While I love the pathfinder system and the horrifying amount of rules and feats, the major thing I find many other RPG's does better is the use of several dice for rolls instead of one.
By using the single d20, attack and skill rolls vary so greatly that it breaks immersion and results in the game depending more on good rolls than anything else.
I've thought of using two d10's instead of the one d20, but I'm wondering how that will affect critical hits/failures.
One could design new critical threats with the same chance as before, but that would become rather complicated when we reach crit threats such as 17-20.
My idea as it stands is to use the two d10's' result as the "hit" roll, and make the d20 the "crit"-die. As in, you roll two d10's and one d20 for every attack/save roll.
A natural 1 or 20 on the d20 are always miss or hit, respectively.
But when you critically threat you don't make a confirmation roll. If you have hit with the result of your two d10's then you automatically crit if you roll your crit range with the d20.
What I can't convert is a bonus to confirmation rolls. Like, would that express itself as a bonus to the crit-die (d20) so that you will crit more often?
*****
Thoughts, criticism?
This has been discussed in detail before. You can probably find threads about it if you search. The issue is that it makes it much harder to overcome DCs if the odds are against you.
A discrete Gaussian distribution won't solve everything.
2d10 had side effects, because the average result of a roll goes up to 11. This gives a small advantage to whoever is rolling, which will add up over time because you're giving out half a point of bonus on more than half of all values.
3d6 is the traditional alternative, though it's an even more extreme change to the game. The old 3e Unearthed Arcana has a threat range conversion table for it.
Both options result in a +1 bonus being more valuable than before, and everything that implies.
3d6 is the preferred choice, as the average is always 10.5, just like with a d20.
This makes combat a lot less swingy. The first attacks in a full attack almost never miss, and the last ones almost never land, if the opponent has any relevant AC
The conversion table for crits is found here:
http://www.d20srd.org/srd/variant/adventuring/bellCurveRolls.htm
Who said that?
It depends on what you're asking exactly.
By preferred I mean compared to 2d10. 2d10 results in a slightly higher average (11 instead of 10.5) which is a slight change to the average game math
If you're asking who said the average is 10.5, that's just math
| 2 people marked this as a favorite. |
I don't understand people who dislike the d20 swing, it's exciting and wholely unique to tabletop.
why beat around the bush with 3d6\2d10
just take 10 on everything. this is what you always secretly wanted.
| 2 people marked this as a favorite. |
So this article from theangrygm.com is pretty topical. It's a great read about probability math in tabletop games and messing around with it:
Just a warning, the AngryGM's adopted tone can be somewhat off putting and his articles are long, but they are very informative.
@Cyrad
Ah, yeah. After switching around with keywords I found one thread named "The Twenty-Sided Die" that revolves around the subject.
I realize the implications of using a Gaussian distribution, with the median becoming more and more expected the more dice you add. However, if you had to use a distribution, would you consider 2d10 the optimal choice?
*****
@Mortuum
I hadn't considered that static bonuses would increase in value, so that is something I will have to deal with.
*****
@Lanitril
While I like the idea of a choice between 2d10 and d20, the group I'm playing with will probably all choose 2d10 anyway.
*****
@D@rK-SePHiRoTH-
While the average with 3d6 is the same, the average also becomes a lot more expected by using 3 dice. That would probably affect the math of pathfinder more that +0.5 in the favour of the one rolling.
*****
@Jader7777
My group have played a whole lot of different RPGs, and pathfinder has unfortunately been left behind in favour of games where there is some sort of Gaussian distribution at work. That's why I'm trying to "steal" the main difference between pathfinder and those games to make it more attractive once again.
Being the one guy who rolls under 8 the whole session can really bum you out.
2d10 had side effects, because the average result of a roll goes up to 11.
It goes up to 11! That alone confirms 2d10 as the most awesome system. Nothing beats going up to 11!
I find that the d20 only matters too much at lower levels, it's quickly diminishes in significance.
