And that is in and of itself an issue. there should be more at stake for some rolls than "Win/Lose". Degrees of success can be as important as simple pass/fail. I know many DMs who use d20 rolls to determine how long a task takes as much as whether the task succeeds, and even in D&D there's often gradations of success or failure. A module may ask for a History check and offer increasingly specific information at DCs 9, 12, and 16.
Nobody's saying there's no place for the d20. Some tasks would more befit a d20 roll than a 2d10 roll. But for basic skills? Those nights where you can't roll above a 5 to save your life and your Awesome Hero can't even put their pants on straight without fumbling a dice check are awful. It can, does, and has thrown people out of the story when they simply can't catch a break. A more stable skill system, even as just a variant rule, might be exactly what the doctor ordered for tables that hate how lame and incompetent the ridiculous swinginess of the d20 makes everybody look.
D&D 5E doesn't use critical successes or failures outside attack rolls. A 20 has the exact same outcome as a 15, as long as that 15 is sufficient to meet the DC. A 1 has the exact same outcome as a 5, as long as that 5 is insufficient to meet the DC. "All possible outcomes" (optional or homebrew rules notwithstanding) is two outcomes: success or failure. And unless you need to roll 11 or higher for success, they're not on exactly equal footing.
That's a good point.
I would love to see a statistical analysis of something relatively simple, like a contested Strength roll. Between two characters, one possessing a str modifier of +0 vs. another character with a str modifier +1. I think that the 2d10 method would more reliably result in a success for the stronger character, but me no math good enough to prove just how much of a difference it would make vs. 1d20.
I think the biggest change in the D&D system would show up in Advantage/Disadvantage rolls. Probably 2d10 results in a smaller likely spread of results in that case as well.
Every time I roll a d20, I am just as likely (percentage-wise) to end up with a critical failure
I would point out that the likelihood of that happening is exactly 0 as there is no such concept as a critical failure in 5e. :p
The hilarious thing is with this, most groups don’t know this. If you have a +9 to your CON save, you can’t lose concentration on any spell that does 20 points of damage or less. If you have expertise with acrobatics/athletics, you can just swim, or do basic flips, without fear of failure once your bonus is a +10. If you get a Nat 1 on an attack, you just miss. You don’t break your sword or throw it into the multiverse, you just miss.
Shit, with a Nat 20 on a skill check, that doesn’t mean you win either. You just add a 20. The -2 Charisma fighter who rolls a 20 then has a 18 persuasion check. Still pretty high and might persuade most commoners, but typically won’t do much to a trained merchant, or noble. Definitely not enough for the memes of romancing a dragon who is proficient and naturally charismatic.
My group does do crit success/fail on saving throws because to us, as a group that came from many older editions it makes sense. That way there is always a minor point of success(to the point earlier, celebrating those nat 20s and cheering them on as a group). We don’t do it on skill checks though, because it doesn’t make sense that the untrained lockpick can get through the vault of the king on pure luck.
By using any method that makes the average more common and reduces the chance of getting a 19 or 20 your making the higher rated armor more powerful. I it takes the fighter a 17 or above to hit armor class X, it becomes more difficult when you make these roles less common. You have to rebalance all of your fights/armor and weapon bonuses. All of the other systems are designed with this in mind.
1) And that is in and of itself an issue. there should be more at stake for some rolls than "Win/Lose". Degrees of success can be as important as simple pass/fail. I know many DMs who use d20 rolls to determine how long a task takes as much as whether the task succeeds, and even in D&D there's often gradations of success or failure. A module may ask for a History check and offer increasingly specific information at DCs 9, 12, and 16.
2) Nobody's saying there's no place for the d20. Some tasks would more befit a d20 roll than a 2d10 roll. But for basic skills? Those nights where you can't roll above a 5 to save your life and your Awesome Hero can't even put their pants on straight without fumbling a dice check are awful. It can, does, and has thrown people out of the story when they simply can't catch a break. A more stable skill system, even as just a variant rule, might be exactly what the doctor ordered for tables that hate how lame and incompetent the ridiculous swinginess of the d20 makes everybody look.
