Nice and very accurate that a d20 is swingy. Your system reminds me a little of Shift and spark bound. Similar ideas I have seen in Dodge, Block, Parry.
I would argue though that a lot of your critique comes down to the fictional positioning of your character. If you are a master thief why do you even need to roll to open the door. Is there danger or context that makes it necessary. If you do fail is it because of skill, luck or something outside your control. The narrative position in this case handles the degree of complications
I agree with this sentiment 100%. If your character is competent, don't roll at all unless there is something endangering/interfering with the process.
I spent some time looking over Block, Dodge, Parry, it’s quite interesting, and I can see the value in your argument. If something is likely given the fiction, then by all means keep the ball rolling.
A master thief doesn’t need to prove themselves on every lock. As a solitaire player though, the roll itself is part of the experience, it’s tactile, and routine. There’s no GM, no table energy, no other players reacting to what happens. The dice are doing some of that work; the anticipation, the result, the texture of the moment. Simply passing because it’s likely is narratively clean, and works quite well. But it can function as a silent drain with solitaire play. The skill curve tries to honor the element of surprise, the roll still happens, but the fiction of who your character is (applied as weight) shapes what’s possible before the dice ever hit the table.
I do like the bell curve you use and stepping down the dice depending on skill is a clean progression.
I think what a lot of skill based resolutions sometimes miss is outcomes besides success and failure. That's what I mean with fictional positioning. I decide if I roll but also what can happen. If a master thief fails do they fail because of skill or do they actually succeed but are undermined because of what they find or they take longer than expected etc. But I prefer pbta style approach to skill based approach
“But if you start applying “weight”, by manipulating the odds, (+1, -2 etc) that perfect curve gets warped. The weight acts like a magnet, dragging the results toward a specific number and skewing the distribution slightly or even so heavily that it lopsides the bell.”
This is flat out wrong. The curve of 2d6+1 is shifted up by 1, but it’s the same shape. Same for any other dice or modifier.
You’re correct that a flat modifier preserves the curve’s shape mathematically; 2d6+1 is just 2d6 shifted up. What I’m saying, and perhaps should have clarified for the purists, is that weight applied against a fixed threshold changes the distribution of outcomes that matter, not the shape of the raw curve itself. Enough negative weight pushes the meaningful success range out of reach entirely, which in practice produces the lopsided effect I was describing, even if the underlying bell curve remains intact.
However, if you visualize the curve as outcomes, and apply a -5 for example, the 2 becomes most common result, skewing the results and warping the curve because we are cutting off the tail end. We do this because we’re not interested in any values above or below a certain threshold.
Here is what you actually did if you look at the math:
2d12 14+ 45.8%
2d10 12+ 43.7%
2d8 10+ 43.7%
2d6 8+ 41.7%
2d4 6+ 37.%
From 2d12 to 2d6, your system basically does not give the PC a meaningful improvement in the chance to succeed. A 4% difference means that one in 25 checks will be changed from failure to succeess.
If I use that skill once per session and play one session per week, I will get the benefit once in six months.
Basically you’re making skill levels irrelevant, players can feel better by improving the numbers on the sheet but in actual play, there is very little measurable difference.
Also, rolling 2d10 vs. 12+ and 2d10+2 vs. 14+ is mathematically equivalent. The modifiers don’t “squish the curve”, that’s nonsense.
So I don’t know why you’re changing the target number. You can keep the target number constant at 14+ and just add a modifier on the PC side. Same thing mathematically but less complicated in play.
But the really important message here is: You need to look at gameplay outcomes. If your system creates a meaningful gameplay outcome between a success on a 10 and a success on 11, great. But if both just mean “you open the lock, moving on”, you are fiddling with something that does not change the game at all.
Yes some designers really care because the die result is something you physically see on the table, but does that translate to something in the fictional layer of the game.
If your game only processes the die result as “pass” and “fail”, it doesn’t matter what the distribution looks like inside those buckets.
Also the more skilled I am, the HIGHER the chance of a critical failure?!
Thank you for your comments, you raise some interested points. You’re right that the raw success percentages are closer than they might feel at the table, that’s a fair reading of the math. At the table the system works quite well and it’s quite functional and realistic, but on paper, where it lacks fiction and storyline, I can see where it seems off.
Where I’d push back is that the system isn’t designed around pass/fail as its primary output. The result ladder between a natural 2 and a natural max is where the meaningful differentiation lives, not just whether you opened the lock, but how cleanly, at what cost, with what consequence. A journeyman scraping through at 10 and an expert landing at 11 might both open the lock, but one of them left marks, made noise, or spent time they didn’t have.
