UA: Why 3d6 for the "Bell Curve" variant, instead of 2d10?
- Thread starter Azlan
- Start date
DonAdam
Explorer
The 3d6 seems to be a smoother conversion of d20.
However, I'll only be using it for ability checks (where the modifiers are relatively small). I prefer combats to be wild and less predictable, but I don't like it when Newmoon the elf rips the gate off its hinges that Rogar the barbarian could not.
However, I'll only be using it for ability checks (where the modifiers are relatively small). I prefer combats to be wild and less predictable, but I don't like it when Newmoon the elf rips the gate off its hinges that Rogar the barbarian could not.
Here's something I posted on another board to someone who was excited about switching to 3d6:
"I grew up playing TFT (proto-GURPS) and was excited to see the 3d6 idea in UA. I did some playtesting, and you have to do a lot of tweaking of DCs with 3d6 to make it work well for D&D. I'm going to approach my group about trying 2d10 for our next campaign. You get a nice pyramid distribution instead of a bell. Here's how the two breakdown:
3d6
Roll # %
3 1 0.5%
4 3 1.4%
5 6 2.8%
6 10 4.6%
7 15 6.9%
8 21 9.7%
9 25 11.6%
10 27 12.5%
11 27 12.5%
12 25 11.6%
13 21 9.7%
14 15 6.9%
15 10 4.6%
16 6 2.8%
17 3 1.4%
18 1 0.5%
2d10
Roll # %
2 1 1%
3 2 2%
4 3 3%
5 4 4%
6 5 5%
7 6 6%
8 7 7%
9 8 8%
10 9 9%
11 10 10%
12 9 9%
13 8 8%
14 7 7%
15 6 6%
16 5 5%
17 4 4%
18 3 3%
19 2 2%
20 1 1%
I'd do bows, etc. with a threat of on 18-20 (6% instead of 5%) with Longswords critical threat on 17-20 (10% vs. 10%), Scimitars critical threat on 16-20 (15% vs. 15%), Keen Longswords critical threat on a 15-20 (21% vs. 20%). It works out pretty nicely.
For skill checks I'll probably have 18-20 "explode" to a 2nd roll, adding to the first, with 2-4 "imploding" to a 2nd roll subtracting from the first.
I really like the idea of highly skilled PCs not getting hosed in opposed contests very often by the fact that a "2' comes up just as often as an "11". Sure if you roll a d20 a zillion times it settles on an average of 10.5, but when you're role-playing and your STR 18 Barbarian with rippling muscles loses an arm-wrestling match to a STR 8 dairy maid just because she rolled a "15" and you rolled a "5", it can really break a mood. Sure, it can happen with a bell or a pyramid, but not as often. With a flat distribution, that rogue with 8 ranks in "Move Silently" can say, "Well, yeah, that sleepy 1st level guard heard me *this* time, but that's just because I rolled a "2" on *average* he wouldn't have heard me," but he can't take much comfort in that - the alarm is raised...
It's also cool how every skill rank, BAB point, 1 on your weapon, and a simple flanking maneuver really seem to matter more with a distribution that isn't flat. Soooo... frustrating sometimes when you get the drop on a foe due to a well executed plan, are flanking, under the effects of a Bless, etc. and *still* have a, say 5-15% chance of failing (1-3)
3d6 might be a little to radical to start out with, though - D&D just isn't set up to handle it very well maybe. It really does hurt the balance for "the little guys" with 3d6 particularly. All those kobolds were really hoping to get a few licks in on that armored Fighter who's got Combat Expertise, by rolling a few 17s,18,19, and 20s (20%ish). It's just not going to happen with the bell - they're screwed. That's not *necessarily* a bad thing if they only have a 4.6% chance 3d6's 'natural 20' of 16,17,1 of hitting or a 6% chance (2d10's 'natural 20' of 18,19,20) of hitting, but the DM has to really keep their eye on the balance and design encounters accordingly *AND* not let the party's AC get too far apart. Something that is going to hit the party mage *every single time* on a bell might not be able to touch the party cleric if he invests heavily in magical defense. That can be a concern with d20 too - it's just aggravated by the 3d6 bell even more.