1) Absolutely. I do some of those things myself. My point is that changing the system from being based on a single die to multiple dice changes nothing about this in and of itself.
2) We've all been there presumably, but I think that's too limited a picture. Yes, D&D being so swingy makes it possible to fail often, more often than statistically likely, but it also makes it possible to try something difficult and still have maybe 25-30% chance of success. And if you fail at basic stuff a lot I would suspect your DM is calling for rolls they should just handwaive, but maybe that's just me. Bell curve outcomes make it less likely to fail easy stuff, but also even more likely to fail at the harder stuff. And in the end it all still and always translates into odds. Your chance of success isn't just determined by what and how many dice you roll, but also by how hard the DM makes the task. And by how hard, I mean how likely to fail. You need 11 or better? 50% chance of failure with 1d20, 45% with 2d10. You need 6 or better? 25% chance of failure with 1d20, 10% with 2d10. However, what you need is up to the DM. If it's something they feel should only fail in 10% of attempts, they a) probably won't (or shouldn't) call for a roll and b) should just set the DC accordingly. Discussions about dice pools vs single die rolls almost always assume DCs (or TNs, or whatever) would be the same in both cases. Why would that necessarily be true? My advice to DMs out there: stop thinking about DCs purely in terms of abstract difficulty and more in terms of odds. DC 15 is 70% chance of failure for Joe Q Average without relevant skill or knowledge (using 1d20). "Seven out of ten attempts fail if you ask a rando you pick up on the street to try it" is meaningful. DC 15 is abstract gibberish.
I'm not saying there is no value in reliability but in D&D, well, there arguably isn't much (at least if the DM sets DCs that translate to realistic odds). Looking at older editions of Legend of the Five Rings (this actually translates reasonably well to the last, 5th, edition as well but since narrative dice take longer to explain I'm passing on that), dice pools of d10s determine outcomes (how many you roll and how many you keep depend on your skill and Ring ranks, and possibly some added effect), with the GM setting a TN according to difficulty. If you roll a handful or more d10s, you can have a fairly reliable idea of your odds of success, and L5R used the Raise mechanic to capitalize on that: if you wanted a better success, you could decide to raise the TN in increments of 5 and get additional benefits if you were successful. We could still do something similar with the d20 system, but the xd10 bell curve makes that a lot easier and more effective. L5R leverages that reliability to enhance its mechanics. That makes it valuable. There is no such thing in D&D though, and thus that value doesn't really exist either.
Rollback Post to RevisionRollBack
Want to start playing but don't have anyone to play with? You can try these options: [link].
I think it makes sense for skill checks. Otherwise you get those situations where the barbarian can't break down the door, but the wizard can.
But combat is supposed to be random. If you use 2d10 for attack rolls, you're going to get into a lot of fights where you either hit nearly every time, or almost never.
> Let's start a fight: > 2d10 🎲 makes more sense than 1d20 - change my mind.
d10s are the ugliest dice* in the pandaverse, whereas the d20 is a beautiful die.
If the title of the post was suggesting 2d12, my interest would have been piqued and I might have bothered actually reading it ;)
*Edit: OK, d8's are possibly on a par, or even worse, but that is an aside. Also d4s are admittedly an acquired taste.
No, you are correct. d10s are objectively the worst because they're non-Platonic. Like, if all the faces aren't congruent regular polygons, what are you even doing?
For a math junkie like me, there is sooo much to comment on here. But let's start with some of the less mathy stuff.
In the original D&D, the D20 was marked 0-9 twice. There were two "0"s, two "1"s, ... We colored half the die with a marker to distinguish which values were the "tens" and left the other half of the faces alone. So although they had twenty faces, they could be used as D10s, and were. I don't believe my original set of dice included a d10, much less the "percentile die." But I like the contemporary dice better.
Lyxen is correct in that RAW 5e does not have a critical failure. But I haven't played with anyone recently that didn't use critical failures, and usually for skills just like combat. Different strokes for different folks. Next time I DM I am going to have to think on whether we'll have them or not. But one thing I like that I am going to do is HDYWTDT (Matt Mercer style). I think that is a wonderful idea.
