2d10 for 1d20: Repercussions, Variants, and MAGIC

[Tinkermancer]

Greetings esteemed colleagues and fellow enthusiasts! LOOOOONG time lurker here, posting for the first time to hopefully glean some useful information from those of you with FAR more experience than myself in the hopes of improving my game. I LOVE 5E, having happily upgraded from 3.5 about six months ago, and after weekly sessions as a player in a friend's game, I'm ready to, once again, take up the mantle of DM and teach my friends a thing or two about DRAMA. :)

I've always liked the 2d10 variant of 3.5, where you replace the d20 roll with 2d10 in order to decrease the swingyness of the game, but it wasn't THAT big of deal in a game where getting bonuses became easier and easier and roll modifiers larger and larger with each level. I can't say I've been unhappy with the lower numbers inherent to 5E, but it HAS proven to be far swingier. My first thought was that the 2d10 variant would help in that area, but not without repercussions.

Crits - On a d20, crits happen 5% of the time and only on a 20. On 2d10, if crits only happen on a 20, that means a crit only happens 1% of the time. If we move it to 19 and 20, it increases to 3%. Still, not quite there. If we add 18 to the mix, it moves to 6%. That's better.

There are probably more, and I'm bringing it here to hopefully encourage someone with a better mind for numbers than myself to show me where I'm going wrong lol. BUT ALSO...I feel like this variant encourages some interesting CREATIVE options, and I'd like some help with that!

Variants - What if instead, crits happen when one of the dice show a 0? That means a crit will happen 5% of the time (I think), and we can make 00 be a crit with max damage for extra fun! EDIT - Another poster pointed out that this is actually 19%...so nope. Not what I was looking for!

Procs - Certain magic items (weapon, armor, even other!) can be set up with procs assigned to certain numbers. I was always hesitant to do this on d20, because I felt like it would happen too rarely if assigning just one number, and if assigned to the wrong number, would only happen during a crit, or only happen when you fail to connect. With 2d10, we could set it to, 7 or 8 let's say, and still have a decent chance of hitting without having to crit. The chance to do so also increases to 10%. ALSO ALSO, it gives us the chance for a SUPER proc on matching numbers.

Proc example: GUIDING LIGHT
This magic weapon exudes a calming yet invigorating glow when grasped. It functions normally as a +1 weapon that does 1d8 radiant damage on a successful hit, and an additional d8 to undead and fiends. When a die shows 8, the wielder is healed by an amount equal to the radiant damage shown. When the two die match, the weapon guides the user toward the target, adding another +2 hit. If the two die match AND both show 8s, the radiant damage is maxed and doubled, plus the user is healed for the radiant amount.

Not TOO much thought put into that, but I'll take some of yours!

Proc example: ARMOR OF REJUVENATION
This armor is flush with radiant energy. When an attack roll is made, if the die shows a 2, the wearer heals 1d6 damage. If both show 2s, that amount is tripled.

These are just shots in the dark, but maybe enough to get some creative juices flowing. Any other thoughts on the repercussions of the 2d10 variant, any thoughts on variations of this variant, and any ideas for magic items? I'd love to steal your ideas for my game! Thanks again!

[Arcangel4774]

I like the d20 to d10 change. Especially as it applies to skills and the everybody roll conundrum. Some people change dis/advantage to roll 3 keep the low/high 2 under this change. How does that work for balance

[Lonely Tylenol]

2d10 would break the game as-is.

Here's an AnyDice link for comparison: http://anydice.com/program/b5b3

The standard assumption, with small variations, is that you hit on an 8 or better, and the enemy hits on a 12. If you look at the create-a-monster table, monsters' AC jumps by 1 basically whenever a character with standard array that pumps their primary to 16 at level 1 and maxes the stat by level 8 *would* get an increase, so "CR-appropriate" always hits around 8. To-hit bonuses scale similarly, though obviously the assumption breaks down a little bit because of varied armor.

So with the current game expectations, 1d20 hits at least a 12 about 45% of the time, and hits an 8 around 65% of the time. Combined with criticals on 20s, it creates an assumption that a level-appropriate monster's DPR is about half of what's listed, and a character will do about 70%.

With 2d20, monsters still hit about 45% of the time, but any increase to AC causes this to drop sharply. Meanwhile, players hit almost 80% of the time, with no extra changes.

The game doesn't sit the standard for players and monsters at the same point. They fall at different points of the curve. It would create nigh-invulnerability for creatures that get minor bonuses above and beyond the norm, because hitting extremes on a curve is much harder than on a single larger dice.

Also: your crit numbers for "one dice is a 0" increase the chances of crit from 1% to 19. If you think of it as percentile dice, you would crit on every number that starts with a 0 (1-9), and every represented number that ends with a 0 (every multiple of 10 up to 100). 19 in total. Would be devastating, especially with the 80% chance to hit created by 2d10 already.

[Zman]

I use the 2d10 varient for skills and ability checks to take the swinginess of the dice out of it, but I caution doing it for attack rolls and saves as that change has massive unforseen effects on balance and can destabilize your game with High and Low ACs and saves.

[Lombra]

What about rolling a d100 (which actually is two d10s in many dice sets) and dividing the result by 5 (rounded up)? It would be the same as rolling a d20 but less swingy. At least I think.

[Lonely Tylenol]

What about rolling a d100 (which actually is two d10s in many dice sets) and dividing the result by 5 (rounded up)? It would be the same as rolling a d20 but less swingy. At least I think.

Exactly as swingy.

[Tinkermancer]

2d10 would break the game as-is.


The standard assumption, with small variations, is that you hit on an 8 or better, and the enemy hits on a 12. If you look at the create-a-monster table, monsters' AC jumps by 1 basically whenever a character with standard array that pumps their primary to 16 at level 1 and maxes the stat by level 8 *would* get an increase, so "CR-appropriate" always hits around 8. To-hit bonuses scale similarly, though obviously the assumption breaks down a little bit because of varied armor.

So with the current game expectations, 1d20 hits at least a 12 about 45% of the time, and hits an 8 around 65% of the time. Combined with criticals on 20s, it creates an assumption that a level-appropriate monster's DPR is about half of what's listed, and a character will do about 70%.

With 2d20, monsters still hit about 45% of the time, but any increase to AC causes this to drop sharply. Meanwhile, players hit almost 80% of the time, with no extra changes.

The game doesn't sit the standard for players and monsters at the same point. They fall at different points of the curve. It would create nigh-invulnerability for creatures that get minor bonuses above and beyond the norm, because hitting extremes on a curve is much harder than on a single larger dice.

Also: your crit numbers for "one dice is a 0" increase the chances of crit from 1% to 19. If you think of it as percentile dice, you would crit on every number that starts with a 0 (1-9), and every represented number that ends with a 0 (every multiple of 10 up to 100). 19 in total. Would be devastating, especially with the 80% chance to hit created by 2d10 already.

Thanks for the break down!

I didn't realize "one dice is a 0" increased the chance so much! That would DEFINITELY break the game. I would do better to just crit on 18-20 then, with a max damage crit on 00.

Also, I had considered only using 2d10 for the players, keeping the d20 for everyone else for unpredictability and lack of training. I would probably use 2d10 for trained NPCs.

Furthermore, I am considering not only for its own sake, but for the procing customization I posted as well. Without that part, I probably wouldn't use it, considering your data.

[Lombra]

Exactly as swingy.

I was obviously missing something

[Lonely Tylenol]

I didn't realize "one dice is a 0" increased the chance so much! That would DEFINITELY break the game. I would do better to just crit on 18-20 then, with a max damage crit on 00.