Anyway, be aware of the pitfalls. I'm going to try it out with 2d10 and see how it goes. Good luck to you!"
*****
I don't understand this comment in this thread, Unseelie (cool name!):
"A two die bell curve is too steep in general, which means that bonuses and penalties have too much of an effect. A 3 die bell curve is much more natural."
The problem that swrushing noted of bonuses being different depending where you fall on the curve is even more pronounced with 3d6 than with 2d10; a +1 sword could give you a 0.5% boost, or a 12.5% boost. In any case, I don't find that argument very persuasive becuase I often feel, as noted above, that your bonuses don't mean *anything* with a flat distribution - "I'm playing the role of a very skillful tracker, and invested in Survival ranks to reflect that, but 15% of the time, I couldn't succeed at a relatively mundane task just because I rolled a 1,2, or 3." Joy.
I like d20 the way it is, but I want to try something different with the dice mechanics and see how it feels for one campaign.
"I grew up playing TFT (proto-GURPS) and was excited to see the 3d6 idea in UA. I did some playtesting, and you have to do a lot of tweaking of DCs with 3d6 to make it work well for D&D. I'm going to approach my group about trying 2d10 for our next campaign. You get a nice pyramid distribution instead of a bell. Here's how the two breakdown:
3d6
Roll # %
3 1 0.5%
4 3 1.4%
5 6 2.8%
6 10 4.6%
7 15 6.9%
8 21 9.7%
9 25 11.6%
10 27 12.5%
11 27 12.5%
12 25 11.6%
13 21 9.7%
14 15 6.9%
15 10 4.6%
16 6 2.8%
17 3 1.4%
18 1 0.5%
2d10
Roll # %
2 1 1%
3 2 2%
4 3 3%
5 4 4%
6 5 5%
7 6 6%
8 7 7%
9 8 8%
10 9 9%
11 10 10%
12 9 9%
13 8 8%
14 7 7%
15 6 6%
16 5 5%
17 4 4%
18 3 3%
19 2 2%
20 1 1%
I'd do bows, etc. with a threat of on 18-20 (6% instead of 5%) with Longswords critical threat on 17-20 (10% vs. 10%), Scimitars critical threat on 16-20 (15% vs. 15%), Keen Longswords critical threat on a 15-20 (21% vs. 20%). It works out pretty nicely.
For skill checks I'll probably have 18-20 "explode" to a 2nd roll, adding to the first, with 2-4 "imploding" to a 2nd roll subtracting from the first.
I really like the idea of highly skilled PCs not getting hosed in opposed contests very often by the fact that a "2' comes up just as often as an "11". Sure if you roll a d20 a zillion times it settles on an average of 10.5, but when you're role-playing and your STR 18 Barbarian with rippling muscles loses an arm-wrestling match to a STR 8 dairy maid just because she rolled a "15" and you rolled a "5", it can really break a mood. Sure, it can happen with a bell or a pyramid, but not as often. With a flat distribution, that rogue with 8 ranks in "Move Silently" can say, "Well, yeah, that sleepy 1st level guard heard me *this* time, but that's just because I rolled a "2" on *average* he wouldn't have heard me," but he can't take much comfort in that - the alarm is raised...
