A thought about dice rolling - contested or target number

[malladin_ben]

Hi folks,

I'm just thinking through a basic dice rolling mechanic. I'd like to get a nice neatuniversal dice mechanic that I can use in the game no matter what is going on.

The mainstream approach seems to be that you can either make a roll against a flat difficulty target number, backed up by making a contested roll if your action targets another character.

However, I would prefer to have one universal way of doing this. Therefore I'm thinking that either every roll involves a set target number, calculated from the target character's stats if appropriate (the one-roll method) OR every roll involves the player making a roll and the GM making another roll to determine success (the opposed roll method).

I can think of the following advantages for the One Roll method:
- faster at the point of resolving the action
- no messy rules needed for determining who wins when the results are tied
but the following problems:
- one party in a contested roll doesn't feel like their character is playing a part
- need to determine which characer rolls and which character provides the target number (the system rules I am developing I think will go to some way of mitigating this by the structure of the action system)
- calculating the target number might be time consuming (although this can be mitigated by using a standard formulae and possibly recording standard target numbers on the character sheet)

I can think of the following advantages of the contested roll method:
- adds a more dramatic flavour to the game
- automatically creates a bell-curve of results
- no need to determine which character is going to make the roll as both characters roll
but the following problems:
- reduces the random spread of results (although this can be mitigated by simply increasing the size of the dice)
- time consuming making two rolls for everything
- problems deciding what happens on a tied result - you can have a partial success, but that may lead to further problems in itself, or a series of tie-breaking rules, which in tern just ends up adding more complexity.

Personally I think the balance of my own opinions is in favour of the one-roll method, but maybe I've missed some issues. What does anyone else think?

Cheerio, and thanks

Ben

[danielsan]

Uh... I'm confused on how your "one roll" method is different than the mainstream method. The way you described both seemed exactly the same.

[Bastoche]

For the system I'm currently working on, it's gonna be pool dice and ALL opposed rolls, even against a "door". There's no reason to make opposed rolls against non-active participant other than a) keeping the mechanic streamlied and b) making the statistics universal no matter what. Think of the non-active "participant" roll as "luck" or something similar.

[malladin_ben]

In the mainstream method you have both types of roll - one-roll for uncontested tests (such as climbing a mountain) and opposed-roll for when you're up against another character.

the one-roll method I suggest always uses a roll vs a target number, calculating the target number from the opposing character's various stats.

hope that clarifies things,

Ben

[malladin_ben]

Ah Bastouche, you've chosen to go down the all opposed rolls route. I have no problem with this thematically, or describing how it works, I'm more concerned with the practicalities of implementing  it. How have you resolved some of the difficulties I mention above?

Thanks,

Ben

[casquilho]

- problems deciding what happens on a tied result - you can have a partial success, but that may lead to further problems in itself, or a series of tie-breaking rules, which in tern just ends up adding more complexity.

One way to eliminate this issue is to not allow ties. I have seen in some games where the person doing the action must have a higher number (or more successes) or it goes to the defender/other party.

By stating that the attacker/initiator must have a better score, you do not need a tie breaker rule as there are no more ties. Even if both end up with the same number it is not a tie, but rather a failure for the attacker/initiator.

Daniel

[Bastoche]

I handle the ties how Daniel suggests.

As for the resolving time, 2 things to consider:

1) the game is rules light so there aren't many rolls to begin with
2) consider I haven't playtested the game yet and I still got some things to think about.

To put things in perspective, the main idea behind the mechanic is a task intent resolution mechanic. Basically the player states the intent/desired goal and what skill or skills are to be used. Then the GM sets the conscequences for failure. If the player agree to the failure terms, he rolls. On a single roll, you role a D10 dice pool against an Ob5 and count the success and the GM rolls the opposed roll. If the player (or more generally the "attacker") gets more success, the intent is accomplished if not, conscequences of failure are accomplished.

You may chain tests and only the last test determines the outcome. The precending tests are used to give bonus (when successful) or penalities (when unsuccessful) for the upcoming rolls.

For example, you could have a character that wants to inflitrate enemy lines using steatlh. He could decide to wear the appropriate diguise (first test), cause a diversion (second test) and then infiltrate (third test). If the first 2 tests are successful, the player gets a cumulative bonus to the task for the intent. If not a penality and if one pass and other fail, a simple unadjusted test.

There are other mechanics to give better chances at succeeding depending on the circumstances etc. The idea is to create a build-up that might give very high chances of success with appropriate strategy.

Furthemore on the time consuming aspect: we'll heavily use the idea of "say yes or roll the dice". Whenever a player attempt to do something "credible" (as in "something the character should be able to achieve fairly easily) and more importantly something at makes the game go along in the proper direction (in terms of plot) the GM will grant automatic successes.

