D&D General How would you redo 4e?

[EzekielRaiden]

Yeah, I'm not saying 4e did, particularly, though the tendency did exist at times. I actually think of the non-stacking advantage/disadvantage thing as a kind of design constraint which spurs one to more creative solutions.

My problem is...I've never actually seen anything I considered a particularly creative solution. I certainly agree that it calls for one. I just have been profoundly disappointed with every solution I've seen proposed.

"Cancellation" stacking is still stacking, and almost as much bookkeeping as just having modifiers directly.
"One-and-done," as 5e does, leads either to Advantage being too special to use, or (as actually happened) massively over-used and thus often worthless. (And then they went and even broke one-and-done with stuff like Elven Accuracy!)
I want to say Level Up uses a "roll more dice, take the highest for your bonus" approach, which doesn't actually seem like a meaningful time savings, obviating the main benefit.

My preferred solution is to make two separate bonus/penalty tracks. You have Advantage/Disadvantage, which affects the die itself, and you have Boosted/Broken (still looking for better terms there), which is a flat +/- 2. Some things merely give that status (e.g. flanking attacks are Boosted, so a Warlord-style "let an ally make a Boosted attack" at-will doesn't further improve it), but a few can go ~~to infinity and beyond~~ past that, allowing limited and special stacking.

E.g.: the rogue is Flanking an orc. Their attack rolls are Boosted. Their Warlord buddy uses the long-rest attack Hammer Into Anvil, which means any ally flanking with the Warlord (and the Warlord flanking with any ally) is Super-Boosted, getting +4 to attacks instead of +2.

"Advantage" is for things that are uncertain, but likely to bring out your best performance. "Boosted" is for things that are certain to be helpful, but to a generally lesser degree. Each has value. Getting Disadvantage-with-Boost would be something like "making a trick shot you've practiced to death....but now there's a chaotic cyclone making things worse." Getting Advantage-with-Broken would be something like making a coup de grace with an actually broken weapon, you're likely to perform well but have to overcome the inherent faults of the tool. Within each track, it cancels normally: Disadvantage would turn "Super-Advantage" into just regular Advantage and vice-versa, and exact opposites always cancel.

The limitations on stacking static bonuses would mostly apply to combat rolls. Non-combat rolls, such as skill checks, would have looser restrictions on incidental bonuses, to reflect their more varied applications. Those are generally a haggle kind of situation anyway, so hashing out the specific bonuses is expected.

 

[Voidmoji]

"Cancellation" stacking is still stacking, and almost as much bookkeeping as just having modifiers directly.

I don't have a problem with stacking modifiers. Like anything, it can get out of control, but when in control? It works just fine,.

 

[EzekielRaiden]

I don't have a problem with stacking modifiers. Like anything, it can get out of control, but when in control? It works just fine,.

Oh, certainly. I mostly said that because the most commonly claimed benefits of things like Advantage (binary, non-stacking) are:
A) it's impossible for stacking to get out of control if things simply can't stack,
B) all the time that would be spent tracking and tallying bonuses now goes to other gameplay, and
C) if there's only three states (bonus, penalty, standard), players can intuitively know what they're getting out of various buffs or risk-taking maneuvers.

I consider the "cancellation stacking" format (my term, meaning "tally up your buffs and debuffs, whichever is greater is what you have") to be a poor compromise that often ends up being the worst of both worlds. It has, by design, all the limitations of the "no stacking at all" rules (lacks nuance, can't represent degrees of benefit/detriment, easy to overuse and thus make worthless, etc.) But it gives up point C entirely (not that, IMO, Advantage was actually that intuitive to begin with; it's easy to know that it's good, but not how good it is), and weakens point B, possibly to the point of not actually saving any meaningful amount of time.

And that's...sort of the rub for this thing. We want to find something that preserves most of the speed, simplicity, and intuitive effect of the "no stacking" method, while regaining some of the nuances, degrees-of-effect, depth of design space, and room to grow so bonuses are at least almost always actually bonuses and not wasted. But I find instead that sacrifices are made on both ends, giving us more than half of the problems and less than half of the benefits of both approaches.

And, to be clear, I don't consider my proposed solution very clever either. It's just trying to...sort of play the weaknesses of the two methods against one another so they end up cancelling out and thus being a minor improvement on most fronts. It's still pretty simple (either ±4, ±2, or 0; either best/worst of 3d20, best/worst of 2d20, or just 1d20), but not quite the ur-simplicity of "nothing stacks." It's nuanced enough to cover unusual situations (like having Super-Disadvantage and Super-Boost, meaning you have a very high floor but it's very unlikely that you'll do much better, or the reverse, where you're likely to get a high roll but you have a small chance of utterly bombing) without being a sprawling morass of modifiers. It allows for at least some design depth, but prevents things going off the deep end. And it should still be fairly quick and not particularly counter-intuitive, though it might still take some time to get used to.