On the critical failure point, you are again correct, and I addressed as well as offered direction and a fix for that in the post. At master tier a natural 2 softens narratively. The fiction overrides the mechanic when it needs to. That’s a design choice, not an oversight.
The fixed modifier versus shifting target debate is mathematically equivalent, we can certainly agree on that. However, I chose the shifting target because it’s more intuitive at the table for solo play. Fewer numbers to track, more tactile clarity about where you stand.
The deeper question you’re raising, at least to me, is does any of this translate to the fictional layer? The answer is yes, but only if the player is engaging with the full result ladder rather than reducing it to binary. That’s a player behavior question as much as a design question. So all valid points, however, when you strip away the narrative and fiction, i.e. the overlay, all that remains is clear math. The input, I.e. narrative and fiction is what turns the wheels and makes it functional.
I still run d20 at my tables but have been translating "Advantage/Disadvantage" into something very like what you describe here and letting players keep *any* 2 of the dice they roll -- meaning two high-rolling Advantage dice in a pool of, say, 4 or 5 dice can save you from a garbage d20 roll! The floor is raised significantly, and the swing of the d20 matters less and less the more you go out of your way to interact with the environment and prepare assets ahead of time.
That said, I've still been using Traits and Skills as flat modifiers to further raise the floor and keep things feeling familiar to my players... But you may have convinced me to translate Skills into dice as well as have Traits be the only flat modifiers left...
Your idea is well suited and thought out. I love learning about how other people navigate these kinds of issues. I’m glad you like this system, I appreciate the comment and I really hope you use it. Let me know how it goes and how you implement it if you do!
Thanks for this! I'm mediocre at envisioning probabilities and this was really clarifying. I've thinking about ways to make dice rolls feel excitingly random without feeling unfair, and the "weight" mechanic is a cool approach.
Glad you found this useful! I try to explain things in a way that I’d like things explained to me. I’m not the sharpest, but I can grasp just about anything as long as I can understand what’s being said about it.
Thats the way to defeat the iron rule of the dice throw: its not whim, likes/dislikes, or freedom from diaster. Winning always gets boreing...but losing also is boreing...
That’s exactly right. When everything get’s boring we can tip the scales and redirect our own intention by throwing all the weight somewhere off center.
Nice and very accurate that a d20 is swingy. Your system reminds me a little of Shift and spark bound. Similar ideas I have seen in Dodge, Block, Parry.
I would argue though that a lot of your critique comes down to the fictional positioning of your character. If you are a master thief why do you even need to roll to open the door. Is there danger or context that makes it necessary. If you do fail is it because of skill, luck or something outside your control. The narrative position in this case handles the degree of complications
I agree with this sentiment 100%. If your character is competent, don't roll at all unless there is something endangering/interfering with the process.
I spent some time looking over Block, Dodge, Parry, it’s quite interesting, and I can see the value in your argument. If something is likely given the fiction, then by all means keep the ball rolling.
A master thief doesn’t need to prove themselves on every lock. As a solitaire player though, the roll itself is part of the experience, it’s tactile, and routine. There’s no GM, no table energy, no other players reacting to what happens. The dice are doing some of that work; the anticipation, the result, the texture of the moment. Simply passing because it’s likely is narratively clean, and works quite well. But it can function as a silent drain with solitaire play. The skill curve tries to honor the element of surprise, the roll still happens, but the fiction of who your character is (applied as weight) shapes what’s possible before the dice ever hit the table.
I do like the bell curve you use and stepping down the dice depending on skill is a clean progression.
I think what a lot of skill based resolutions sometimes miss is outcomes besides success and failure. That's what I mean with fictional positioning. I decide if I roll but also what can happen. If a master thief fails do they fail because of skill or do they actually succeed but are undermined because of what they find or they take longer than expected etc. But I prefer pbta style approach to skill based approach
“But if you start applying “weight”, by manipulating the odds, (+1, -2 etc) that perfect curve gets warped. The weight acts like a magnet, dragging the results toward a specific number and skewing the distribution slightly or even so heavily that it lopsides the bell.”
This is flat out wrong. The curve of 2d6+1 is shifted up by 1, but it’s the same shape. Same for any other dice or modifier.
You’re correct that a flat modifier preserves the curve’s shape mathematically; 2d6+1 is just 2d6 shifted up. What I’m saying, and perhaps should have clarified for the purists, is that weight applied against a fixed threshold changes the distribution of outcomes that matter, not the shape of the raw curve itself. Enough negative weight pushes the meaningful success range out of reach entirely, which in practice produces the lopsided effect I was describing, even if the underlying bell curve remains intact.