You could also do something like "doubles crit if they hit," which jumps to about a 10% crit chance but realistically is more like 5-6%, since double 9s would hit, but double 1s wouldn't. Becomes difficult to include expanded crit ranges for Champions and Hexblades, though.


Also, I had considered only using 2d10 for the players, keeping the d20 for everyone else for unpredictability and lack of training. I would probably use 2d10 for trained NPCs.

This is the opposite of fixing the problem. The math is more or less static for the monsters and NPCs if they switch to 2d10 (about 45% to hit either way; only gets off if the PCs have bloated ACs), but swings wildly out of control if PCs switch to 2d10 in combat (from 65% chance to hit to 79%).


I was obviously missing something

So if you did 1d100/5 rounded up, then the numbers breakdown would be:

1 = 1-5
2 = 6-10
3 = 11-15
...
19 = 91-95
20 = 96-100

Each result has exactly 5 outcomes, out of a possible 100, and 5/100 is equal to 1/20 chance of success. No change on the odds at all.

[BurgerBeast]

This doesn't strike me as being as "game breaking" as some are saying. It looks like it can have serious effects, but as long as you are wary of the extremes and know the math, it could be very interesting.

If you're used to the idea of 2d10 in 3.5, then you have some idea of this. I hesitate to recommend it if you aren't well-versed in the math of probabilities, but I also have zero experience in trying it, whereas you do have that experience.

One thing to watch out for, I would have thought, is how things like the chanpion's improved critical interacts with your choices. It's actually not that bad if you make the first increase 17+ (10%) and the second increase 16+ (15%). So nearly spot on (just the first increase is not as significant as it is under the 1d20 system).

[Tinkermancer]

You could also do something like "doubles crit if they hit," which jumps to about a 10% crit chance but realistically is more like 5-6%, since double 9s would hit, but double 1s wouldn't. Becomes difficult to include expanded crit ranges for Champions and Hexblades, though.

I suppose I could invoke the 'cross that bridge when we get to it' clause of the campaign. I'm not sure if any of my players are interested in those classes yet. Also, considering this is a custom setting with a custom story, it may not matter past this campaign. Thank you for the heads up though. I hadn't considered that.



This is the opposite of fixing the problem. The math is more or less static for the monsters and NPCs if they switch to 2d10 (about 45% to hit either way; only gets off if the PCs have bloated ACs), but swings wildly out of control if PCs switch to 2d10 in combat (from 65% chance to hit to 79%).

Well, unless the problem is the swingyness of the PCs, then it appears to solve it. I'm not sure I have TOO big of a problem with that, as it means I can send larger amounts of easier critters after them, and not have to worry about a series of poor rolls preventing them from hitting. Couple that with the easier enemy's lower HP, and it should lend a very heroic feel to the game.


Now, how would I figure out how much of a bonus Advantage adds? Two ways: 2d10 OR 1d20, or 3d10 drop lowest?

[Lonely Tylenol]

If you're used to the idea of 2d10 in 3.5, then you have some idea of this. I hesitate to recommend it if you aren't well-versed in the math of probabilities, but I also have zero experience in trying it, whereas you do have that experience.

3.5 and 5 are fundamentally different systems. In 3.5, creatures could get from +2 to +6 to certain skills just from racial bonuses, and the expectation is that a 20th-level character will have a +23 just from skill ranks, and a martial-focused character at least a +16 just from BAB. Static bonuses are trivial to come by, and can easily get to +50 from ability bonuses, ranks/BAB/saving throw bonuses from levels, magic items, and spells. Hell, multiple spells gave bonuses upwards of +30 or more (Jump and Glibness come to mind, just from core). The difference between 2d10 and 1d20 is a small fraction of the difference provided by static bonuses, which make up the bulk of pass/fail margins before any dice are ever rolled, and 2d10 is mostly just insurance against fumbles, which is a useful application for it.

In 5e, numbers are much more tightly controlled. For most characters, the difference in numerical bonuses between the start and end of a 20-level build is six (two from ability scores, from +3 to +5, and four from proficiency, from +2 to +6). Magic items inflate this slightly, but not nearly to the extent that they do in 3.5; in 5e, the maximum numerical bonus to attacks is +3, ability items set scores instead of providing numeral bonuses, and the few skill items that exist provide advantage rather than bonuses of any kind). Hitting 30 in something reliably is something legitimately difficult to do, unless you specifically build toward that thing. For the vast majority of characters, the sum of all bonuses provided to the foremost expert in their field at the pinnacle of their natural ability, assisted by magic items, are dwarfed by an average commoner with skill ranks alone, or by a single magical item. This includes attack rolls (thanks to BAB) and saves, both good and poor (poor saves roughly match proficiency bonuses, starting slower and then outpacing them with levels). The dice roll is always relevant. And not only is the dice roll always relevant, but in combat, it is designed to hit certain break points: a CR-appropriate creature should generally always hit on a 12 and be hit on an 8. They fit creatures of appropriate level at roughly the same point in the curve for all levels, which works because of the regular distribution of a single dice. When you throw an actual curve into the equation and ask them to fall on it, the math changes drastically, which we can already see with advantage. 2d10 changes the basic assumptions of the game's math and provides the player with an advantage for no better reason than the system has changed around them.

[Kane0]

Seems like a fine thing to test out. Give it a try and tell us how it goes!

From what I can see it will make easy rolls easier and hard rolls harder because of the increase in consistency. If you are able to do it you have significantly lower chances of failure, but you pay for that with a much higher chance of failure when attempting to do something unlikely. So if your aim is to reduce the swing, should do that pretty well.

[Zman]

Think this kind of change through. From experience I can say it works very well for ability and skill checks ie Initiative, Skills, etc.

But, it really breaks down for to hit rolls and saves. For instance, say you need a 17 to hit an enemy. With a D20 you have a 20% chance of hitting them, but with 2d10 you only have a 10% chance. That makes having a high AC where someone needs to roll a 20 twice as valuable. This is very common for saves.

Or say they have a low a AC, only needing a 7 to hit. With a d20 you have a 70% chance to hit them, now with 2d10 you have an 85% chance of hitting them. The low AC is now worthless twice as often as before. This is similar if you have strong modifiers for saves, it makes hitting even very high ACs easy.

It will also really slow your game down, but when used for Ability checks and Skills while keeping the d20 for Attack rolls and saves the games works very well and it is one of my major recommended house rules. The game as written is not balanced around using 2d10 for attacks and saves, now the kind of bell curve it creates is nice and I like it, but how attacks and saves work in the game it is not really built for it where ability checks gain some reliability and it feels right.

[Laurefindel]

you could try to trigger critical hits on doubles(e.g. Results of 1 and 1, 6 and 6, 0 and 0 etc). That's 10 combinations out of 100, of a 10% chance if you like.

If that's too generous, you could try 'double that hits'. Based on the postulate that PCs hit on a 8+, that would bring it to a double 4 (or above), or 6% crit chance.

It would mean that enemies with higher AC would yield less cricical hits, and characters with high attack values get more crits, which is unorthodox for D&D but not unreasonable for a RPG.

Champions could get a critical hit on any double pair.

Haven't thought of all implications her, just thinking out loud.

[Kane0]

you could try to trigger critical hits on doubles(e.g. Results of 1 and 1, 6 and 6, 0 and 0 etc). That's 10 combinations out of 100, of a 10% chance if you like.

If that's too generous, you could try 'double that hits'. Based on the postulate that PCs hit on a 8+, that would bring it to a double 4 (or above), or 6% crit chance.