So about the math ... Everything about this thread is about the shape of the probability curve. Do we want a flat curve, a linear peak or something that looks like a bell shaped curve? Let's not get into whether it is a "standard" probability curve or distribution. It isn't and nobody I know wants it to be a standard distribution. That's too much work.
If you want to have a bell shaped curve I think there is a lot of merit in that, but I suspect it will drive down the successful hits (both ways) and this will cause combat to last longer. So if you want to have the bell shaped curve (using three or more dice) just remember this will also happen. And if you like the excitement of the crits, understand this will reduce the number of crits. If you start monkeying around with the system so the number of crits stays about the same, then you have subverted the reason you went down the bell shaped curve pursuit at all.
The 1d20 produces 20 results. 2d10 produces 19 results. 3d6 produces 16 results. 4d6 produces 21 results. 1d10&2d6-2 produce 20 results. This combination also produces a good bell shaped curve with a median value of 11. But in my experience, many of the folks in the target age don't want to have to add the dice together, and then add or subtract modifiers. So now we're getting back to the idea of just rolling a single die, a d20.
And in the end, I think folks are having fun with the simple d20 regardless of how many times it comes up 20 or 1 in an evening.
If you have a table that wants to have more "expected" results, then I think you should look into playing with a dice pool. A dice pool also allows a lot of other fun with modifiers. But I don't see the game being reworked for dice pools. Like Yuri said, "It's a tradition."
Rollback Post to RevisionRollBack
Cum catapultae proscriptae erunt tum soli proscript catapultas habebunt
I think more possible outcomes are good for some results table, which can also be placed on a curve. In just regular rolling whether for table results or a task check, wouldn't the extremes be less frequent? But this isn't my expertise, so I'm writing more to test my assumptions rather than really answer your query.
If you're rolling to look up a result on a table (especially a table with a set number of results expecting a flat curve), use the dice the table was designed for.
Well, yes; but I'm thinking about table design where the middling numbers would be occupied by more middling encounters and the extreme ends of the array would indicate the more exceptional encounters (or item, or what have you). Pooled dice over one dice, I believe would allow for a more dynamic range of possibilities, if I'm following the discussion.
Rollback Post to RevisionRollBack
Jander Sunstar is the thinking person's Drizzt, fight me.
2d10 only gives you 19 possible outcomes with a standard bell curve ass opposed to the 1d20’s 20 variables with the flat line.
3d6 only gives you 15 possible outcomes with a very steep curve.
2d12 gives 23 possible variables with the same bell curve as 2d10.
Why is “more possible outcomes” an inherent good?
If you don't want more possible outcomes, you can just not use dice at all. The more dice you use, the closer you get to that.
Sorry, but that's completely nonsensical. D&D tends toward only two outcomes, success or failure. Many GMs will expand this slightly, but no one actually has something different in mind for each one of the twenty possible results of a d20 roll. So 3d6 has 16 degrees of graduation rather than 20. Why is that bad? What do those four additional numbers do for the game?
2d10 only gives you 19 possible outcomes with a standard bell curve ass opposed to the 1d20’s 20 variables with the flat line.
3d6 only gives you 15 possible outcomes with a very steep curve.
2d12 gives 23 possible variables with the same bell curve as 2d10.
Why is “more possible outcomes” an inherent good?
If you don't want more possible outcomes, you can just not use dice at all. The more dice you use, the closer you get to that.
Sorry, but that's completely nonsensical. D&D tends toward only two outcomes, success or failure. Many GMs will expand this slightly, but no one actually has something different in mind for each one of the twenty possible results of a d20 roll. So 3d6 has 16 degrees of graduation rather than 20. Why is that bad? What do those four additional numbers do for the game?
They push the result into 20 equally random results. Having a +5 modifier on a roll of 1d20 is dramatically different from having a +5 modifier on 3d6. That sort of thing would necessitate a significant rebalancing of the game.
Rollback Post to RevisionRollBack
Find your own truth, choose your enemies carefully, and never deal with a dragon.
