Anyone ever used Bell Curve Rolls?

[tedcahill2]

I'm interested in using the bell curve variant from Unearth Arcana, rolling 3d6 in place of 1d20.

Unlike in UA, where they suggest to increase the crit range from 20 to 16-20 to keep it at roughly 5%, I plan to require an 18 but make them more devastating.

Anyway, does anyone have experience using 3d6 in place of 1d20?

[The_Jette]

Out of curiosity, what makes this system worthwhile? It would still give an average of 10.5, and drop the crit rate in either direction from 1 in 20 to 1 in 316.

[Jay R]

Easy things get easier; hard things get harder.

The biggest effect it will have is that differences in skill level, especially in opposed rolls, will be much more important, since you will roll 9-12 almost half the time (104/216 = 48%).

With a skill mod of 9, you will make a DC15 skill roll 93.5% of the time, a DC20 skill roll 50% of the time, and a DC25 skill roll about 6.5% of the time, rather than 75%, 50%, and 25%.

Similarly, depending on the attacker's BAB, a +4 increase to your armor class could protect you up to 48% more often, if it changes needing a 9 to needing a 13.

[Nifft]

Dungeon World uses 2d6+Stat, and success is 7+, so it's effectively bell curve rolling.

For D&D, I'd suggest trying out 2d10 instead of 3d6 -- a crit is 2 points higher (and still a natural 20), and the low end is 1 lower, so you've got a bell curve centered close to the same point, but it's not quite as steep. The expected value is 11 (instead of the d20's 10.5 expected value), but you're presumably planning to change up some of the DCs anyway, since you're deliberately changing the probability distribution.

[Florian]

Not for D20, no. I´m also playing L5R and that has a very good bell-curve mechanic, but the core system is also geared towards more than binary (pass/fail) checks. So I wouldn't´t recommend trying it for D20.

[Zanos]

Out of curiosity, what makes this system worthwhile? It would still give an average of 10.5, and drop the crit rate in either direction from 1 in 20 to 1 in 316.
http://anydice.com/program/508

1d20 each dice face is equally likely. 3d6 clusters around average results, creating a bell curve where extremes are less likely.

If you like to emphasize character skill over random chance it's a pretty good system, but you'll have to use some duct tape and elbow grease to make it work in certain spots.

[emeraldstreak]

Extensively in other systems, and my conclusion is they offer inferior gameplay to the d20 swing.

[The_Jette]

http://anydice.com/program/508
1d20 each result is equally luckily. 3d6 clusters around average results, creating a bell curve where extremes are less likely.

Yes, I understand that. The question was why that would be a good thing, since it essentially nullifies critical hits. Also, how would weapons with bigger crit ranges be handled?

[Zanos]

I edited my post while you were quoting me. If results cluster around the average, skill matters more and the system "swings" less than with a 1d20.

[The_Jette]

I edited my post while you were quoting me.

I see that. Thank you for clearing that up. Seems like a system built around preventing critical successes/failures. That's not bad. It's just a little bland, imo.

[Florian]

Yes, I understand that. The question was why that would be a good thing, since it essentially nullifies critical hits. Also, how would weapons with bigger crit ranges be handled?

You´d have to change the whole system to work with it, then you´d have some gain. A high crit range weapon would mean you´d have to modify it in a way that, say, using a scimitar would mean "No die result below 3". Same with most DCs, you´d have to keep them on a static level, and so on.

[tedcahill2]

I see that. Thank you for clearing that up. Seems like a system built around preventing critical successes/failures. That's not bad. It's just a little bland, imo.

Well, the system as outlined in UA states that critical hits should be scaled, so a 16-18 (collectively a 4.63% chance) should all threaten a critical on a weapon that normally has a 20 critical range. You also can't take 20, but you can take 16 which takes 10 times as long, or take 18 which takes 100 times as long.

As mentioned, the system clusters rolls into an average, with about 50% of all rolls falling between 9 and 12. It puts a heavier emphasis on raw skills and stats than luck. I was planning on coupling this variant with the injury system (in place of hit points), so having a critical failure on 1 in every 200 rolls, instead of 1 in 20, is better when a critical failure could result in you dropping to 0 hit point in a single hit.