It would mean that enemies with higher AC would yield less cricical hits, and characters with high attack values get more crits, which is unorthodox for D&D but not unreasonable for a RPG.

Champions could get a critical hit on any double pair.

Haven't thought of all implications her, just thinking out loud.

Sounds solid to me.

Edit: Wait, what about champion 18-20 crits?

[Laurefindel]

Sounds solid to me.

Edit: Wait, what about champion 18-20 crits?

The player toss a coin. If the coin was minted before 1967 it's a critical hit!

Ahem, I don't know... throw a third die, ignore it if it doesn't create a double? Haven't worked out the maths on that...

[Kane0]

Or just change the ability so it does max damage + roll when you crit, or straight double damage, or EXPLODING CRIT DAMAGE.

[Laurefindel]

Champions with 18-20 critical range get x3 damage on a crit?

[Theodoxus]

Not sure if this is what you're looking for, but one guy at our table routinely rolls a d10 and a d6, as a control die. So, 1-10, and then 1-3 = +0, 4-6 = +10 to the d10 roll.

I'm not a statistician, but I think it's the same odds as a d20, just 'feels' better to roll 2 dice...

My, I'm a fan of the 3d20 where you take the middle die; take the high on advantage and low on disadvantage. The only real problem is lack of crits... still haven't figured out how to make up for that... But it does make combat a lot less swingy - and more tactical, as you're looking for anything that'll boost your rolls out of the 8-12 average.

[BurgerBeast]

3.5 and 5 are fundamentally different systems. In 3.5, creatures could get from +2 to +6 to certain skills just from racial bonuses, and the expectation is that a 20th-level character will have a +23 just from skill ranks, and a martial-focused character at least a +16 just from BAB. Static bonuses are trivial to come by, and can easily get to +50 from ability bonuses, ranks/BAB/saving throw bonuses from levels, magic items, and spells. Hell, multiple spells gave bonuses upwards of +30 or more (Jump and Glibness come to mind, just from core). The difference between 2d10 and 1d20 is a small fraction of the difference provided by static bonuses, which make up the bulk of pass/fail margins before any dice are ever rolled, and 2d10 is mostly just insurance against fumbles, which is a useful application for it.

Understood. Agreed.


In 5e, numbers are much more tightly controlled. For most characters, the difference in numerical bonuses between the start and end of a 20-level build is six (two from ability scores, from +3 to +5, and four from proficiency, from +2 to +6). Magic items inflate this slightly, but not nearly to the extent that they do in 3.5; in 5e, the maximum numerical bonus to attacks is +3, ability items set scores instead of providing numeral bonuses, and the few skill items that exist provide advantage rather than bonuses of any kind). Hitting 30 in something reliably is something legitimately difficult to do, unless you specifically build toward that thing. For the vast majority of characters, the sum of all bonuses provided to the foremost expert in their field at the pinnacle of their natural ability, assisted by magic items, are dwarfed by an average commoner with skill ranks alone, or by a single magical item.

Lost me here. Some awkward language at the end, I think.


This includes attack rolls (thanks to BAB) and saves, both good and poor (poor saves roughly match proficiency bonuses, starting slower and then outpacing them with levels). The dice roll is always relevant.

Understood. Agreed.


And not only is the dice roll always relevant, but in combat, it is designed to hit certain break points: a CR-appropriate creature should generally always hit on a 12 and be hit on an 8.

On this point the difference is minimal. It's outside these ranges that problems arise.


They fit creatures of appropriate level at roughly the same point in the curve for all levels, which works because of the regular distribution of a single dice.

As I say, this is the acceptable part of the change. It's the edges of the curve that present the problem.


When you throw an actual curve into the equation and ask them to fall on it, the math changes drastically, which we can already see with advantage. 2d10 changes the basic assumptions of the game's math and provides the player with an advantage for no better reason than the system has changed around them.

It's a difference that can be adjusted for, though, is my point. It changes the math, and requires adjusting.

My point is that the more tightly the ranges are controlled, the less of a problem the switch to 2d10 is. So, in this regard, a system that is more likely to use the 8-12 range is less affected by the switch. As long as the DM is aware of the problem of extremely narrow success ranges, he can work with it.

[Pex]

What will Bless do with its +1d4 to hit and saving throws and its counterpart Bane?

[BurgerBeast]

What will Bless do with its +1d4 to hit and saving throws and its counterpart Bane?

Substantially increase in value.

The good news is that there's no discrepancy in how it scales by level. (i.e. It doesn't increase or decrease in value as you level.)

[rbstr]

Not sure if this is what you're looking for, but one guy at our table routinely rolls a d10 and a d6, as a control die. So, 1-10, and then 1-3 = +0, 4-6 = +10 to the d10 roll.

I'm not a statistician, but I think it's the same odds as a d20, just 'feels' better to roll 2 dice...


I didn't believe that'd be the case for some reason...but it looks like it is equivalent to a d20. ! If i've done it right.
You can actually roll that with anydice, which is pretty cool http://anydice.com/program/b5c5

But given that they're identical, it's really only fun since it's more dice. It doesn't do anything vs. a d20.

[Thrudd]

Here's how you do -

Figure out the percentages associated with rolling each number in the rolling method you are thinking about - looks like done already.

Decide how often you think certain things should be accomplished in the game, and then alter the numbers to fit what you want. You may need to change AC values around to make sure characters are hitting as often as you think they should. Spell DC's and other saving throws as well will need to be adjusted. Pretty much anything that requires rolling a d20 to determine you should rewrite with a value that is in line with the percentage of success/failure you think appropriate. Go through the entire book, all the classes, all the spells, all the monsters and check for anything that needs readjusting.

[BurgerBeast]

(Unsolicited, sarcastic, and entirely wrong comments.)

Yeah, except that some of the changes are the desired changes. Only the undesirable changes need be made.

So, yeah... no.

You're probably really only looking at the situations where the d20 roll is required to be 16+ (maybe 17+) to hit and 5- (maybe 4-) to miss.

[Steampunkette]

The ultimate result of 2d10 in place of 1d20:

Combat becomes more predictable. Average rolls are more common than outlier rolls and thus more simple hits are achieved. Very difficult tasks become even more difficult as rolling a 17 or higher is progressively more rare. Critical successes are minimally more common.

And... that's it.

You'll get hit by enemies more often, you'll hit enemies more often, and beings with very high ACs will become harder to hit.

You'll need a few mechanical shifts to make sure things like Advantage and Disadvantage work reasonably well (personally I'd suggest just adding a third d10 and taking the two highest/lowest).

Be wary of bonus-granting abilities (Bless, Guidance, Bardic Inspiration) as those can use this setup to become much more powerful.

Suddenly, the dice roll is vastly less important than the bonuses you have stacked onto it. This can lead to powergaming pretty badly. But it does better represent a world where a skilled swordsman is better at hitting someone without some lowly peon showing them up through a series of wildly RNJesus rolls... It will still happen, just much less often.

[Kane0]

I didn't believe that'd be the case for some reason...but it looks like it is equivalent to a d20. ! If i've done it right.
You can actually roll that with anydice, which is pretty cool http://anydice.com/program/b5c5

But given that they're identical, it's really only fun since it's more dice. It doesn't do anything vs. a d20.

The same can be done with any die or coin plus the d10, no? Use the d10 as normal and high/low, odds/evens, heads/tails add either +0 or +10. 50% chance of 1-10, 50% chance of 11-20, 10% chance of each result within so 5% all around just like a d20.