"Canon" is what's factual to D&D lore. "Cannon" is what you're going to be shot with if you keep getting the word wrong.
2d10 only gives you 19 possible outcomes with a standard bell curve ass opposed to the 1d20’s 20 variables with the flat line.
3d6 only gives you 15 possible outcomes with a very steep curve.
2d12 gives 23 possible variables with the same bell curve as 2d10.
Why is “more possible outcomes” an inherent good?
If you don't want more possible outcomes, you can just not use dice at all. The more dice you use, the closer you get to that.
Sorry, but that's completely nonsensical. D&D tends toward only two outcomes, success or failure. Many GMs will expand this slightly, but no one actually has something different in mind for each one of the twenty possible results of a d20 roll. So 3d6 has 16 degrees of graduation rather than 20. Why is that bad? What do those four additional numbers do for the game?
They push the result into 20 equally random results. Having a +5 modifier on a roll of 1d20 is dramatically different from having a +5 modifier on 3d6. That sort of thing would necessitate a significant rebalancing of the game.
I disagree. A +5 modifier on a roll only shifts the curve, it doesn't change it. It's also functionally not really different from shifting the DC range down 5 points. 3d6 is different from 1d20 because of the curve, not because of modifier effects.
edit: also, 3d6 gives 16 distinct possible results, not 15. That's going to annoy me every time I see it.
Why is “more possible outcomes” an inherent good?
Also, two dice isn’t enough to approximate a normal distribution, so I’m not really sure 2d10 could be said to give a “standard bell curve.” 3d6 does a better job at achieving that, if “standard bell curve” is something you’re after.
Because part of the reason D&D feels a little samesame after a while is because of the bounded accuracy since everything has to fit in a relatively cramped space. With an approach like this, everything gets pulled towards the middle. The more dice (3dX vs 2dX), the gentler the pull, but still. In addition, more dice means a smaller probability it hitting a crit either way.
Systems with Dice Pools have two to three different scales. One of them dictates how many dice in a pool, the other dictates how many successes are required to reach which tier of success, and occasionally another scale which dictates the over/under on the roll. But D&D doesn’t have pools. In D&D, everything is tied into the one scale through modifiers. So 1d20 gives you a 20 point spread with an even 5% chance to hit any result. the modifiers give you between -1 to 17 additional points to add to the result, a 19 point spread. So everything fits between 0-37 because it has to. That makes it simple, but what is both the best and worst part of 5e in a great many ways, bounded accuracy. Any system with bounded accuracy has the same problem. It also means that Stats are just under half of the input, and luck is just over half. That makes if swingy. What are the most common comments about the d20 system? Simple, but swingy.
By moving to a 2d or 3d system, because of how much it pulls everything towards the middle, the total number of potential results is important inso much as it determines how many results there will be closest to the middle. Because of the spike/curve, the results in the middle 1/3 will always be the most likely. So, for example, in a 2d10 system there are 19 potential results so 1/3 is a 6 point spread. The vast majority of results will be between 8.5 and 13.5 , a spread of 6. So even though the modifiers and luck would share a 50/50 distribution for total influence over any potential roll, in reality luck will actually have way less influence on the vast majority of rolls. With almost 2/3 of the rolls going to that middle 1/3 of the spectrum, but with the modifiers staying the same, then everything ends up distilling to result in Expertise being even more powerful, and lack of proficiency bearing more of a hindrance towards ever getting higher than just below average. In addition, the probability of rolling a double 1 or a double 10 drops to just 1% for each. With 3d6 it’s slightly worse because the curve is even more predictable and the extremes even less likely at under 1/2% each.
But with 2d12 it shifts the other way and rounds up to an 8 point spread which means that, by simply changing Expertise to 1.5x Proficiency instead of 2x and now that spread shifts to -1 to 14 and now luck accounts for around or just over 1/3 of the likely outcomes, but the extreme highs and lows are also still somewhat likely at over 2/3% This makes it more consistent and less swingy, but also still keeps luck relevant enough in the game. (3d8 compares to this approximately as well as 3d6 compares to 2d10.)
