tweaking the resolution and ranges?

Has anyone tweaked the resolution system at all for this game? For example, expanded the range of partial success, raised the threshold of total success, used different dice with different ranges (i.e. 2d8 or 1d12). If so, how did it work for you?

I'm toying around this system for a campaign for my group. Two things that I am focused on are the partial success element and the skill range element. Specifically, I think partial successes are far more interesting and am considering broadening the range. As for skill ranges, this is for a longer term campaign I'm concerned that the players will be all but immune to failure once they get to higher skill levels (i.e. eventually a +5).

Has anyone hacked their system along these lines? I'd appreciate any input. Thanks.

The short range of outcomes can be limiting in terms of differentiating between character abilities, e.g. anything more than a +3 bonus is almost a guarantee of non-failure. If you're really keen on extending the range, try upping the dice size first, replacing the d6s with d12s for example: that gives you a lot more room in which to play around with the results and have larger bonuses as you describe.

The downside is that it drastically waters down character competence: a +1 bonus is only @25% as significant if you're rolling 2d12 instead of 2d6, so stat increments are less of a reward than new moves.

The alternative to extending the range of outcomes is to give the characters tactical options to change one type of success into another, e.g. if they accept a lesser result in advance, then any miss becomes a 7-9 instead. I think one of the hacks uses this option to add skills to the system, so that when you are using a skill, you never miss, you just get a 7-9 result instead. If that seems too effective, then a skill could extend the mid-range when used, making it 6-9, 5-9, 6-10 or whatever.

Yeah I worked up some probability charts for my game along these lines. Here's the probabilities for AW + a "roll 3, 4, 5, pick 2 highest" system:

===

Probability of failure (6-), partial success (7-9), success (10+)

2d6-2: 72.22% 25% 2.78%

2d6-1: 58.83% 33.33% 8.33%

2d6: 41.66% 41.66% 16.66%

2d6+1: 27.77% 44.46% 27.77%

2d6+2: 16.66% 41.66% 41.66%

2d6+3: 8.33% 33.33% 58.33%

---

3d6, 2 high: 19.44% 44.91% 35.65%

4d6, 2 high: 9.03% 38.81% 52.16%

5d6, 2 high: 4.22% 30.62% 65.16%
===

Eventually I settled on 3d6(+/- n) - firstly because you still don't need any weird dice, secondly because it's got a much bigger partial success range. My system as is has static stats (ie. You can't improve them) ranging from -1 to +2, situational and move-based advantages disadvantages ranging from -2 to +2. Feel free to steal it!

===

Probability of failure (9-), partial success (10-13), success (14-17), critical success (18+, triple ones)

3d6-3: 73.61% 24.08% 1.85% 0.46%

3d6-2: 62.04% 32.87% 4.63% 0.46%

3d6-1: 49.54% 40.74% 9.26% 0.46%

3d6: 37.04% 46.3% 15.74% 0.92%

3d6+1: 25.57% 48.14% 23.98% 2.31%

3d6+2: 15.74% 46.3% 32.87% 5.09%

3d6+3: 8.8% 40.74% 40.74% 9.72%

3d6+4: 4.17% 32.87% 46.3% 16.66%
===

I did all this with anydice

The short range of outcomes can be limiting in terms of differentiating between character abilities, e.g. anything more than a +3 bonus is almost a guarantee of non-failure. If you're really keen on extending the range, try upping the dice size first, replacing the d6s with d12s for example: that gives you a lot more room in which to play around with the results and have larger bonuses as you describe.

The downside is that it drastically waters down character competence: a +1 bonus is only @25% as significant if you're rolling 2d12 instead of 2d6, so stat increments are less of a reward than new moves.

The alternative to extending the range of outcomes is to give the characters tactical options to change one type of success into another, e.g. if they accept a lesser result in advance, then any miss becomes a 7-9 instead. I think one of the hacks uses this option to add skills to the system, so that when you are using a skill, you never miss, you just get a 7-9 result instead. If that seems too effective, then a skill could extend the mid-range when used, making it 6-9, 5-9, 6-10 or whatever.

What I'm tempted to do is go with something like 1dX rather than 2dX. With a linear distribution, I figure I can extend the value of higher skills while also being able to avoid the whole auto-success thing.