[Ruslan]

Crits - On a d20, crits happen 5% of the time and only on a 20. On 2d10, if crits only happen on a 20, that means a crit only happens 1% of the time. If we move it to 19 and 20, it increases to 3%. Still, not quite there. If we add 18 to the mix, it moves to 6%. That's better. If you want exactly 5%, then crits happen on any roll of 18+, but not 9/9. So, 10/8, 8/10, 10/9, 9/10 and 10/10 are crits. Easy enough to remember.



Variants - What if instead, crits happen when one of the dice show a 0? That means a crit will happen 5% of the time (I think), and we can make 00 be a crit with max damage for extra fun!Actually 19%.
10/1, 10/2, 10/3 .... 10/10, 9/10, 8/10 ... 2/10, 1/10. Nineteen combinations out of 100.



These are just shots in the dark, but maybe enough to get some creative juices flowing. Any other thoughts on the repercussions of the 2d10 variant, any thoughts on variations of this variant, and any ideas for magic items? I'd love to steal your ideas for my game! Thanks again!
It will make combat (and skills, I guess) more predictable and less swingy. A lot more chances to get a result 'around the middle' than an outlier.

[JellyPooga]

Glibness will become worth its 8th level slot. :smallbiggrin:

[skaddix]

As noted in 3.5 you could hit much higher values.

Level 20 lets say with a Belt of Giant Strength (29) and a +3 Weapon gets you what

9+6+3=18 that is about as good as you can get before die as noted above that be laughable in 3.5 for a lvl 20.

[Lonely Tylenol]

Lost me here. Some awkward language at the end, I think.

TL;DR the way this system is limited in its bonuses, no amount of cheese and optimization will ever give you a bigger bonus than Joe Shmoe in 3.5 gets in skill ranks just for leveling up.


On this point the difference is minimal. It's outside these ranges that problems arise.

"Outside these ranges" is the default assumption for a large part of the game, though. The game is designed so that monster hordes are always threatening in numbers, so you can't rush headlong into an army and just win, like you can in 3.5. But if you move outside the default expectations of the game even slightly for weaker creatures, such as a Knight attacking an 8th-level party member with AC 19, their to-hit bonus plummets seventeen percent for being just two away from the expectations for a single challenging creature, rather than the ten percent they'd drop normally. If you think that's supposed to be normal, because a CR 3 Knight is five "levels" below an 8th-level adventurer, note that three Knights are considered a Medium-Hard encounter for four such PCs. And it gets worse before it gets better: the difference between a CR 1/2 Deep Gnome with 2d10 and 1d20 is a nine-point swing, with the 2d10 Deep Gnome hitting about two thirds as often as the 1d20 (21% vs 30%).


As I say, this is the acceptable part of the change. It's the edges of the curve that present the problem.

"The edges of the curve" make up the vast majority of all games. What do you expect to happen if you endeavor to challenge a group with an encounter that doesn't consist of a single level-appropriate monster? Three Challenge 4 Gnoll Fang of Yeenoghu (a Hard encounter for a party of four 9th-level adventurers) would take damage from any adventurer who diligently maxed their primary stat a shocking 94% of the time (an 18 primary stat, from, say, a feat-taker, would still hit 90% of the time), while the Fang would hit a Fighter in +1 plate 28% of the time, or 15% of the time if that Fighter wore a nonmagical shield.

To put into perspective how skewed those numbers really are: the effective AC of the Gnoll Fang of Yeenoghu in this encounter (to the PCs) is 11. That is an AC matched, and often beaten, by CR 1/4 and 1/8 creatures, who are individually not meant to be a threat even to a first-level party (these are your boars, mastiffs, and bandits). They have the same effective to-hit bonus against this party as most CR 1/8, with the 14 they require to hit a 19 AC being worth about as much, statistically speaking, as the 16 a CR 1/8 Bandit would need to hit, and the 16 they need to hit a 21 AC being worth as much as the Bandit's 18 to-hit threshold with a d20. These aren't even horde encounters (this is one above a pair, and already classed as Hard), but because they fall off the top of the probability curve, 2d10 absolutely trivializes them.


It's a difference that can be adjusted for, though, is my point. It changes the math, and requires adjusting.

My point is that the more tightly the ranges are controlled, the less of a problem the switch to 2d10 is. So, in this regard, a system that is more likely to use the 8-12 range is less affected by the switch. As long as the DM is aware of the problem of extremely narrow success ranges, he can work with it.

It cannot be adjusted for unless you specifically give bonuses to monsters outside their normal CR range, like "+1 for pairs, +2 for groups of 3-6, +3 to groups of 7-12," and so on, with the expectation that these bonuses would only be used for groups of weaker monsters meant to constitute a standard encounter, which requires an eyeball test. Additionally, ACs would need to be flatly increased across the board, but only for monsters, so that adventurers don't just get a free effective +3 to hit against CR-appropriate monsters (which is what changing from 1d20, with its 65% to hit, to 2d10, with its 79% to hit, does). Oh, but these ACs should only be adjusted by about 2, because that's a 15% difference in to-hit on its own. And also you need to watch out if your party decides to punch up (to, say, a monster 1 or 2 above their normal Challenge) or fights an encounter whose shtick is having a good AC, because that AC bump could cause dramatic changes in difficulty if it moves monsters across the top of the bell curve, which is actually the most volatile part of every bell curve (for example, moving from "hitting on an 8" to "hitting on a 12" is a 34% swing for 2d10, as opposed to a 20% swing for 1d20). And then you also need to do this for all monsters' to-hit and save bonuses, lest you make the make the mistake of giving a CR 1 Quickling a strong curve and accidentally TPKing your first-level party.

(This is NOT hyperbole: the Quickling has 16 AC and imposes disadvantage on attacks against it, so however you rule disadvantage you're likely to have something like a 30% chance to hit it, but it has a +8 to hit and a +6 modifier on damage rolls, so it'd have upwards of an 80% chance to hit with each of its three attacks against a Chainmail-wearing Fighter, higher against everyone else.)



(Unsolicited, sarcastic, and entirely wrong comments.)

(Quoted portion of post)

This is really misplaced. This is an open forum, which means that his input is every bit as "solicited" as yours—you have no right to post here that everyone else does not have, Thrudd included.

He's also right, as it were—changing the dice used from regular distribution to a probability curve challenges or outright changes a lot of basic assumptions about the game, such as what it means to "hit on an 8" and "be hit on a 12," and what monsters should constitute a threat to the party, and whether or not D&D 5e is a game that should have hordes or group encounters challenge parties. In order to avoid challenging all of these very basic assumptions about how the game is meant to be played, you need to manually adjust all combat values of all monsters that might be affected by this change, and decide about how often each monster should be affected by X ability which targets Y defense, and similarly, how often it should wound the party in such a fashion. At least as much care and attention should go into revising them as it did creating these values in the first place, or you create edge cases like a Quickling TPK, or Gnoll Fang Commoners, that the game clearly did not intend.

There is no need for you to be flippant or dismissive to him with your "Yeah... No" response—he's probably the most correct person in this whole thread, and deserves at least as dignified a response as you would ask for yourself.

[Unoriginal]

Not sure how 5e is "swingy".


3.X pretty quickly became "dice doesn't matter, I have +20 everywhere I need." Guess 5e is swingier, but it's like saying that a snail is faster than a rock.

[JellyPooga]

@Tylenol: Yes the game will change dramatically, but it changes for everyone. That just means the flavour of the game changes, much as using the Gritty Realism rules does. Using 2d10 will make the game more lethal, due to more hits going through, making combat more risky and defence and tactical thinking paramount over charging in to encounters blindly.