It's just not really all that elegant, and still leaves things fairly shallow in terms of design space.

 

[Voidmoji]

I consider the "cancellation stacking" format (my term, meaning "tally up your buffs and debuffs, whichever is greater is what you have") to be a poor compromise that often ends up being the worst of both worlds. It has, by design, all the limitations of the "no stacking at all" rules (lacks nuance, can't represent degrees of benefit/detriment, easy to overuse and thus make worthless, etc.)

Not sure how it doesn't allow for degrees, but otherwise, this has not been a problem for me in 4e or my own game, really. I consider it a perfectly fine compromise. YMMV.

 

[EzekielRaiden]

Not sure how it doesn't allow for degrees, but otherwise, this has not been a problem for me in 4e or my own game, really. I consider it a perfectly fine compromise. YMMV.

Er, 4e does not use "cancellation stacking," as I defined the phrase. 4e just has ordinary modifiers. To whit...

4e: You can get many different bonuses of different sizes or which affect how many dice you roll.

5e: Non-stacking. Either you have Advantage, or you don't, same for Disadvantage. If you have both Advantage and Disadvantage, then you have no effect.

Other systems/supplements others have described: "Cancellation stacking." Either you have Advantage or you don't, same for Disadvantage. If you have both, you must count how many sources of Advantage and Disadvantage you have. Whichever one is larger, is what you have. Only if they are precisely equal do you end up with no effect.

(IIRC) Level Up/A5e: "Best of" stacking (again, my phrase.) You can stack up Expertise Dice, but you only get one die to add. More dice simply give you a better chance of getting the best possible bonus (+6.) Some versions of this could include cancellation as well, I don't know the specific implementation in A5e.

In my opinion, "non-stacking" (how 5e does it) throws out the baby with the bathwater, and the efforts at compromise feel like they end up throwing out the baby and keeping some bathwater. 4e does not use "cancellation stacking" nor "best-of" stacking. It just lets modifiers stack as they are, good or bad.

 

[AbdulAlhazred]

Hmm, lets think about this...

My problem is...I've never actually seen anything I considered a particularly creative solution. I certainly agree that it calls for one. I just have been profoundly disappointed with every solution I've seen proposed.

"Cancellation" stacking is still stacking, and almost as much bookkeeping as just having modifiers directly.

Well, it avoids all the adding and subtracting and looking endlessly for one more thing. Generally on the GM side if the bad guys have advantage or conversely grant the PCs disadvantage, there's a short list of ways that can happen (if your monster/scenario design makes that a long list maybe you should sharpen up the design). In practice we have found it speeds up play when it replaces all situational modifiers, but YMMV I guess?

"One-and-done," as 5e does, leads either to Advantage being too special to use, or (as actually happened) massively over-used and thus often worthless. (And then they went and even broke one-and-done with stuff like Elven Accuracy!)
I want to say Level Up uses a "roll more dice, take the highest for your bonus" approach, which doesn't actually seem like a meaningful time savings, obviating the main benefit.

I hear what you are saying, OTOH we decided way back when it would be utilized in certain specific situations and not in basically any others, though there is some slight wiggle room outside of combat. It hasn't presented an issue at this point. It comes up enough to matter, but not enough to become just an ordinary thing. Given that I am the one designing my game, I don't have to let things get borked up.

My preferred solution is to make two separate bonus/penalty tracks. You have Advantage/Disadvantage, which affects the die itself, and you have Boosted/Broken (still looking for better terms there), which is a flat +/- 2. Some things merely give that status (e.g. flanking attacks are Boosted, so a Warlord-style "let an ally make a Boosted attack" at-will doesn't further improve it), but a few can go ~~to infinity and beyond~~ past that, allowing limited and special stacking.

E.g.: the rogue is Flanking an orc. Their attack rolls are Boosted. Their Warlord buddy uses the long-rest attack Hammer Into Anvil, which means any ally flanking with the Warlord (and the Warlord flanking with any ally) is Super-Boosted, getting +4 to attacks instead of +2.