You could call that 'working like intended'. At the begin you play an adventurer who is more or less at the mercy of chance, but later on his abilities will become more reliable. Which gives you a feeling of achievement - you wouldn't get this if you would start out being excellent.
Further, a d20 gives you a better chance to succeed against really bad odds. Want to escape a grapple as a low level wizard? Want to hit that overwhelming boss (with thick plot armor)? Want to attempt a desperate climb to save a mate? You are better off with a d20 here, because an average roll won't help you.
-----------------
Ok, this is the suggestions section, so I tried to find an alternate way to d20. I found nothing convincing, so I propose a different, already existing approach: Hero points. They don't change the type of dice rolled (avoiding a lot of side effects), they partially help against bad rolls and they are rather simple (for Pathfinder standards).
How about Double 1s through Double 5s being a potential fumble, and an auto-miss
---) If the roll total is a miss as well, the fumble is confirmed. (vs DC 10 recommended)
-----) Double 1's is always a fumble
Double 6s through Double 0s are a potential critical hit, and an auto-hit
---) If the roll total is also a hit, the critical is confirmed
-----) Double 0's is always a critical
Everything is covered with a single roll.
@THUNDER_Jeffro
While I have knowledge of the subject, It's always nice to have a fresh reminder and charts of the suject at hand.
*****
@SheepishEidolon
By using 2d10's and one d20, I've managed to give a slight boost to the probability of succeeding against bad odds, As you can both succeed with the 2d10's and by rolling a 20 on the d20.
Req Roll : . d20 : 2d10 + d20
. . . . 20 : 5% . : 6% (1+5)
. . . . 19 : 10% : 8% (3+5)
. . . . 18 : 15% : 11% (6+5)
. . . . 17 : 20% : 15% (10+5)
. . . . 16 : 25% : 20% (15+5)
. . . . 15 : 30% : 26% (21+5)
. . . . 14 : 35% : 33% (28+5)
. . . . 13 : 40% : 41% (36+5)
At 14 and higher, you're better off with the d20. With the exception of 20, which has a higher chance of occuring.
The easiest way to allow players to succeed against bad odds would be to make the 2d10+d20 optional, with players using simply the d20 for rolls they think they'll have a hard time beating. But then again, the "new" odds aren't that far away at higher target rolls.
Oh, and we play with Hero Points. I've noticed that most players don't use them if they got 2 or lower. (Since 2 will allow you to cheat death)
*****
@Daw
The idea of placing higher significance on double rolls is great, but I'm having trouble converting threat ranges to the same concept.
@Daw
The idea of placing higher significance on double rolls is great, but I'm having trouble converting threat ranges to the same concept.
I assume you are asking about weapons with expanded threat ranges.
How about either dice rolling a 10 being a threat for 19-20 and 18-20 weapons/attacksAnd either dice rolling a 9 or 10 being a threat for 17-20 and wider threat range weapons/attacks.
The percentages aren't too far off and it is still quick and easy.
How about either dice rolling a 10 being a threat for 19-20 and 18-20 weapons/attacks
And either dice rolling a 9 or 10 being a threat for 17-20 and wider threat range weapons/attacks.The percentages aren't too far off and it is still quick and easy.
With all due respect, I think this skews the critical threat percentages too far to remain fair.
So, for our basic d20, adding 1 to the crit threat increases the possibility of a crit roll by 5% (20=5%, 19-20=10%, 18-20=15%, 17-20=20%).
With your proposal, we get the following spread:
20 equivalent (both roll 10) = 1% (1 possible out of 100 outcomes)
19-20 & 18-20 equivalent (1 rolls 10) = 19% (1 of 10 outcomes + 1 of 10 outcomes - 1 of 100 outcomes (both are 10))
17-20 equivalent (1 rolls >8) = 36% (2 of 10 + 2 of 10 - 4 of 100)
I think that's too steep of a curve to keep the weapons balanced without changing any other factors.
| 2 people marked this as a favorite. |
A true Gaussian d20 would be 2d6+1d8-2, just to mention.