[Tinkermancer]

@Tylenol: Yes the game will change dramatically, but it changes for everyone. That just means the flavour of the game changes, much as using the Gritty Realism rules does. Using 2d10 will make the game more lethal, due to more hits going through, making combat more risky and defence and tactical thinking paramount over charging in to encounters blindly.

This is what I was going for, amongst other things. Regardless, I plan on running this way for a bit to see what the true outcome will be. There is no substitute for experience after all!

Now, I suppose I should start a different thread for my magic item question, as it seems like everyone is too focused on the 2d10 here lol

[JellyPooga]

Do note that using 2d10 will also make high-end armour very good. Plate armour will be well worth it's weight, for example. As such, you may want to make the game gritty and the setting relatively low-market i.e. only nobles (and high-level adventurers!) wear much above Hide Armour and wield "advanced" weapons like swords (particularly Rapiers and Two-handers). Magic is also something you may want to consider reining in because, as others have noted, it's effects will be more dramatic; using the Gritty Realism rules might be an idea.

[Sir cryosin]

Thanks for the break down!

I didn't realize "one dice is a 0" increased the chance so much! That would DEFINITELY break the game. I would do better to just crit on 18-20 then, with a max damage crit on 00.

Also, I had considered only using 2d10 for the players, keeping the d20 for everyone else for unpredictability and lack of training. I would probably use 2d10 for trained NPCs.

Furthermore, I am considering not only for its own sake, but for the procing customization I posted as well. Without that part, I probably wouldn't use it, considering your data.

If your criting on a 18-20 thats stepping on the champions toes. So you would have to adjust to this new range. Making the champion ridiculous in damage output.

[Tinkermancer]

If your criting on a 18-20 thats stepping on the champions toes. So you would have to adjust to this new range. Making the champion ridiculous in damage output.

I would likely also make Champions crit (if they hit) on any double number, and then later make them do max damage on every crit. I believe they get two crit upgrades, of the top of my head.

[Tanarii]

The disadvantage of this is when you're outmatched, you're really outmatched. (Or possibly that's an advantage, depending on how you want the game to play.) It makes cakewalks even cake-walkier, and turkey-shoots even turkey-shootier.

Advantage and disadvantage get pretty screwy btw:

Advantage:
http://anydice.com/program/b5d0

Disadvantage:
http://anydice.com/program/b5d1

Edit: to be clear, what's screwy is the highest/lowest die roll are no longer the most likely result by a large margin. Even given you're trying to move things to the center, you'd expect them still to be an equally likely result to retain intended purpose.

[Tinkermancer]

The disadvantage of this is when you're outmatched, you're really outmatched. (Or possibly that's an advantage, depending on how you want the game to play.) It makes cakewalks even cake-walkier, and turkey-shoots even turkey-shootier.

Advantage and disadvantage get pretty screwy btw:

Edit: to be clear, what's screwy is the highest/lowest die roll are no longer the most likely result by a large margin. Even given you're trying to move things to the center, you'd expect them still to be an equally likely result to retain intended purpose.

It goes up if you do 'highest 2 of 3d10' like you'd expect. I'm trying to figure out to program in 'highest ?? of 1d20, 2d10' but having no luck.

[Tanarii]

It goes up if you do 'highest 2 of 3d10' like you'd expect. I'm trying to figure out to program in 'highest ?? of 1d20, 2d10' but having no luck.Yeah I've never figured out how to do a highest or lowest of two different kinds of dice. Pretty sure it'd require a custom function.

[Laurefindel]

If your criting on a 18-20 thats stepping on the champions toes. So you would have to adjust to this new range. Making the champion ridiculous in damage output.

The OP was crit on a 18-20 as the sum of 2d10 (6% crit range), not 18-20 on a d20 (15% crit range).

I presume the champion would get proportionally higher ranges (on 2d10) to match its 10% crit range at 3rd level and 15% crit range at 17th level.

[Ruslan]

The OP was crit on a 18-20 as the sum of 2d10 (6% crit range), not 18-20 on a d20 (15% crit range).

I presume the champion would get proportionally higher ranges (on 2d10) to match its 10% crit range at 3rd level and 15% crit range at 17th level.
It's simple:

For everyone: Crit on 18+, except 9/9 (exactly 5% chance) (if that's too complicated, remove the 'except 9/9' clause making it 6%, not much of a difference)
For a mid-level Champion: Crit on 17+ (exactly 10% chance)
For a high-level Champion: Crit on 16+ (exactly 15% chance)

[Grod_The_Giant]

5e really wants to be on a bell curve. It tells you "this +1 bonus is very significant!" when no, it's really just 5% and the d20 roll is the main thing that matters. 2d10 is a nice shallow curve; it helps in nicely unobtrusive ways. It'll change 5e in certain ways, sure; your higher numbers will become more meaningful, and if you want hordes of weak monsters to remain threatening you might need to pump up the numbers a bit, make more use of Help actions, or mix in more leader-type monsters. Or you could tweak their stats; it looks like a +2 AC (and save DC) across the board would help make up for the higher average roll.


(Suggestion for Advantage/Disadvantage: roll 3d10 and take the best/worst 2; it shifts the mean from around 11 to 14 or 8, which is acceptably close to the original)

[Snails]

I think 2d10 makes sense for skills. I think that it will tend to make very high ACs superb and less than high ACs worthless -- I do not like that, but it is perfectly playable.

I really do not think 2d10 works for Saves. IMHO 5e already has a scaling issue in saves as you climb up the levels.


Consider a basic Fighter.

Fighter 1 -- Will +0 ; typical spell DC ~12 ; d20 --> 45% success ; 2d10 --> 45% success
Fighter 5 -- Will +0 ; typical spell DC ~14 ; d20 --> 35% success ; 2d10 --> 28% success
Fighter 10 -- Will +0 ; typical spell DC ~16 ; d20 --> 25% success ; 2d10 --> 15% success
Fighter 15 -- Will +0 ; typical spell DC ~18 ; d20 --> 15% success ; 2d10 --> 6% success
Fighter 20 -- Will +0 ; typical spell DC ~19 ; d20 --> 10% success ; 2d10 --> 3% success

With standard 5e, we can see that success rate of saves just falls off a cliff once we hit the someteenth levels. That is probably a minor issue, as we can reasonably expect that PCs of those lofty levels could eventually gain some kind of magic items to mitigate this issue. That argument does not hold in the single digit levels, where expectations of the system working for low or no magic items should hold.

With 2d10 for saves, you will start seeing effectively "hopeless" saves down at 10th, 9th, 8th level, maybe even 7th level, well before it is reasonable to hope magic items will help the PCs enough.

When theorycrafting, it is easy to say "Well, I guess you could boost your Wis and have a +1 Will, so you would have a 21% of saving, instead of 15%. And you can save against that Will spell every round." At the gaming table, it means if get targeted by that save on Round 1, my PC is probably effectively locked for the rest of the combat -- it is easy to fail 5 or 6 or 7 saves in a row.

EDIT: Corrected typo on list. (Thanks, Zman.)

[Zman]

I think 2d10 makes sense for skills. I think that it will tend to make very high ACs superb and less than high ACs worthless -- I do not like that, but it is perfectly playable.

I really do not think 2d10 works for Saves. IMHO 5e already has a scaling issue in saves as you climb up the levels.


Consider a basic Fighter.

