[Andrew's Unnamed Game] Dice Mechanic

[Andrew Cooper]

In my playtest, one of the slightly off things I noticed was that keeping up with how many and what type of dice to roll from the beginning of a Conflict to the end was a bit of a chore.  It might have just been because the game was new and we where used to keeping up with the information but I can definately see it as a hinderance to players new to the system.  Thus, I'm considering changing the mechanic.  Let me give a brief rundown of the current system vs what I'm thinking about for the new one.

Current

Players buy dice pools that default to d6's.  Target number is always a 4.
Activation of Traits by the Player moves the dice up 1 die type.  d6 -> d8
Activation of Traits by an Opponent moves the dice down 1 die type. d10 -> d8
Convictions add +1 success to the roll for each Conviction that applies.
Players roll their final dice pools.  Most successes win.

Example:

Player A rolls 4d6+1 and get results of 2,3,4,5.  That's 2 rolled successes plus 1 for the Conviction.  Total of 3.
Player B rolls 3d8 and gets results of 1,5,6.  That's 2 successes.
Player A wins and gets to narrate his success.

Proposed

Players start with 1d6.
Players buy +1 to the roll for each Power spent.  (Maximum equal to Attribute)
Activation of Traits by the Player moves the die up 1 die type.  d6 -> d8
Activation of Traits by an Opponent moves the die down 1 die type. d10 -> d8
Convictions add +1 die to the roll for each Conviction that applies.
Players roll their final dice pools and add the value of the dice and the pluses.

Example:

Player A rolls 1d6+6 and get results of 4+6.  Total of 10.
Player B rolls 2d6+4 and gets results of 2+3+4.  Total of 9.
Player A wins and narrates results.

Questions

I know there are some statistical geniuses on these forums but I'm not one of them.  Is my proposal changing the dynamics of the probabilities significantly?  In my one playtest the current system seemed to give me the kind of feel probability-wise that I was aiming at.  I'm not sure I'd want to really screw it up too much.

Does the Proposed system seem simpler to others, or is it just me?  There's less dice counting, I think.  In the playtest we had pools of 7 and 8 dice due to grouping a couple of times.  Is 1d6+7 easier to deal with than 7d6 vs a TN of 4 to other people?

Thanks for taking a look.

[Jasper]

Yeah, you are definitely changing things. First, you'd be devaluing the bonus from Convictions. Before, a Conviction was worth more than a purchased die (depending on the die size), but now it's much less. Second, you're making traits more important and generally changing the dynamic they produce. Before, the range of results didn't change based on traits; now it does: the more traits used, the higher your possible roll. Before, the average expectation (for a single die) changed only a little with different die sizes (eg. 0.5 -> 0.63 -> 0.7), now it changes much more (3.5 -> 4.5 -> 5.5).

So, all told, you're devaluing convictions, increasing the value of traits, and creating a larger range of results. Whether any of these things are good or bad is up to you -- and they could be compensated for by changing the scale of Convictions and Traits.

Adding dice up is generally slower and less well liked than simply counting them (as with a target number) so I don't think you gain anything there.


Why do you want differing die sizes at all? Why not, as an alternative, change the target number? This would give traits about the same kind of weight and wouldn't mess with the distribution at all. It also cuts down the search-time of finding differing dice to roll. Everyone's very familiar with target number systems, and since you already have one (as it stands) why not make a little more use of it? (Or are you already using it somehow and I missed it?)

[Andrew Cooper]

Jasper,

Thanks for the reply!  Definately some food for thought.

Not to be argumentative but could you explain how the Proposed change would devalue Convictions?  I thought it might actually make them more powerful than they were.  If I can use a Conviction I get to add the value of another whole die to my score.  That's a potential increase of 4,6,8,10 or 12 points.  Whereas with a Trait I only get a potential gain of 2 points per Trait.  Can you show me where I'm screwing this up so I know what I'm talking about?

Your question about why not just use a standard dice pool and vary the Target Number is a good one.  Mostly it is just an aesthetic choice of mine.  I like the different die types.  I noticed that unless one is playing D&D one rarely uses the d4, d8 or d12.  Everyone seems to like d6's and d10's for some reason.  I figured I'd try to come up with an interesting mechanic that actually used all those wierd dice that gamers have lying around.  I also really, really liked the basic Alternity dice mechanic and I wanted to do something like that without being a complete rip-off.  I'm not sure that's a great reason to design the system like it is but it's really the only reason I have. :-/

[Walt Freitag]

There are differences, but not exactly as Jasper has characterized them. The change in expectation for different die sizes (0.5 -> 0.63 -> 0.7) vs. (3.5 -> 4.5 -> 5.5) is actually pretty much the same, relative to the overall scale of the outcome. For instance, 0.5 -> 0.625 (d6 to d8 under the old system) is a x1.25 increase, and 3.5 -> 4.5 (d6 to d8 under the new system) is a x1.29 increase. The expectations for Convictions also works out about the same, relative to the overall scale.