For example, if the players need to hunt a criminal escaping from the law, the players may agree on a certain way to track down the guilty felon. If the route the players take is coherent enough and that the game cannot move anymore foward until the players find the guy, then we cut to the chase and just let the player have him. You might imagine that the GM asks for a single roll from the players to determine if they find him or he finds them with whatever advantages or disadvantages it may give to any party. It's intended to be nar game with cinematic (frequent) combat action. Hope it helps.

[J. Scott Timmerman]

Aloha, Ben.

Yeah, as long as your game is rules-light and employs "Say Yes or Roll the Dice," I'd say go with contested rolls here.  With SYoRtD, every roll is opposed by another player simply because that player (usually the GM) didn't say yes.  That player is taking up the yoke of representing the opposition to that task, even if the opposition is inanimate.  Therefore (s)he can make the die roll.  Note that Bastoche mentions "consequences of failure".  This is important.  Failure in itself is boring.  I think most people would agree with me here that simply taking longer at a task for failure is also boring.  If there are no real consequences, why roll?  And since there are consequences, why not just justify the opposing roll based on whether the consequence itself succeeds or not?

With opposed rolls, I stick with "ties mean both players succeed" - whatever that happens to mean.  I just hate it when a roll ends up meaning nothing.  In my game, one of the primary philosophies is that every round something happens to change the state of things.  An important facet of this is that Defense never succeeds - characters must do something.  I have been implementing some nifty zodiac paper-rock-scissors mechanics just to break the ties - just to make when characters actually do tie more dramatic.

Another thing - You don't have to increase the die size on opposed rolls to increase the spread.  Opposed rolls increase the spread by default.  Yes, 1d vs. 1d does create a pyramid 'curve' with a peak in the middle, however this peak is no higher statistically than any one result on 1d alone!  And everything else to the sides of the peak is actually a lower probability and therefore more spread out than a single roll against a target number.

Just to play devil's advocate for target numbers, though - if you desire a more organic and unfiatty way of arbitrating whether something should be a challenge or not, particularly if you're going for a Simulationist agenda, you may want to avoid SYoRtD.  In that case, it may be best to stick with a target number for everything to speed things up through all the rolling.  IMHO, however, I don't think that system in-and-of-itself has traditionally done a good job of providing interesting opposition.  The players come up with more interesting stuff, and strong-system Sim can be a bit limiting there.

-JT

[Bastoche]

I just want to further emphasize the "consequences" bit. I meant it exactly how Jason expand on it.

[Hereward The Wake]

Jason
Would you mind going into a bit more detail on the "nifty zodiac paper-rock-scissors" rules you mention?

If this is going of topic for this thread we can split on to a new one.

Cheers
Jonathan

[Jack Aidley]

When I do dice vs. dice in my own system I don't get the GM to roll the opposing dice. Get the player to roll two different colours of dice and they can do it all themselves. I find this to be quicker and neater.

[casquilho]

When I do dice vs. dice in my own system I don't get the GM to roll the opposing dice. Get the player to roll two different colours of dice and they can do it all themselves. I find this to be quicker and neater.

To a point I agree thsi would be easy. However there are times (based on the game of course) where I do not want the player to know right away if they failed or not. By allowing the player to know for sure if he failed or not could lessen the game drama and impact.

Example: Playing a game based on military actions. The player asks if he can see anyone hiding in the small village he is about to enter. I have him roll his spot check. I roll some dice. I tell him he does not see anyone. Does that mean there is someone? Did he "fail" in his contested roll? etc. His not knowing he succeeded but there is not someone helped me keep some drama in the moment.

Daniel

[J. Scott Timmerman]

Would you mind going into a bit more detail on the "nifty zodiac paper-rock-scissors" rules you mention?

My current project involves Lucky Numbers, based on a character's date of birth/zodiac, which is in turn decided alongside character personality during character creation.  These lucky numbers are already used in the event resolution mechanic.  By setting up paper-rock-scissors relationships between the numbers, ties only occur when the lucky numbers AND the Cipher (=Threshold/Result/Margin) are the same.  I'd love to go OT and talk about how all this ties in with the game and setting I'm creating, but I'll start a thread about that later. ^^

You can do tiebreakers all kinds of ways, though.  Hell, if nothing else, you could have a decimal component stat or one for each stat that breaks ties, and isn't used for anything else.  If all rolls in your system have an active and a passive participant, you could arbitrarily decide between the two as others have suggested.  But you don't want to force yourself into a corner by saying that "all rolls have an active and passive participant" unless that is congruent with what your game is about and how you want your game to feel.

I mean, when every roll is about "success vs. failure for one task/action," what happens when the character is "on the line"?  That's easy - just arbitrate that ties are always success, always failure, or something in-between.  More importantly, Ben, if you eventually do decide on a mechanic which isn't just "success vs. failure for one task," you've got to think about when and why ties become problematic.  What types of events do you want to encourage to happen in your game?  What type of action do you want to drive toward? 

-JT