Fighter 1 -- Will +0 ; typical spell DC ~12 ; d20 --> 45% success ; 2d10 --> 45% success
Fighter 5 -- Will +0 ; typical spell DC ~14 ; d20 --> 28% success ; 2d10 --> 35% success
Fighter 10 -- Will +0 ; typical spell DC ~16 ; d20 --> 15% success ; 2d10 --> 25% success
Fighter 15 -- Will +0 ; typical spell DC ~18 ; d20 --> 6% success ; 2d10 --> 15% success
Fighter 20 -- Will +0 ; typical spell DC ~19 ; d20 --> 3% success ; 2d10 --> 10% success

With standard 5e, we can see that success rate of saves just falls off a cliff once we hit the someteenth levels. That is probably a minor issue, as we can reasonably expect that PCs of those lofty levels could eventually gain some kind of magic items to mitigate this issue. That argument does not hold in the single digit levels, where expectations of the system working for low or no magic items should hold.

With 2d10 for saves, you will start seeing effectively "hopeless" saves down at 10th, 9th, 8th level, maybe even 7th level, well before it is reasonable to hope magic items will help the PCs.

When theorycrafting, it is easy to say "Well, I guess you could boost you Wis and have a +1 Will, so you would have a 21% of saving, instead of 15%. And you can save against that Will spell every round." At the gaming table, it means if get targeted by that save on Round 1, my PC is probably effectively locked for the rest of the combat -- it is easy to fail 5 or 6 or 7 saves in a row.

It does work very well for skills, I've been using it for over a year.


You mixed up the 2d10 and the d20 in your list. Your point is extremely valid, the 2d10 greatly alters the game when used for to hit rolls and especially for saves. It requires a lot of alterations to the game to maintain balance.

[Snails]

Not sure how 5e is "swingy".

3.X pretty quickly became "dice doesn't matter, I have +20 everywhere I need." Guess 5e is swingier, but it's like saying that a snail is faster than a rock.

Well, yes. "Swingy" is just a statement that with flatter math it is inevitable to see lots good runs and bad runs by the luck of the dice. Such was always true in earlier editions for low levels, but much less so as you climbed up levels. 5e preserves this "early level" flatness much, much further.

[Tinkermancer]

It does work very well for skills, I've been using it for over a year.


You mixed up the 2d10 and the d20 in your list. Your point is extremely valid, the 2d10 greatly alters the game when used for to hit rolls and especially for saves. It requires a lot of alterations to the game to maintain balance.

Is this more a commentary on how poorly saving throws work out as levels increase?

[Zman]

Is this more a commentary on how poorly saving throws work out as levels increase?

Not just as levels progress, but any time a high or low roll is needed for attacks and saves, the 2d10 method greatly alters expectations. Non proficient saves are just a prime example of this.

2d10 is amazing for skills and ability checks, but IMO is not well suited for attack rolls and saves. Now, had 5e been designed with the light curve of 2d10 in mind it could have been made to handle it well, but stock there are problems with it.

[Ruslan]

2d10 is amazing for skills and ability checks, but IMO is not well suited for attack rolls and saves. I agree with this very much. Combat is chaotic [everyone has a chance to hit or miss everyone], but out-of-combat skills are supposed to be predictable - if you're good at a task, you're good. If you're bad, you're bad.

[Thrudd]

Yeah, except that some of the changes are the desired changes. Only the undesirable changes need be made.

So, yeah... no.

You're probably really only looking at the situations where the d20 roll is required to be 16+ (maybe 17+) to hit and 5- (maybe 4-) to miss.

Yeah, you won't need to change everything because some are within the percentages you want. But you still look at everything, to make sure there won't be any weird surprises or points in the game where things become totally broken with neither players nor monsters being able to touch each other, for example. The range of potential bonuses may need to be shifted, AC increased or decreased, saving throws increased or decreased, depending on what you want. You're changing the math of the primary game mechanic. It affects a lot of things, is all I'm saying, don't underestimate it's overall effects. It pays to be thorough.

[Snails]

I agree with this very much. Combat is chaotic [everyone has a chance to hit or miss everyone], but out-of-combat skills are supposed to be predictable - if you're good at a task, you're good. If you're bad, you're bad.

Part of this is the relative richness of the sub-systems. Every class gains specific meaty abilities/resources that tweak or mitigate the random chaos of battle -- combat is a very detailed sub-system. But most classes offer zilch when it comes to skills -- your DM can throw you a bone if your background or roleplaying happens to suggest it, but otherwise you have your +n and a d20 and there is nothing more to say.

[Grod_The_Giant]

Part of this is the relative richness of the sub-systems. Every class gains specific meaty abilities/resources that tweak or mitigate the random chaos of battle -- combat is a very detailed sub-system. But most classes offer zilch when it comes to skills -- your DM can throw you a bone if your background or roleplaying happens to suggest it, but otherwise you have your +n and a d20 and there is nothing more to say.
The single largest determining factor, I think, is that combat takes multiple rolls. Lots of them. Even a detailed skill challenge has fewer rolls than, like, one round of combat. Maybe even one turn.

[Zalabim]

You also aren't supposed to die as the result of one d20 roll, but people don't follow the suggestions in the book and then say they want to effectively change every number in the game before trying the game as its presented.

"Swingy" to me means either "I fail too much" or "other people don't fail enough" which always strikes me as a rather selfish complaint either way. You are less likely to fail three times in a row on a 79% chance than on a 65% chance (target number is 8), but we can't just give you an 80% chance (lower the DC by 3) because then the guy who's supposed to suck has a 55% chance instead of a 36% chance (5 less bonus) and people who suck have to know it. Or something.

It really boils down to an attitude about how much "unusual" is acceptable. Or maybe I just have an irrational response to bell curves.

Yeah I've never figured out how to do a highest or lowest of two different kinds of dice. Pretty sure it'd require a custom function.

Here: http://anydice.com/program/b606 and if you want to go deeper, you can create nested [highest ] functions. Like [highest of [highest 1 of 2d20] and [highest 2 of 3d10]].

[Spiritchaser]

I've played a few games that used two dice for hits... car wars, and the dreaded hellbore in SFB, a I think battletech, and a few others...

Very generally I preferred it. I preferred a certain degree of randomness, but not too much randomness.

I'd expect you'd need to re-assess CR of most things, and carefully consider each encounter for the first few games.

After a while you'd get a general sense of what needed modification before each fight, and be able to wing it more, with the occasional outlier where a certain encounter would be crazier than you expected.

I think it's a great and noble goal to try... and I'd love to hear how it goes, but... I cannot see myself spending that kind of effort when what's there is functional.

I'd also note that 5e is NOT perfectly balanced (though it's certainly not the worst), and that to a certain degree, a failure in balance is mitigated by fairly random dice. Reducing the swingyness of the dice roll would likely demand more precision in balance and playtest, which is even more time that I cannot imagine spending.

[Malifice]

Advantage could be modeled at 3d10 drop the lowest die.

Ditto but reversed for disadvantage.

I would leave critical hits at a 19 or 20. With the champion upgrading that to an 18 and then a 17.

[Malifice]

Also; expertise would be too powerful in this variant.

I would reduce it to a flat bonus of +2, increasing to +3 at 11th level.

Similarly I would nerf archery style to only grant plus 1 to hit.

Magic Shields and Magic armor should not be allowed to stack.

[Grod_The_Giant]

You also aren't supposed to die as the result of one d20 roll, but people don't follow the suggestions in the book and then say they want to effectively change every number in the game before trying the game as its presented.