The main difference is the reliability of the effects of traits. When rolling, six dice or more in the old system, it becomes more likely than not that going from d6s to d8s will increase the number of successes. (How likely? 55% of the time for six dice, 66% of the time for eight dice.) When rolling just one die, going from a d6 to a d8 increases the number rolled just 25% of the time. For two dice it's 44% of the time. The average improvement (relative to the overall scale) is about the same but with the old system when there are six dice or more you get reliable small improvements compared to the new system when there are one or two dice, where you get less-reliable bigger improvements. In plain English: if you're rolling just one or two dice, when you increase the die size it's fairly likely that you'll get unlucky and roll low anyway. So if use of multiple Convictions is rare, Traits will be less effective under the new system -- that is to say, a difference in Traits between the two sides will be less likely to swing the outcome in favor of the superior Traits than it was for the old system.

As for the amount of labor in the process (the "handling time" as we call it here), it depends on how often and how many Convictions are typically used. Rolling one die and adding +6, or rolling two dice and adding them and adding +6, is probably a bit quicker than rolling six dice and comparing them each to a target number and adding the successes. But when you get to summing up three dice, there's not much advantage either way, and summing four dice or more is probably worse than rolling the pool.

- Walt

[Jasper]

I didn't read your post closely enough, Andrew. Listen to Walt.

[Mikael]

Does the Proposed system seem simpler to others, or is it just me?  There's less dice counting, I think.  In the playtest we had pools of 7 and 8 dice due to grouping a couple of times.  Is 1d6+7 easier to deal with than 7d6 vs a TN of 4 to other people?

Others have already provided some excellent statistical analysis, so I'll just go with the gut feeling - which says that:

o As you stated in the rules, I think, there is something primitively satisfying in rolling a large number of dice vs. just getting a big bonus on one die.
o Adding three or more dice is more cognitive effort than counting the successes.
o In the new, suggested approach, you somewhat obscure the concept of successes, which I would like to utilize further than just for declaring the winner (see below for this suggestion).

So, in terms of limiting the handling time, I support the props approach:

o Some sort of buckets for the different types of dice.
o Definitely something else than dice as Power markers.
o An easily printable sheet to place the Power markers on, with one half for each side of the conflict, and clear markers places for dice bought and type of die.

Suggestion for using the successes generated in the conflict:

o Buy off opponent's successes (so the basic assumption would be that you buy off all their successes and are left with the remainder to use as you please).
o Cause damage to the opponent.
o Remove damage from yourself or the opponent. If you have escalated to damage during the buying process, you HAVE to try to cause damage.
o Change a trait for yourself (1 success).
o Buy a trait for yourself (2 successes).
o Change or buy a trait for your opponent (3 successes).
o Change a Conviction (2 successes).
o Buy a Conviction (3 successes).

If you left your opponent with successes, they can then do the same with those. (This might be too easy to exploit between co-operative players, though, so perhaps we just say that the winner gets to use the successes she had in excess of the loser, and loser always loses. But I do like the opportunity for leaving some of the opponents successes alone, in order to buy that trait you've been waiting for - and then hoping that the opponent can do you no significant harm with the successes you left him. Hmm, perhaps just rule that this option is available only in conflicts where one side is the GM?)

This idea begs for the current resolution system, since you can then just easily remove the success dice from the pile as you use them.

I have an idea for the damage mechanic, but perhaps it is too influenced by TSOY and DitV. You might also dislike damage as a principle, or have a clear mechanics idea in mind already.

Cheers,
+ Mikael

[Andrew Cooper]

Others have already provided some excellent statistical analysis, so I'll just go with the gut feeling - which says that:

o As you stated in the rules, I think, there is something primitively satisfying in rolling a large number of dice vs. just getting a big bonus on one die.
o Adding three or more dice is more cognitive effort than counting the successes.
o In the new, suggested approach, you somewhat obscure the concept of successes, which I would like to utilize further than just for declaring the winner (see below for this suggestion).

I think I agree with everything you just listed.  I love dice pools and the sensation of the dice hitting the table.  Not everyone feels this way, of course.  But I like it.  I'm thinking about putting the Proposed System in as an alternate dice mechanic like they do at the end of the rules in FATE.  This way those players that aren't big into large numbers of various types of dice can adjust the game to their tastes a bit.

I have an idea for the damage mechanic, but perhaps it is too influenced by TSOY and DitV. You might also dislike damage as a principle, or have a clear mechanics idea in mind already.

I'd love to hear your idea.  Feel free to PM me or email me.  I don't think it would fall within the scope of this particular thread.