[Elkad]

Yes. In Gurps and Traveller anyway.

The curve means a single point has massive effect at the ends of the scale. Which encourages min-maxxing even more than a flat roll does.
And the more dice you add, the worse it gets. 2d10 is kinda OK. 3d6 isn't. 5d4 would be terrible.

[Gildedragon]

I used it... For one session
My players revolted against my heresy against the sacred d20. This was a group that when I joined them would roll for stats using 1d20

[heavyfuel]

I've tried once with everything except Attack Rolls. They still used d20

It's pretty great, actually. Would definitely recommend for more serious games, where you don't want an unlucky die to screw you up. Stuff like "The goblin resisted your dungeoncrasher's bullrush" just doesn't happen anymore, and high base numbers become so much more important (making feats like Iron Will actually pretty good pickups)

[Jay R]

I've tried once with everything except Attack Rolls. They still used d20

It's pretty great, actually. Would definitely recommend for more serious games, where you don't want an unlucky die to screw you up. Stuff like "The goblin resisted your dungeoncrasher's bullrush" just doesn't happen anymore, and high base numbers become so much more important (making feats like Iron Will actually pretty good pickups)

I'd analyze this idea a little further. In situations in which the higher ability virtually always wins, like opposed strength rolls, use the 3d6. In situations in which extreme swings are possible, like knowledge checks, use the d20. [It's quite possible to know a great deal about a subject and still not know the one fact needed right now. Or for the person who knows the least to still have the one fact currently needed.]

[tedcahill2]

I'd analyze this idea a little further. In situations in which the higher ability virtually always wins, like opposed strength rolls, use the 3d6. In situations in which extreme swings are possible, like knowledge checks, use the d20. [It's quite possible to know a great deal about a subject and still not know the one fact needed right now. Or for the person who knows the least to still have the one fact currently needed.]

This sounds interesting. Where do you think a d20 would apply and where would 3d6 apply.

Knowledge checks are a great example of where a d20 is a good choice, for the reasons you mentioned, but it seems nearly everything else would make more sense to use 3d6, where skill is more important than chance.

[The_Jette]

This sounds interesting. Where do you think a d20 would apply and where would 3d6 apply.

Knowledge checks are a great example of where a d20 is a good choice, for the reasons you mentioned, but it seems nearly everything else would make more sense to use 3d6, where skill is more important than chance.

That's actually a great question that leads to something else: class features or feats that let you swap out the 3d6 for 1d20. It would allow for a situation that normally would make skill more important than luck change for that one person. A fighter who relies on his luck manages to defeat far superior opponents because he keeps getting high rolls, instead of relying on the middle ground. That would completely change the dynamics of luck in a game.

[Mendicant]

You could also use it as a middle ground between taking ten and just rolling for it, if some amount of random chance still seems appropriate.

One other meta thing to keep in mind: probabilities are a lot harder to do mentally with 3d6. A d20 or percentile roll is much more intuitive for a player who needs to know how risky a course of actuon might be.

[johnbragg]

That's actually a great question that leads to something else: class features or feats that let you swap out the 3d6 for 1d20. It would allow for a situation that normally would make skill more important than luck change for that one person. A fighter who relies on his luck manages to defeat far superior opponents because he keeps getting high rolls, instead of relying on the middle ground. That would completely change the dynamics of luck in a game.

That's how I see rogue-types.

Characters in a D&D-verse exist as a set of interrelated numbers, a little like the people in the MAtrix. Inspired by Haley's speech in Azure City where she deduced that none of the Xykons were the real Xykon, the rogue cheats at reality. So when it's to the rogue's advantage, he or she can roll 3d6 instead of d20. In fact, plagiarizing the Bard Spell points table, when she absolutely needs to make an impossible roll, the rogue can roll 4d6 or 5d6 or heck 12d6 if he wants to burn 17 points.