|
5 months ago ::
Jan 16, 2013 - 4:40PM
#11
|
Date Joined:
Jul 27, 2009
|
Interesting thread. Some (random) thoughts: Using 3d6 (works less well with 2d10), instead of having a +1 or +2 or whatever to hit, just increase the number of dice but drop the lowest dice. (This is how the 4d6 method of rolling for attributes works.) Penalties work the same way, 'cept you drop the highest die. (Probably works out to about +1d6 per +2.) Some sample distributions can be found on anydice.com (wow, thanks for that link, lok_soldier): output [highest 3 of 4d6] named "highest 3 of 4d6"
output [highest 3 of 8d6] named "highest 3 of 8d6"
output 3d6 named "3d6"
output [lowest 3 of 4d6] named "lowest 3 of 4d6"
output [lowest 3 of 8d6] named "lowest 4 of 8d6"
I'm guessing that skill check results should be distributed log-normally (normally, but with a limited downside). Anybody have any idea how to roll that?
|
|
|
|
5 months ago ::
Jan 16, 2013 - 6:21PM
#12
|
|
|
So no one mentioned 2d10 instead (though it's in the initial table the OP provided)...
It's between the single d20 and the 3d6 and should add more "average" rolls without making extreme rolls that much rarer.
I could see 2d10 for skills and 1d20 for combat working. It's something worth trying out in a playtest.
Although it takes away some of the flavour of the d20 system... there's that. Some may dislike, some not care.
|
|
|
|
5 months ago ::
Jan 17, 2013 - 3:34AM
#13
|
Date Joined:
Jul 27, 2009
|
You'd also have to redefine the mechanism for crits. To me easiest is just rolling another d20 and check for 1s and 20s. Or a d10: rolling a 0 is a crit. Goodness or badness is determined if you would've hit or not.
|
|
|
|
5 months ago ::
Jan 17, 2013 - 4:43AM
#14
|
|
|
You'd also have to redefine the mechanism for crits. To me easiest is just rolling another d20 and check for 1s and 20s. Or a d10: rolling a 0 is a crit. Goodness or badness is determined if you would've hit or not.
Yeah, I wouldn't use 2d10 for combat.
As someone said, maybe combat is just more volatile.
Skills, though, don't have criticals... so maybe.
|
|
|
|
5 months ago ::
Jan 17, 2013 - 5:16AM
#15
|
Date Joined:
Oct 27, 2007
|
I'm toying with the idea to have skill checks be a 3d20 roll vs an attribute. One success out of 3 for simple checks (equivalent of low DC's), 2 out of 3 for medium difficulty, 3 out of 3 for ard, 4 out of 3 for very hard. being skilled nmeans that the difficulty shift one category (simpole checks become automaitic, medium 1 out of 3, hard 2 out of 3, very hard 3 out of 3).
The more dice you roll the less chance there is that an unskilled weakling outperforms a strong and skilled climber.
|
|
|
|
5 months ago ::
Jan 17, 2013 - 5:33AM
#16
|
Date Joined:
Jul 27, 2009
|
Yeah, I wouldn't use 2d10 for combat. As someone said, maybe combat is just more volatile.
Agreed. Maybe there can be a simple rule where if something takes less than 10 seconds to do than you must use 1d20. But if something takes longer than that (say, a Disguise to keep in Character, or Knowledge[whatever] to figure something out) then you should use something which more normally distributed (Median of 3d20 sounds good, and works stylistically). This idea was already present in v4.0 with skill contests to the point that the outcomes were almost deterministic.
Another possibility to get closer to a normal distribution is to get rid of fixed modifiers and use dice in their place. So rather than 1d20 + 5 to hit, you could use 1d20 + 1d10. This is pretty close to how skill dice work in the current iteration of Next. (Though I think the idea was pineered by FASA's Earthdawn RPG).
|
|
|