However, if you visualize the curve as outcomes, and apply a -5 for example, the 2 becomes most common result, skewing the results and warping the curve because we are cutting off the tail end. We do this because we’re not interested in any values above or below a certain threshold.
Here is what you actually did if you look at the math:
2d12 14+ 45.8%
2d10 12+ 43.7%
2d8 10+ 43.7%
2d6 8+ 41.7%
2d4 6+ 37.%
From 2d12 to 2d6, your system basically does not give the PC a meaningful improvement in the chance to succeed. A 4% difference means that one in 25 checks will be changed from failure to succeess.
If I use that skill once per session and play one session per week, I will get the benefit once in six months.
Basically you’re making skill levels irrelevant, players can feel better by improving the numbers on the sheet but in actual play, there is very little measurable difference.
Also, rolling 2d10 vs. 12+ and 2d10+2 vs. 14+ is mathematically equivalent. The modifiers don’t “squish the curve”, that’s nonsense.
So I don’t know why you’re changing the target number. You can keep the target number constant at 14+ and just add a modifier on the PC side. Same thing mathematically but less complicated in play.
But the really important message here is: You need to look at gameplay outcomes. If your system creates a meaningful gameplay outcome between a success on a 10 and a success on 11, great. But if both just mean “you open the lock, moving on”, you are fiddling with something that does not change the game at all.
Yes some designers really care because the die result is something you physically see on the table, but does that translate to something in the fictional layer of the game.
If your game only processes the die result as “pass” and “fail”, it doesn’t matter what the distribution looks like inside those buckets.
Also the more skilled I am, the HIGHER the chance of a critical failure?!
Thank you for your comments, you raise some interested points. You’re right that the raw success percentages are closer than they might feel at the table, that’s a fair reading of the math. At the table the system works quite well and it’s quite functional and realistic, but on paper, where it lacks fiction and storyline, I can see where it seems off.
Where I’d push back is that the system isn’t designed around pass/fail as its primary output. The result ladder between a natural 2 and a natural max is where the meaningful differentiation lives, not just whether you opened the lock, but how cleanly, at what cost, with what consequence. A journeyman scraping through at 10 and an expert landing at 11 might both open the lock, but one of them left marks, made noise, or spent time they didn’t have.
On the critical failure point, you are again correct, and I addressed as well as offered direction and a fix for that in the post. At master tier a natural 2 softens narratively. The fiction overrides the mechanic when it needs to. That’s a design choice, not an oversight.
The fixed modifier versus shifting target debate is mathematically equivalent, we can certainly agree on that. However, I chose the shifting target because it’s more intuitive at the table for solo play. Fewer numbers to track, more tactile clarity about where you stand.
The deeper question you’re raising, at least to me, is does any of this translate to the fictional layer? The answer is yes, but only if the player is engaging with the full result ladder rather than reducing it to binary. That’s a player behavior question as much as a design question. So all valid points, however, when you strip away the narrative and fiction, i.e. the overlay, all that remains is clear math. The input, I.e. narrative and fiction is what turns the wheels and makes it functional.
I still run d20 at my tables but have been translating "Advantage/Disadvantage" into something very like what you describe here and letting players keep *any* 2 of the dice they roll -- meaning two high-rolling Advantage dice in a pool of, say, 4 or 5 dice can save you from a garbage d20 roll! The floor is raised significantly, and the swing of the d20 matters less and less the more you go out of your way to interact with the environment and prepare assets ahead of time.
That said, I've still been using Traits and Skills as flat modifiers to further raise the floor and keep things feeling familiar to my players... But you may have convinced me to translate Skills into dice as well as have Traits be the only flat modifiers left...
Your idea is well suited and thought out. I love learning about how other people navigate these kinds of issues. I’m glad you like this system, I appreciate the comment and I really hope you use it. Let me know how it goes and how you implement it if you do!
Thanks for this! I'm mediocre at envisioning probabilities and this was really clarifying. I've thinking about ways to make dice rolls feel excitingly random without feeling unfair, and the "weight" mechanic is a cool approach.
Glad you found this useful! I try to explain things in a way that I’d like things explained to me. I’m not the sharpest, but I can grasp just about anything as long as I can understand what’s being said about it.
Thats the way to defeat the iron rule of the dice throw: its not whim, likes/dislikes, or freedom from diaster. Winning always gets boreing...but losing also is boreing...
That’s exactly right. When everything get’s boring we can tip the scales and redirect our own intention by throwing all the weight somewhere off center.
Might just bulid musicle mass yeeting the weights...:)