2d10 only gives you 19 possible outcomes with a standard bell curve ass opposed to the 1d20’s 20 variables with the flat line.
3d6 only gives you 15 possible outcomes with a very steep curve.
2d12 gives 23 possible variables with the same bell curve as 2d10.
Why is “more possible outcomes” an inherent good?
If you don't want more possible outcomes, you can just not use dice at all. The more dice you use, the closer you get to that.
Sorry, but that's completely nonsensical. D&D tends toward only two outcomes, success or failure. Many GMs will expand this slightly, but no one actually has something different in mind for each one of the twenty possible results of a d20 roll. So 3d6 has 16 degrees of graduation rather than 20. Why is that bad? What do those four additional numbers do for the game?
They push the result into 20 equally random results. Having a +5 modifier on a roll of 1d20 is dramatically different from having a +5 modifier on 3d6. That sort of thing would necessitate a significant rebalancing of the game.
I disagree. A +5 modifier on a roll only shifts the curve, it doesn't change it. It's also functionally not really different from shifting the DC range down 5 points. 3d6 is different from 1d20 because of the curve, not because of modifier effects.
edit: also, 3d6 gives 16 distinct possible results, not 15. That's going to annoy me every time I see it.
The issue is the curve itself. On a roll of 1d20, each +1 modifier increases your chance to succeed by 5% (barring, of course, having a modifier that's so high that you can't fail). On a curve, pushing each +1 does not grant the same percent chance modifier to your odds of success. Additionally, switching over to a bell curve dramatically weakens the Advantage/Disadvantage system: if your odds of rolling a 4 are lower than your odds of rolling a 10, having disadvantage on a roll suddenly becomes dramatically less of a threat. Likewise, having Advantage becomes less of a benefit.
Edit: or let me put it this way: with a d20, if you need a 14+ to succeed your odds are 35%. With 3d6 they drop down to 16.2%, more than 50% lower. With a bell curve, normal tasks become easier, hard tasks become harder, and easy tasks become trivial.
And that is in and of itself an issue. there should be more at stake for some rolls than "Win/Lose". Degrees of success can be as important as simple pass/fail. I know many DMs who use d20 rolls to determine how long a task takes as much as whether the task succeeds, and even in D&D there's often gradations of success or failure. A module may ask for a History check and offer increasingly specific information at DCs 9, 12, and 16.
Nobody's saying there's no place for the d20. Some tasks would more befit a d20 roll than a 2d10 roll. But for basic skills? Those nights where you can't roll above a 5 to save your life and your Awesome Hero can't even put their pants on straight without fumbling a dice check are awful. It can, does, and has thrown people out of the story when they simply can't catch a break. A more stable skill system, even as just a variant rule, might be exactly what the doctor ordered for tables that hate how lame and incompetent the ridiculous swinginess of the d20 makes everybody look.
Please do not contact or message me.
That's a good point.
I would love to see a statistical analysis of something relatively simple, like a contested Strength roll. Between two characters, one possessing a str modifier of +0 vs. another character with a str modifier +1. I think that the 2d10 method would more reliably result in a success for the stronger character, but me no math good enough to prove just how much of a difference it would make vs. 1d20.
I think the biggest change in the D&D system would show up in Advantage/Disadvantage rolls. Probably 2d10 results in a smaller likely spread of results in that case as well.
Oh.
Well never mind then.
I’m not going to fight you. Genesys is a far better RPG system than D&D; it works with a dice pool, and is totally worth a try! Would recommend.
That said, the fun of D&D is founded as much on nostalgia as on anything else. For that reason, I do like the d20!
Wizard (Gandalf) of the Tolkien Club
> Let's start a fight:
> 2d10 🎲 makes more sense than 1d20 - change my mind.
d10s are the ugliest dice* in the pandaverse, whereas the d20 is a beautiful die.
If the title of the post was suggesting 2d12, my interest would have been piqued and I might have bothered actually reading it ;)
*Edit: OK, d8's are possibly on a par, or even worse, but that is an aside. Also d4s are admittedly an acquired taste.
