely to miss the torso, say)  I can choose it by out-of-character logic, but it often seems jarring.  

Two alternatives that let you vary risk but still give a range of possible outcomes:

1) Fudge lets you do this very easily by choosing the number of Fudge dice to roll.  (Each Fudge die has two -1s, two zeroes, and two +1s.)  It always has zero as the average, but more dice allow greater success but risk worse failure.  

2) The Aurora RPG (a self-published hard SF game) used failure dice, where you could choose how many to roll.  You could roll one to nine d10.  If the number rolled is N, you got a base of +N to your skill.  But for every die that was N or less, you subtracted 2 from your total.  So if my skill was 5, and I chose to roll 3d10.  Then my total would be 8, minus 2 for every die that was 3 or less.  

The result was that 2d or 3d were tied for highest average.  However, you still might want 0d or 1d if you wanted to be sure of not getting too low a roll.  And higher dice were if you really wanted to gamble on getting a great success.  The probabilities work out like this:

Dice Avg RMS
0d - 0.0 0.0
1d - 0.8 0.6
2d - 1.2 1.1
3d - 1.2 1.6
4d - 0.8 2.0
5d - 0.0 2.4

I ran three sessions of StarCluster 2 using Risk Dice. The players liked it, and would use them again. This set of characters are what we use for experimenting with different TR methods - they were last used to run a few sessions diceless - so my group are familiar enough with different StarCluster TRs.

We used a modified version of Kyle's system. Stats in SC 2 run theoretically to infinity, but we had capped them at 15 max for the pervious testing. Skills are also theoretically infinite, but practically seldom get over +5. Players could take an automatic success for one success, or roll the dice and risk failure. That meant three or even four dice minimum. There were no botches. Each success was worth 20 points base, with modifiers added.

During the three sessions, only once, in a rather unimportant choice, did anyone choose the auto success. The players really prefered risking the dice for better return!

Initiative was also risk based. Roll as many dice as you want, with each additional die after the first costing one success off your roll. Init goes highest to lowest. The players never chose the risk dice here, always choosing to take the safe initiative. Perhaps the cost was too high? Maybe I should have gone with a +1 per die rolled later for each additional init die?

Anyway, it was fun, and successful! I'll be adding Risk Dice into the StarCluster System Toolbox!

-clash

Marginal failure? Eh? This is the same as almost every system out there, you can get,

• Critical success - roll all 1s
• Success - roll xd6, under Attribute + Skill
• Failure - roll xd6, over Attribute + Skill
• Critical failure - roll all 6s

Almost all systems have those four basic results. That's what's imposed by the GM. For degrees of success beyond critical success, the player chooses what they're going for. Rather than the GM telling the player the difficulty and the modifiers to pass/fail, the player decides how much to push their character, and then rolls the appropriate number of dice.

The more dice you roll, the less likely you are to get a critical failure, and the better your success if you do succeed.

A character may do things well, or do them quickly.

• With 1 success (rolling under Attribute + Skill on 1d6), they do the task neither well nor quickly, but it's done.
  
• With 2 successes, they do it either well or quickly.
  
• 3 successes, both well and quickly.
  
• 4 successes, and either very well or very quickly.
  
• 5 successes, both very well and very quickly.  

So for example a character is in the middle of a firefight and trying to pick a lock to escape. The player knows that with just 1 success, they might take something like ten minutes to do it - so even someone with no skill can do it, but not quickly or well enough given the firefight happening. To do it very quickly, they'll need 4 successes. So the player chooses how much the character is pushing themselves. That reflects taking risks with your skill to be impressive.

For example, if I'm driving, I can take that hairpin turn at 15km/hr, I won't impress anyone but I'll certainly be able to do it. Or I can be a smartarse and hit it at 90km/hr, in which case I either crash or look very impressive.

Because it's a cinematic game system, higher Risk Dice increase the chance of simple failure, but decrease the chance of critical failure. That's rather like the way when Jason Bourne drives his car backwards off a building, he either lands on all the wheels and drives away happily, or he lands upside-down but shakes his head and climbs out; he never crashes out and needs the jaws of life to prise his twisted remains from the smashed car hours later. It's only his foes chasing him who crash out and are seriously injured. That's because (in game terms), Jason bourne always uses at least 3d6 Risk Dice, whereas his more cautious foes chasing him use 1d6 or 2d6; they're much more likely to roll all 1s.  

For targeting hit locations (though I don't know what this has to do with Risk Dice), so far I've just allowed "aim high", "aim low" and "aim middle." In any case the first 2d6 of your to hit roll is also read as your hit location roll. I can't show you that because I don't know how to do tables in forum posts. But the basic thing is that if you aim high, you won't hit anything lower than the groin, and have a much improved chance of striking the head; if you aim low, you won't anything lower than the chest (it's assuming you're using a weapon with your hands), and have a much improved chance of hitting the lower limbs, groin and abdomen.

If you want aiming at hit locations to not have a lesser chance of hitting anything at all, then you need to either have go for an overlay with groupings the way Millennium's End did it, or else a table equivalent of that. It's also pretty easy to do things like, "for each 1 margin of success, you can move the hit location one row in the table" etc, so that the player rolls the attack and only after chooses the hit location.

Still don't know what hit locations have to do with this, though.

I guess I wasn't clear.  Let me try again.  

So in my example of, say, GURPS hit locations.  If you try to hit a target right in the heart, then you are more likely to miss him altogether, but there is no chance of hitting him not in the heart.  So if I was shooting at a target, the hits would all be grouped over the heart and then a bunch missed the target - but no hits anywhere else on the body.  If there is really an in-character choice to get such a grouping, that would be strange.  