Rawhead,
You left out the Doubles part of the Threat chance, so the base chances are 5% threat and 5% fumble.
If you think the additional threat range effect is too high, how about using a red and a white d10,
For 18-20 and 19-20 threat ranges: if the red is a 10, it is also a threat, even tho not doubles
For 17-20 or wider, if either dice is a 10, it is also a threat, even tho not doubles
So a pretty large amount of potential critical threats aren't actually going to be good enough rolls to hit their target? That is pretty screwy if I am understanding right.
How would 2d20/2 or 2d20 and double all DCs affect the odds? I guess fumbles and crits would be ridiculously rare...
Kahnya and Snowblind,
Almost no one will ever fumble, the attack/confirm/fumble-avoid roll will always be from 11 to 15. 01-05 will be 10 + 1 to 5.
Actually, a fairly high amount threats will confirm as well.
For 96-00 the hit/confirm roll will always be from 15 to 20 plus your attack bonuses.
For 91-00, you are still talking about a roll of 10+, which with a decent attack bonus will still likely crit against most foes.
Even with keen weapons with a wide crit range, your minimum unadjusted roll will be 8 or higher.
It is actually worse if you used my doubles mechanic though.
I don't think you can get away with two separate rolls, which is no worse than what we have now, really.
The things you think of when you are trying to fall asleep.
3d6 has the same average (10.5) but has a more bell-curved standard distribution.
2D10 gets rid of fumbles and nat 1 misses. Having a character roll 1 after casting true strike ruins the enjoyment of the game. A distortion of the bell curve seems a small price to pay.
2D10 gets rid of fumbles and nat 1 misses. Having a character roll 1 after casting true strike ruins the enjoyment of the game. A distortion of the bell curve seems a small price to pay.
That is assuming a 2d10 conversion doesn't include auto-fail on 4 or less and auto-success on 18 or more, which gets you roughly the same chances of hitting an auto-success/fail
Why not just remove auto-failure on a natural 1, if the auto-failure thing is the issue. It is rarely relevant, anyway, so why not? Throw out auto-success on a natural 20 while you are at it, as well. Peasants can't touch heroes and have a 0% chance of resisting a succubus, while mid level Paladins are immune to low level caster SoS and can laugh at minor minions who can't even scratch them through their armor.
I was at a convention and my ranger had the spell true strike. 2 times in a row a nat 1 ruined true strike. I buried the D20 in the bottom of a plastic trash can. Devs tried to solve the problem by ruling that only sorcerers and wizards can have true strike. There's another wrong answer to add to the list.
If a lich and several noble drow houses start cranking out rings of spell storing that have three true strike spells in them, that will give you your minimum daily requirement of game balance.
You don't need to move the autofail up to 3, you don't need to get rid of crits, and you certainly never need to ruin the game and blame the dice.
Fumbles are a variant rule or a house rule, and if you dislike auto-fails, you can just remove it.
As it is, rolling 2d10s is a massive change that's not really relevant to removing the natural 1 rule, and they're not incompatible with auto-failing on the minimum natural roll.
|
To reiterate:
All results will be more "average."
Higher DCs will be harder to acheive.
Lower DCs will be harder to miss.
"Crits" and "fumbles" will be rarer.
You should really look at Green Ronin's AGE system, the engine behind the Dragon Age RPG. It used a 3d6 mechainic, but had some interesting combat maneuver-esque things you could do when you roll 2 6's on two particular dice.
-Skeld
I don't know if I simply made it too unclear, but I have no intention of lowering the risk/chance of auto-fail and auto-success. That is still governed by the d20.
*****
Still, I've considered what has been written here and now fully know the implications of the change I'm about test.
So I'll just do what Jader7777 proposed and take 10 on everything. Hopefully, that'll decrease the amount of randomness.
/JK.
I'll try to keep everything in mind and test out the 2d10+d20. With some adjustment here and there I should be able to make it work. And if not, then I'll look for other alternatives.
Thank you for all your insight.