I think the d12 is the prettiest die. I mean d20 is OK, but the d12 surfaces are pentagons, and pentagons > triangles all day long.
WOTC lies. We know that WOTC lies. WOTC knows that we know that WOTC lies. We know that WOTC knows that we know that WOTC lies. And still they lie.
Because of the above (a paraphrase from Orwell) I no longer post to the forums -- PM me if you need help or anything.
The hilarious thing is with this, most groups don’t know this. If you have a +9 to your CON save, you can’t lose concentration on any spell that does 20 points of damage or less. If you have expertise with acrobatics/athletics, you can just swim, or do basic flips, without fear of failure once your bonus is a +10. If you get a Nat 1 on an attack, you just miss. You don’t break your sword or throw it into the multiverse, you just miss.
Shit, with a Nat 20 on a skill check, that doesn’t mean you win either. You just add a 20. The -2 Charisma fighter who rolls a 20 then has a 18 persuasion check. Still pretty high and might persuade most commoners, but typically won’t do much to a trained merchant, or noble. Definitely not enough for the memes of romancing a dragon who is proficient and naturally charismatic.
My group does do crit success/fail on saving throws because to us, as a group that came from many older editions it makes sense. That way there is always a minor point of success(to the point earlier, celebrating those nat 20s and cheering them on as a group). We don’t do it on skill checks though, because it doesn’t make sense that the untrained lockpick can get through the vault of the king on pure luck.
By using any method that makes the average more common and reduces the chance of getting a 19 or 20 your making the higher rated armor more powerful. I it takes the fighter a 17 or above to hit armor class X, it becomes more difficult when you make these roles less common. You have to rebalance all of your fights/armor and weapon bonuses. All of the other systems are designed with this in mind.
1) Absolutely. I do some of those things myself. My point is that changing the system from being based on a single die to multiple dice changes nothing about this in and of itself.
2) We've all been there presumably, but I think that's too limited a picture. Yes, D&D being so swingy makes it possible to fail often, more often than statistically likely, but it also makes it possible to try something difficult and still have maybe 25-30% chance of success. And if you fail at basic stuff a lot I would suspect your DM is calling for rolls they should just handwaive, but maybe that's just me. Bell curve outcomes make it less likely to fail easy stuff, but also even more likely to fail at the harder stuff. And in the end it all still and always translates into odds. Your chance of success isn't just determined by what and how many dice you roll, but also by how hard the DM makes the task. And by how hard, I mean how likely to fail.
You need 11 or better? 50% chance of failure with 1d20, 45% with 2d10. You need 6 or better? 25% chance of failure with 1d20, 10% with 2d10. However, what you need is up to the DM. If it's something they feel should only fail in 10% of attempts, they a) probably won't (or shouldn't) call for a roll and b) should just set the DC accordingly. Discussions about dice pools vs single die rolls almost always assume DCs (or TNs, or whatever) would be the same in both cases. Why would that necessarily be true? My advice to DMs out there: stop thinking about DCs purely in terms of abstract difficulty and more in terms of odds. DC 15 is 70% chance of failure for Joe Q Average without relevant skill or knowledge (using 1d20). "Seven out of ten attempts fail if you ask a rando you pick up on the street to try it" is meaningful. DC 15 is abstract gibberish.
I'm not saying there is no value in reliability but in D&D, well, there arguably isn't much (at least if the DM sets DCs that translate to realistic odds). Looking at older editions of Legend of the Five Rings (this actually translates reasonably well to the last, 5th, edition as well but since narrative dice take longer to explain I'm passing on that), dice pools of d10s determine outcomes (how many you roll and how many you keep depend on your skill and Ring ranks, and possibly some added effect), with the GM setting a TN according to difficulty. If you roll a handful or more d10s, you can have a fairly reliable idea of your odds of success, and L5R used the Raise mechanic to capitalize on that: if you wanted a better success, you could decide to raise the TN in increments of 5 and get additional benefits if you were successful. We could still do something similar with the d20 system, but the xd10 bell curve makes that a lot easier and more effective. L5R leverages that reliability to enhance its mechanics. That makes it valuable. There is no such thing in D&D though, and thus that value doesn't really exist either.