By parallel, I think that in some circumstances it is strange that trying to do the job well means that you are more likely to fail altogether, but you can't end up doing the job poorly.  Similarly, trying to do the job quickly means that you can't get it done more slowly.  In real life, say, there are sometimes tasks which I try to get done quickly -- but I get bogged down.  I do complete the task, but only after longer time than I had tried for.  

So, for example, I might try to climb over a fence quickly.  I get to the top but my clothes get caught.  I try to just pull them but they don't come.  However, after some time carefully pulling them loose, I make it over.  I had tried to be quick, but I succeeded, but only slowly.  

However, in this system, if I as a player try to get the task done quickly, then I am less likely to succeed at all and there is no chance that I will succeed more slowly than I had attempted.

Responding mostly to the OP...

I was thinking of doing something similar for a supers game.

Characters would have ability scores of 5-9 and skills of 5-9 and it would be a roll under on 4D(zero to five).

You could roll under as normal or add dice to do more extraordinary things.

Also, things like superstrength would allow you to stunt without adding dice up to a certain point. Because superpowers don't make you more reliable so much as they make you more extraordinary.

And damage would just be whatever you rolled + mods.

I got my idea less from dicepools and action points and more from Iron Heroes skill challenges and Alternity situation dice.

Quote from: jhkim....  
By parallel, I think that in some circumstances it is strange that trying to do the job well means that you are more likely to fail altogether, but you can't end up doing the job poorly.  Similarly, trying to do the job quickly means that you can't get it done more slowly.  In real life, say, there are sometimes tasks which I try to get done quickly -- but I get bogged down.  I do complete the task, but only after longer time than I had tried for.  
....

I find that odd as well.  
Would it work to have a critical failure be two or more 6's instead of all 6's and critical success two or more 1's?  So if you take the easy way out, 1D6, no critcal failures but no critical successes either.  

Another wonky idea off the top of my head.  Potentially a player could use skill somehow to convert a 6 to a 5, so the James Bond types use skill to perform risky actions without carshing and burning.  The poor smucks chasing him still have to take the turn at 90km/hr, but they don't have the skill to pull it out if they fail.  That is on 5D6 the odds of getting two 6's is failry high.  Old James can convert a couple of 6's to 5 fives to get regular failures (you know like when he meshes up his car just gets more trashed with bits flying off).  Opposed to the smucks, when they get two 6's not such luck they flip and burn.   I guess James would have enough skill to convert four 6's to 5's so here never gets killed doing these things.

Quote from: jhkimBy parallel, I think that in some circumstances it is strange that trying to do the job well means that you are more likely to fail altogether, but you can't end up doing the job poorly.  Similarly, trying to do the job quickly means that you can't get it done more slowly.  In real life, say, there are sometimes tasks which I try to get done quickly -- but I get bogged down.  I do complete the task, but only after longer time than I had tried for.

That's why the GM needs to have a couple of brain cells to rub together to interpret the results. A failed driving roll doesn't automatically mean you crash and burst into flames, a failed lockpicking roll doesn't automatically mean the door is stuck forever.

That's what the GM's there for, to narrate results. A GM should in general not say, "nothing happens." That's boring.

It's also often the case, as jhkim describes, that sometimes while you're in the middle of doing something you realise it's more difficult than you thought. In game terms, what's happening is that the player chose lots of Risk Dice and didn't roll under, and the GM then either says, "okay, you can do it but it'll take longer than you planned," or "you're a bit into it then you realise you can't do it quickly unless you try a different approach. Reroll!"

After nine sessions, we have some playtest results. Lessons learned:-

Players like gambling
Before most rolls there was a moment where a player hesitated, and was egged on by the other players. "Go on... four!" Players enjoyed this.


Gambling hurts "getting into the game"
The focus on numbers and dice before the actual description of the character's action and results take players away from getting into the game world and their characters. Hacking into a foe's computer system or deciding to torture an NPC for quick information isn't so interesting or big a decision when you've spent some moments talking about how many dice to roll.


High numbers break the Risk Dice system
We learned that the attribute + skill total is very important - obvious, I know, but still.

We were using a 2-12 scale for Attributes, and a 0/2-12 scale for Skills. An averagely competent person will have 7, so the total will be 14. That means the player can roll 2d6 with certainty of success, and 3d6 with a very high chance of success. 4d6 is exactly 50-50.

This means that people averagely competent in their job can do the job either very well, very quickly, or else both well and quickly.

It also means that with the use of complementary traits (decide which trait supports it, roll xd6 Risk Dice, get +x to primary trait, keep going till you fail a roll or run out of things to apply, then roll the primary trait) averagely competent PCs can fairly regularly have 4d6 rolled against their traits. 5d6 then doesn't seem that impressive.

What we were originally aiming for was to have 2d6 be most common, 3d6 be reserved for masters of the area (high attribute or high skill), 4d6 for high attribute and masterful skill PCs, and 5d6 only with lots of complementary traits and/or desperation. Rolling 5 Risk Dice should be a notable and memorable event, the way it normally feels to get a critical success in most games. Instead it was quite regular.

This was also a result of being generous with xp, so that within a few sessions PCs had 9+ in their character's area of focus traits. Totals of 17+ and thus rolls of 4d6 or 5d6 were common.

In using a 1d6 to 5d6 Risk Dice system, then, it would be better to have traits rated 1-6 rather than 2-12, or else to base rolls on one trait only.


Conclusion
The Risk Dice system is good for "adventure" or "hack" games, and not at all for "thespy" games. Traits need to be rated on a smaller scale, or else not much xp given, to ensure that rolling a lot of Risk Dice is unusual.

Why not switch to d10?

Or was it a specific design decision to stick with d6's ?