"Swingy" to me means either "I fail too much" or "other people don't fail enough" which always strikes me as a rather selfish complaint either way. You are less likely to fail three times in a row on a 79% chance than on a 65% chance (target number is 8), but we can't just give you an 80% chance (lower the DC by 3) because then the guy who's supposed to suck has a 55% chance instead of a 36% chance (5 less bonus) and people who suck have to know it. Or something.

It really boils down to an attitude about how much "unusual" is acceptable. Or maybe I just have an irrational response to bell curves.
"Dying in one d20 roll" isn't great, but it isn't really what the swingy complaint is about. At least for me, because swinginess is one of my biggest complaints about 5e. My complaint is that having small bonuses relative to the flat curve of the d20 means that the line between "good at something" and "not good at something" gets blurred. I want to reliably succeed at the things I'm supposed to be good at, and struggle at the things that I'm not, because of the choices I made, not because of the luck of the die. I want my Barbarian to pass Str checks and fail Int ones, and I want my friend's Wizard to do the opposite. Adding in a bell curve helps make that happen, while still keeping numbers low enough that you don't entirely lose the potential for unexpected success or failure.

[Tanarii]

"I want to reliably succeed at the things I'm supposed to be good at, and struggle at the things that I'm not, because of the choices I made, not because of the luck of the die.
I prefer to have success depend on the choices I make during the game, not during character creation. That requires a smaller difference between adventures who are good at an adventuring task (low bonus), and adventurers who are really good at an adventuring task (high bonus), and adventures who are experts at an adventuring task (expertise for very high bonus).

What I don't want is an adventurer who struggles to do the adventuring tasks he's not really good at. Been there in 3e, done that with frequently called for checks for everything, with only one character maybe having a high enough bonus, the rest have no chance because low bonus, over it.

[Snails]

I prefer to have success depend on the choices I make during the game, not during character creation. That requires a smaller difference between adventures who are good at an adventuring task (low bonus), and adventurers who are really good at an adventuring task (high bonus), and adventures who are experts at an adventuring task (expertise for very high bonus).

What I don't want is an adventurer who struggles to do the adventuring tasks he's not really good at. Been there in 3e, done that with frequently called for checks for everything, with only one character maybe having a high enough bonus, the rest have no chance because low bonus, over it.

Good decisions sometimes do just solve the problem, or make some kind of success near inevitable, albeit at a possible cost in time or other resources.

For me, I would call that immersion crushing that PCs must be stuck in a box of such modest competence, that character concepts are subservient to keeping everyone on the same very small playing field. If I chose to create a robust outdoorsman (high Str Barbarian with Athletics), at the point in his career that his traveling companion can literally bring the dead to life, do we need him to have a good chance to fall off the castle wall while climbing? W. T. F.

An aside...

3e made an ill-conceived design decision to pit non-magical "skill" classes directly against non-magical "fighter" classes, in terms of areas of competence. As a result, most classes were stuck with such a painfully small number skill points, that building up even a +5 in a cross class skill to round out the character concept seemed too costly and expensive to bother with.

5e gives PCs a very reasonable 5+ skills out of the starting gate -- that rounds out characters pretty well. What 5e lacks is an appropriate means to become genuinely good at any one thing (other than a one level dip into Rogue), or an organic means to acquire a new skill (other than invest an entire Feat to acquire three skills).

Back to the OP topic, I think that 2d10 does serve to reward very good skills with many more successes for somewhat difficult tasks (~DC 15), and also rewards PCs with modest skill (say, +2 on the skill) with more successes for normal tasks (~DC10). That works for me (even if it does not address every concern I might have about the skill system).

[Tanarii]

Good decisions sometimes do just solve the problem, or make some kind of success near inevitable, albeit at a possible cost in time or other resources.

For me, I would call that immersion crushing that PCs must be stuck in a box of such modest competence, that character concepts are subservient to keeping everyone on the same very small playing field. If I chose to create a robust outdoorsman (high Str Barbarian with Athletics), at the point in his career that his traveling companion can literally bring the dead to life, do we need him to have a good chance to fall off the castle wall while climbing? W. T. F.Sounds like a DM that hasn't read the DMG section on Running the Game, which goes into the use of dice and the idea that you don't always need to make a check. In other words, he's still running the game just like 3e, which as I just mentioned isn't my ideal any more. Not everything requires a check. Just things that have a chance of failure and a chance of success.

I'm not trying to contend that the 5e system doesn't have it's down-sides, as others have often pointed out. Nor even that the 3e system doesn't have it's up-sides, I loved it when I played it! But trying to use the 5e system in a 3e-like manner quickly breaks the system.

And also that I personally am tired of the down-sides of the 3e system, mainly the discrepancy in ability where you can do a small number of things (those skills you've maxed), and not do any other thing at at all (everything else). I prefer the 5e system, where adventurers can do the majority of non-class adventuring things fairly well, and all of them can do portion of those things considerably better.

[Snails]

And also that I personally am tired of the down-sides of the 3e system, mainly the discrepancy in ability where you can do a small number of things (those skills you've maxed), and not do any other thing at at all (everything else). I prefer the 5e system, where adventurers can do the majority of non-class adventuring things fairly well, and all of them can do portion of those things considerably better.

Agreed. A PC with 4+ broadly defined skills and a set of tools feels more like a real person who once had a life, and is stepping out into the adventuring world already somewhat competent.

[Zalabim]

Good decisions sometimes do just solve the problem, or make some kind of success near inevitable, albeit at a possible cost in time or other resources.
This is really sound advice.


For me, I would call that immersion crushing that PCs must be stuck in a box of such modest competence, that character concepts are subservient to keeping everyone on the same very small playing field. If I chose to create a robust outdoorsman (high Str Barbarian with Athletics), at the point in his career that his traveling companion can literally bring the dead to life, do we need him to have a good chance to fall off the castle wall while climbing? W. T. F.
Several things don't stick with this though. The casters are bringing the dead back to life in some form at level 5. Just falling off a climbable wall is such a catastrophic failure it strains belief, but +7 on the Strength check is going to produce that result noticeably less often than +0 will. It's mostly an example of questionable game mastering. Questions like "is that really necessary? is that the kind of narrative you want?"


5e gives PCs a very reasonable 5+ skills out of the starting gate -- that rounds out characters pretty well. What 5e lacks is an appropriate means to become genuinely good at any one thing (other than a one level dip into Rogue), or an organic means to acquire a new skill (other than invest an entire Feat to acquire three skills).
The baseline for skills is actually 4 proficient, though a lot of races and classes give more, and indeed a lot of races, classes, and feats can affect skills or ability checks, and it's open to rewards in play as well. There's just no assumption that you can spend some time and money and acquire proficiency, or the equivalent, with every skill in the game. It also might be a little weird to gain a new skill proficiency every 6 levels, for example, because it assumes some particular things about what proficiency means in the game.

[Knaight]

Sounds like a DM that hasn't read the DMG section on Running the Game, which goes into the use of dice and the idea that you don't always need to make a check. In other words, he's still running the game just like 3e, which as I just mentioned isn't my ideal any more. Not everything requires a check. Just things that have a chance of failure and a chance of success.

I'm not trying to contend that the 5e system doesn't have it's down-sides, as others have often pointed out. Nor even that the 3e system doesn't have it's up-sides, I loved it when I played it! But trying to use the 5e system in a 3e-like manner quickly breaks the system.

And also that I personally am tired of the down-sides of the 3e system, mainly the discrepancy in ability where you can do a small number of things (those skills you've maxed), and not do any other thing at at all (everything else). I prefer the 5e system, where adventurers can do the majority of non-class adventuring things fairly well, and all of them can do portion of those things considerably better.