I just find this to be too much mental effort for my 60-year-old brain to put it bluntly. When I play I want things to MOVE and FLOW and just not get hung up on if something did or didn't stack and focused on adding up 6 different tiny little bonuses. So, maybe that's part of why my design feels right for HoML, because it is not a game of little nitty gritty stuff. The heroes are not grubbing for points. They are being clever, and deploying skill in terms of obtaining advantage whenever they can, and in how they deploy their power points, etc. Combat is supposed to be tactical, but you are more likely to do crazy stuff and focus less on small numbers of opponents. The '5x5' format of 4e combats is fine, but I find it too limited.

"Advantage" is for things that are uncertain, but likely to bring out your best performance. "Boosted" is for things that are certain to be helpful, but to a generally lesser degree. Each has value. Getting Disadvantage-with-Boost would be something like "making a trick shot you've practiced to death....but now there's a chaotic cyclone making things worse." Getting Advantage-with-Broken would be something like making a coup de grace with an actually broken weapon, you're likely to perform well but have to overcome the inherent faults of the tool. Within each track, it cancels normally: Disadvantage would turn "Super-Advantage" into just regular Advantage and vice-versa, and exact opposites always cancel.

The limitations on stacking static bonuses would mostly apply to combat rolls. Non-combat rolls, such as skill checks, would have looser restrictions on incidental bonuses, to reflect their more varied applications. Those are generally a haggle kind of situation anyway, so hashing out the specific bonuses is expected.

Well, I have a bit more specifically specified process for how a situation is mapped onto a roll, though I don't think it really is much different from what you are saying in practice.

In any case, you'd have to play it to see, as in some sense there is terminological variation here too. Like, your 'fixed' bonuses CAN actually be dependent on situation. Permanent bonuses in particular frequently have things like keyword bindings, so "Cold Iron Knife: +3 permanent bonus to attack and damage rolls vs fey creatures" is a thing. Now, you might have some other +3 permanent bonus and not care about this, but chances are you don't, so its likely there's going to be a bit of figuring at the start of a fight, though I tend to stick to a theme for at least the arc of a challenge or quest, so you probably just write down your total adjusted base bonus the first time this comes up. That's a bit like your 'boosted' kind of idea, but more restricted. I guess technically powers could play games with this, but I have not really wanted to go down that path, beyond maybe some that do extra damage vs certain keywords.

 

[AbdulAlhazred]

Er, 4e does not use "cancellation stacking," as I defined the phrase. 4e just has ordinary modifiers. To whit...

4e: You can get many different bonuses of different sizes or which affect how many dice you roll.

5e: Non-stacking. Either you have Advantage, or you don't, same for Disadvantage. If you have both Advantage and Disadvantage, then you have no effect.

Other systems/supplements others have described: "Cancellation stacking." Either you have Advantage or you don't, same for Disadvantage. If you have both, you must count how many sources of Advantage and Disadvantage you have. Whichever one is larger, is what you have. Only if they are precisely equal do you end up with no effect.

(IIRC) Level Up/A5e: "Best of" stacking (again, my phrase.) You can stack up Expertise Dice, but you only get one die to add. More dice simply give you a better chance of getting the best possible bonus (+6.) Some versions of this could include cancellation as well, I don't know the specific implementation in A5e.

In my opinion, "non-stacking" (how 5e does it) throws out the baby with the bathwater, and the efforts at compromise feel like they end up throwing out the baby and keeping some bathwater. 4e does not use "cancellation stacking" nor "best-of" stacking. It just lets modifiers stack as they are, good or bad.

The only thing 4e did, which might hint at another possible general solution is to define certain 'conditions' that exist, mostly CA, which always grant a specific bonus (+2). You could get CA from 12 sources, but you still only have CA(+2). Not that 4e really developed that idea, but it is kind of a nascent version of your 'boosted/busted' thing. The only difference being, in principle you could have a dozen of these 'conditions', though obviously more than a very few would probably be kind of silly. I mean, most of the formal 4e conditions work similarly, you can only be 'dazed' and get the penalty from it one time.

So, is there a more generalizable concept there?

 

[W'rkncacnter]

Level Up/A5e: "Best of" stacking (again, my phrase.) You can stack up Expertise Dice, but you only get one die to add. More dice simply give you a better chance of getting the best possible bonus (+6.) Some versions of this could include cancellation as well, I don't know the specific implementation in A5e.

expertise dice in a5e start at 1d4 and stack up to 1d8 (or 1d12 if you have certain features)

 

[EzekielRaiden]

expertise dice in a5e start at 1d4 and stack up to 1d8 (or 1d12 if you have certain features)

Ah, I must have conflated that with some other thing then. I know I've seen the described approach somewhere.

 

[Atomoctba]

Ah, I must have conflated that with some other thing then. I know I've seen the described approach somewhere.

Perhaps conflating it with boons and banes from the Shadow of Demon Lord?