Want to start playing but don't have anyone to play with? You can try these options: [link].
I think it makes sense for skill checks. Otherwise you get those situations where the barbarian can't break down the door, but the wizard can.
But combat is supposed to be random. If you use 2d10 for attack rolls, you're going to get into a lot of fights where you either hit nearly every time, or almost never.
If you don't want more possible outcomes, you can just not use dice at all. The more dice you use, the closer you get to that.
No, you are correct. d10s are objectively the worst because they're non-Platonic. Like, if all the faces aren't congruent regular polygons, what are you even doing?
For a math junkie like me, there is sooo much to comment on here. But let's start with some of the less mathy stuff.
In the original D&D, the D20 was marked 0-9 twice. There were two "0"s, two "1"s, ... We colored half the die with a marker to distinguish which values were the "tens" and left the other half of the faces alone. So although they had twenty faces, they could be used as D10s, and were. I don't believe my original set of dice included a d10, much less the "percentile die." But I like the contemporary dice better.
Lyxen is correct in that RAW 5e does not have a critical failure. But I haven't played with anyone recently that didn't use critical failures, and usually for skills just like combat. Different strokes for different folks. Next time I DM I am going to have to think on whether we'll have them or not. But one thing I like that I am going to do is HDYWTDT (Matt Mercer style). I think that is a wonderful idea.
So about the math ... Everything about this thread is about the shape of the probability curve. Do we want a flat curve, a linear peak or something that looks like a bell shaped curve? Let's not get into whether it is a "standard" probability curve or distribution. It isn't and nobody I know wants it to be a standard distribution. That's too much work.
If you want to have a bell shaped curve I think there is a lot of merit in that, but I suspect it will drive down the successful hits (both ways) and this will cause combat to last longer. So if you want to have the bell shaped curve (using three or more dice) just remember this will also happen. And if you like the excitement of the crits, understand this will reduce the number of crits. If you start monkeying around with the system so the number of crits stays about the same, then you have subverted the reason you went down the bell shaped curve pursuit at all.
The 1d20 produces 20 results. 2d10 produces 19 results. 3d6 produces 16 results. 4d6 produces 21 results. 1d10&2d6-2 produce 20 results. This combination also produces a good bell shaped curve with a median value of 11. But in my experience, many of the folks in the target age don't want to have to add the dice together, and then add or subtract modifiers. So now we're getting back to the idea of just rolling a single die, a d20.
And in the end, I think folks are having fun with the simple d20 regardless of how many times it comes up 20 or 1 in an evening.
If you have a table that wants to have more "expected" results, then I think you should look into playing with a dice pool. A dice pool also allows a lot of other fun with modifiers. But I don't see the game being reworked for dice pools. Like Yuri said, "It's a tradition."
Cum catapultae proscriptae erunt tum soli proscript catapultas habebunt
🎵 TRADITION!🎵
Well, yes; but I'm thinking about table design where the middling numbers would be occupied by more middling encounters and the extreme ends of the array would indicate the more exceptional encounters (or item, or what have you). Pooled dice over one dice, I believe would allow for a more dynamic range of possibilities, if I'm following the discussion.
Jander Sunstar is the thinking person's Drizzt, fight me.
Sorry, but that's completely nonsensical. D&D tends toward only two outcomes, success or failure. Many GMs will expand this slightly, but no one actually has something different in mind for each one of the twenty possible results of a d20 roll. So 3d6 has 16 degrees of graduation rather than 20. Why is that bad? What do those four additional numbers do for the game?
They push the result into 20 equally random results. Having a +5 modifier on a roll of 1d20 is dramatically different from having a +5 modifier on 3d6. That sort of thing would necessitate a significant rebalancing of the game.
Find your own truth, choose your enemies carefully, and never deal with a dragon.
"Canon" is what's factual to D&D lore. "Cannon" is what you're going to be shot with if you keep getting the word wrong.