There's more than just these two options though - it's entirely possible to have a game where you can get 3.x style extreme competence at the skills you most favor while not having 3.x style ineptitude at basically everything else. For instance, most skill based systems accomplish this by spending experience/advancements/whatever on skills directly and it being cheaper to improve low skills than high skills, thus favoring broadly competent characters who have some weaknesses and some strengths. 3.x doesn't do this, partially because the way skills scale linearly, partially because the skill points per level is staggeringly stingy ((2+int)? Really? On anything that isn't int based?). There's a few systems that require that you have at least as many skills at 1 as you do at 2 as you do at 3*, etc.

*Where 1, 2, and 3 represent actual meaningful values and not the equivalent of +1-3 on a d20.

[Tanarii]

There's more than just these two options though - it's entirely possible to have a game where you can get 3.x style extreme competence at the skills you most favor while not having 3.x style ineptitude at basically everything else. For instance, most skill based systems accomplish this by spending experience/advancements/whatever on skills directly and it being cheaper to improve low skills than high skills, thus favoring broadly competent characters who have some weaknesses and some strengths. 3.x doesn't do this, partially because the way skills scale linearly, partially because the skill points per level is staggeringly stingy ((2+int)? Really? On anything that isn't int based?). There's a few systems that require that you have at least as many skills at 1 as you do at 2 as you do at 3*, etc.

*Where 1, 2, and 3 represent actual meaningful values and not the equivalent of +1-3 on a d20.Most of the systems I'm familiar with that have cheap --> expensive scaling for increasing skill chance of success don't have the 3e system of extreme competence. Very few systems do that I've played with, Runequest comes closest with it's well over 100% skill checks on a theoretical 5%-100% range. Contrast that with WHFP (typically 30% to 70%, +10% to -30% modifier for difficulty). Or Palladium (~50%-98%). (All ranges assume you've leveled the skill considerably.) OTOH all three of those have a system where some skills can be done untrained, and others can't be done at all without training, similar to 3e.

Personally I find a range of 'unskilled = 50% chance of success' to 'skilled = 70% chance of success' to 'skilled & expert = 90% chance of success' to be sufficient at level 1. With a -25% chance for more difficult tasks, and a -50% for particularly hard tasks, and up to +30% for leveling.

[TentacleSurpris]

One way to do crits would be if you rolled doubles on the dice and you also hit. Just like in Monopoly.

Doubles occurs 10% of the time. If hitting occurs about half the time, then that approximates the current 5%.

[Knaight]

Most of the systems I'm familiar with that have cheap --> expensive scaling for increasing skill chance of success don't have the 3e system of extreme competence. Very few systems do that I've played with, Runequest comes closest with it's well over 100% skill checks on a theoretical 5%-100% range. Contrast that with WHFP (typically 30% to 70%, +10% to -30% modifier for difficulty). Or Palladium (~50%-98%). (All ranges assume you've leveled the skill considerably.) OTOH all three of those have a system where some skills can be done untrained, and others can't be done at all without training, similar to 3e.
I wouldn't assume leveling the skill though, and 10% vs. 90% is a pretty huge gap. With that said - as another example consider dice pool systems. Shadowrun runs from about 4 to 30 dice, Burning Wheel about 2-12 (plus something that makes the dice 33% better, so more like 2-16). On a binomial distribution that's pretty huge, and both have mechanisms that favor building at least somewhat broadly skilled characters, although the BW one is a bit more convoluted than just more points for higher skills.



Personally I find a range of 'unskilled = 50% chance of success' to 'skilled = 70% chance of success' to 'skilled & expert = 90% chance of success' to be sufficient at level 1. With a -25% chance for more difficult tasks, and a -50% for particularly hard tasks, and up to +30% for leveling.

In d20 terms, if we assume that unskilled is +1 (average attribute of 12-13 on 4d6k3) that makes skilled +5 and skilled & expert +9 at level one, with +12 at the high range. The level one numbers there are much bigger than the level one numbers in 5e, and I'd argue that the 30% there doesn't make a whole lot of sense given that magic and combat get way more than +30%. Take combat - you start at about +5, and end at +11 with pure stats. That +11 can easily turn into +13 with the various combat styles and the like, and +15 with a magic item that isn't hugely impressive thrown in. That's a 50% jump, for seemingly 1.5 times the effectiveness. A second attack bumps it up to 3 times the effectiveness (200% jump), damage increases push it up yet further, if you're a fighter there's another jump from attacks alone, the list goes on. Magic is even more blatant.

[Tanarii]

In d20 terms, if we assume that unskilled is +1 (average attribute of 12-13 on 4d6k3) that makes skilled +5 and skilled & expert +9 at level one, with +12 at the high range.I was talking about 5e numbers ... ability score 8 (-1), ability score 14 15 with prof (+4) and expert (+6). vs DC 10, 15 and 20.

Although I agree the average skill level for a PC to make a check is +1 for a 12-13 based on the standard array, not -1.

[Akolyte01]

2d10 would break the game as-is.

Here's an AnyDice link for comparison: http://anydice.com/program/b5b3

The standard assumption, with small variations, is that you hit on an 8 or better, and the enemy hits on a 12. If you look at the create-a-monster table, monsters' AC jumps by 1 basically whenever a character with standard array that pumps their primary to 16 at level 1 and maxes the stat by level 8 *would* get an increase, so "CR-appropriate" always hits around 8. To-hit bonuses scale similarly, though obviously the assumption breaks down a little bit because of varied armor.

So with the current game expectations, 1d20 hits at least a 12 about 45% of the time, and hits an 8 around 65% of the time. Combined with criticals on 20s, it creates an assumption that a level-appropriate monster's DPR is about half of what's listed, and a character will do about 70%.

With 2d20, monsters still hit about 45% of the time, but any increase to AC causes this to drop sharply. Meanwhile, players hit almost 80% of the time, with no extra changes.

The game doesn't sit the standard for players and monsters at the same point. They fall at different points of the curve. It would create nigh-invulnerability for creatures that get minor bonuses above and beyond the norm, because hitting extremes on a curve is much harder than on a single larger dice.

Also: your crit numbers for "one dice is a 0" increase the chances of crit from 1% to 19. If you think of it as percentile dice, you would crit on every number that starts with a 0 (1-9), and every represented number that ends with a 0 (every multiple of 10 up to 100). 19 in total. Would be devastating, especially with the 80% chance to hit created by 2d10 already.


Okay, so if monsters are balanced to hit players on a 12 and be hit on an 8, that means they hit players 45% of the time and are hit by players 65% of the time. It's possible to use across-the-board monster adjustments to maintain stats near this ratio.

The chance of rolling a 12 or above with 2d10 vs. 1d20 is actually the same, so no adjustment would be needed here.
Giving +2 AC to every monster means they would be balance to be be hit on 10, which means they are hit by players 64% of the time rolling 2d10.

So in an encounter between a party of average PCs vs a monster with a matching challenge rating, +2 to monster AC would keep the encounter near statistically equivalent.

So what *would* change? The biggest differences would be that comparative advantages and disadvantages would carry a lot more weight. Players or monsters with higher than average AC for their level/CR become comparatively much harder to hit in a 2d10 system--1 point over the average means a player would be hit by monsters 9% less in a 2d10 system vs 5% less in a d20 system. The converse is true of players/monsters with lower than average AC. This also means small differences in comparative level become much more important. Groups of monsters, with lower stats than a single monster of equivalent CR, become weaker.

Additional bonuses beyond 'average' become less useful. A +1 to AC makes the biggest difference to a player with average AC, and much less to a player who already has very high or very low AC.