It's also cool how every skill rank, BAB point, 1 on your weapon, and a simple flanking maneuver really seem to matter more with a distribution that isn't flat. Soooo... frustrating sometimes when you get the drop on a foe due to a well executed plan, are flanking, under the effects of a Bless, etc. and *still* have a, say 5-15% chance of failing (1-3)
3d6 might be a little to radical to start out with, though - D&D just isn't set up to handle it very well maybe. It really does hurt the balance for "the little guys" with 3d6 particularly. All those kobolds were really hoping to get a few licks in on that armored Fighter who's got Combat Expertise, by rolling a few 17s,18,19, and 20s (20%ish). It's just not going to happen with the bell - they're screwed. That's not *necessarily* a bad thing if they only have a 4.6% chance 3d6's 'natural 20' of 16,17,1 of hitting or a 6% chance (2d10's 'natural 20' of 18,19,20) of hitting, but the DM has to really keep their eye on the balance and design encounters accordingly *AND* not let the party's AC get too far apart. Something that is going to hit the party mage *every single time* on a bell might not be able to touch the party cleric if he invests heavily in magical defense. That can be a concern with d20 too - it's just aggravated by the 3d6 bell even more.
Anyway, be aware of the pitfalls. I'm going to try it out with 2d10 and see how it goes. Good luck to you!"
*****
I don't understand this comment in this thread, Unseelie (cool name!):
"A two die bell curve is too steep in general, which means that bonuses and penalties have too much of an effect. A 3 die bell curve is much more natural."
The problem that swrushing noted of bonuses being different depending where you fall on the curve is even more pronounced with 3d6 than with 2d10; a +1 sword could give you a 0.5% boost, or a 12.5% boost. In any case, I don't find that argument very persuasive becuase I often feel, as noted above, that your bonuses don't mean *anything* with a flat distribution - "I'm playing the role of a very skillful tracker, and invested in Survival ranks to reflect that, but 15% of the time, I couldn't succeed at a relatively mundane task just because I rolled a 1,2, or 3." Joy.
I like d20 the way it is, but I want to try something different with the dice mechanics and see how it feels for one campaign.
Matthew Gagan said:
This is a logical error.
A d216 has just as much chance of rolling a 1-1-1 as a 3-4-3, just as a d10 has the same chance of rolling a 3 as an 11.
How likely you are to succeed in either case depends on how many of those potential rolls you assign "succeeds" and how many you assign "fails".
if rolling 1d20 you chose to say "1-3 fails" you are saying "you will succeed 85% of the time on average." On any given roll, the outcome is not certain... because you said so.
if rolling 3d6 you chose to say "you will succeed 85% of the time" you will assign a fail range of 3-7 and the outcome will, for any given roll, be uncertain.
What is happening when you say "roll 3d6, you fail on a 3" is that you are saying the chance of success is NOT 85% but 99+%. In d20 you would have frankly just said "you succeed" and not needed a roll.
3d6 wont produce more predictable "results" in terms of success fail than d20 will unless you start assigning greater chances of success for 3d6 than for the d20. Simply put, if you assign a d20 task a 15% chance of failure and the same task under 3d6 a 5% chance of failure, its not the DICE that are making you more likely to succeed under 3d6, its the greater chance of success you CHOSE!
swrushing said:
No.
You misunderstood me. I probably did not state it very clearly.
I'm not clear on your "d216" or your d10 rolling "11s", but I think I know what you are trying to say, and you are arguing against a point I wasn't trying to make.
I wasn't describing a die roll mechanic in my d20 ranger example where you predetermine that a roll of 1,2, or 3 (and therefore 85%) equals automatic failure. Leave whatever "auto-success" number you choose, be it "20" (5%) for d20 or "16,17, and 18" (4.6%) for 3d6 , or "18,19,20" (6%) for 2d10 aside for a moment. I am perfectly happy for those to be relatively equivalent %s regardless of mechanic.
You don't "assign" likelyhoods of success in a role-playing game in a static state (or shouldn't). Success is relative to character skills and situational modifiers, not just task difficulty.
I just like the feel of a skill determination system that recognizes a greater measure of repeatability. You roll 9s, 10s, 11s, and 12s more often with 3d6 or 2d10 than with 1d20. It allows the skill ranks to play more of a determination than the roll.
As I mentioned, I don't have a problem with d20 as it is; I understand probability and the impacts of the different mechanics. I just want to try something different and see how it goes.