I disagree. A +5 modifier on a roll only shifts the curve, it doesn't change it. It's also functionally not really different from shifting the DC range down 5 points. 3d6 is different from 1d20 because of the curve, not because of modifier effects.
edit: also, 3d6 gives 16 distinct possible results, not 15. That's going to annoy me every time I see it.
Want to start playing but don't have anyone to play with? You can try these options: [link].
Because part of the reason D&D feels a little samesame after a while is because of the bounded accuracy since everything has to fit in a relatively cramped space. With an approach like this, everything gets pulled towards the middle. The more dice (3dX vs 2dX), the gentler the pull, but still. In addition, more dice means a smaller probability it hitting a crit either way.
Systems with Dice Pools have two to three different scales. One of them dictates how many dice in a pool, the other dictates how many successes are required to reach which tier of success, and occasionally another scale which dictates the over/under on the roll. But D&D doesn’t have pools. In D&D, everything is tied into the one scale through modifiers. So 1d20 gives you a 20 point spread with an even 5% chance to hit any result. the modifiers give you between -1 to 17 additional points to add to the result, a 19 point spread. So everything fits between 0-37 because it has to. That makes it simple, but what is both the best and worst part of 5e in a great many ways, bounded accuracy. Any system with bounded accuracy has the same problem. It also means that Stats are just under half of the input, and luck is just over half. That makes if swingy. What are the most common comments about the d20 system? Simple, but swingy.
By moving to a 2d or 3d system, because of how much it pulls everything towards the middle, the total number of potential results is important inso much as it determines how many results there will be closest to the middle. Because of the spike/curve, the results in the middle 1/3 will always be the most likely. So, for example, in a 2d10 system there are 19 potential results so 1/3 is a 6 point spread. The vast majority of results will be between 8.5 and 13.5 , a spread of 6. So even though the modifiers and luck would share a 50/50 distribution for total influence over any potential roll, in reality luck will actually have way less influence on the vast majority of rolls. With almost 2/3 of the rolls going to that middle 1/3 of the spectrum, but with the modifiers staying the same, then everything ends up distilling to result in Expertise being even more powerful, and lack of proficiency bearing more of a hindrance towards ever getting higher than just below average. In addition, the probability of rolling a double 1 or a double 10 drops to just 1% for each. With 3d6 it’s slightly worse because the curve is even more predictable and the extremes even less likely at under 1/2% each.
But with 2d12 it shifts the other way and rounds up to an 8 point spread which means that, by simply changing Expertise to 1.5x Proficiency instead of 2x and now that spread shifts to -1 to 14 and now luck accounts for around or just over 1/3 of the likely outcomes, but the extreme highs and lows are also still somewhat likely at over 2/3% This makes it more consistent and less swingy, but also still keeps luck relevant enough in the game. (3d8 compares to this approximately as well as 3d6 compares to 2d10.)
Make sense?
Creating Epic Boons on DDB
DDB Buyers' Guide
Hardcovers, DDB & You
Content Troubleshooting
The issue is the curve itself. On a roll of 1d20, each +1 modifier increases your chance to succeed by 5% (barring, of course, having a modifier that's so high that you can't fail). On a curve, pushing each +1 does not grant the same percent chance modifier to your odds of success. Additionally, switching over to a bell curve dramatically weakens the Advantage/Disadvantage system: if your odds of rolling a 4 are lower than your odds of rolling a 10, having disadvantage on a roll suddenly becomes dramatically less of a threat. Likewise, having Advantage becomes less of a benefit.
Edit: or let me put it this way: with a d20, if you need a 14+ to succeed your odds are 35%. With 3d6 they drop down to 16.2%, more than 50% lower. With a bell curve, normal tasks become easier, hard tasks become harder, and easy tasks become trivial.
Here's a good summary I found regarding the rules published in 3.5 Edition's Unearthed Arcana.
Find your own truth, choose your enemies carefully, and never deal with a dragon.
"Canon" is what's factual to D&D lore. "Cannon" is what you're going to be shot with if you keep getting the word wrong.