D20 Variance in practice

[noob]

I'm sure that is the only reason.

It is the main reason: dnd being famous means that
1: When people starts playing rpgs they start with dnd
2: The games you can find the most tables for are dnd games
3: Most people playing an rpg excepts a dnd like experience.
So it is normal to play dnd over other rpgs because it is way more convenient(you find a table easier) due to the game being famous and people except to play that rpg (so when you introduce them to rpgs you start by dnd because it is what they heard about).

It is how famous things works: famous things are more talked about and known and so stays or become more famous.

The fact there is one rpg that is famous is actually a good thing because it means there is more people playing rpgs in general: if dnd was not so much represented in fiction people would probably play rpgs in general less.

In terms of design dnd is in many ways horrible or over-convoluted or lacking in advice for the gm and players but design does not matters because people playing dnd often do not even read half of the rules and people probably disregard any advice the game gives them.

[Anonymouswizard]

5e works at level 20, 3.5 breaks before level 20.

-DF

Honestly, I'm not sure 5e works at level 1.

Although I feel 5e's problem is closer to 'numbers are designed for situations where you roll multiple times, most skill uses are once and done'.

[Telok]

Found my stuff on rolls. Still rough draft, missing full suite of examples.

1. Who rolls
One person
Several people in order
Several people at once
Everyone in order
Everyone at once

2. Target numbers (may combine)
Over/under a static number
Opposed (Amy vs Bob)
Sum total (add all rolls until the number is something)
Sum total being reduced by something

3. Outcome (may combine)
Binary success/failure
Continue rolling until something
Best X out of Y rolls (a.k.a. X success before Y+1-X failures)
By amount (degrees of success/failure)

Examples:
A single basketball half court free throw for $100,000. One person, one roll, binary outcome.
Tug-of-war game. All at once, sum total reduced by opposed team roll, continues until amount of success is reached by one side.
Chess match, play by play. Each person in order by side, could be opposed by the previous persons roll, could be roll until 6 success before 3 losses with possible success/loss removed for rolling a sufficient degree of success.
Open giant door. Several people at once, each person has to get over a minimum number, sum rolls until over a large number, reduce by opposed check of person on other side of door trying to keep it closed.

Frankly it all comes down to whether you're rolling to find success/failure, how long it takes to succeed, or how much you succeed/fail by.

[Quertus]

In 5e, the GM can involve both the +11 specialist and the +0 unskilled character in a challenge.

In 3.5 the GM picks either the +22 specialist or the +0 unskilled character: Only one of their rolls will matter.

There is a lot more variety to 3.5 than 5.e, but I figure the "best" of the two systems has got to be the one that is playable by the most players.

-DF

My main disagreement would be that rolling dice != playing the game.
Only one of them will be rolling, yes. And in some scenes, nobody rolls anything. Are those scenes "not playing the game"? Is planning a heist in Shadowrun not part of the game? Is a climactic negotiation to make peace between the PCs' city and the Drow city underneath it, so they can together resist the invading Mind Flayer empire, "barely playing the game" because there may be only a couple rolls involved? Is a caster who only uses spells without attack rolls not playing until they need to roll something?

The other angle is challenge, and while I'd agree that the given skill check is only a challenge for one of them, I don't see that as a problem. Lots of things are only conditionally challenging:
* Crossing a big chasm is difficult ... unless you can fly.
* Getting inside the court is difficult ... unless you're a noble.
* Crossing Fever Swamp is difficult ... unless you're immune to disease and poison.
* Getting passage to Middle-of-Nowhere Island is difficult ... unless you're rich enough to charter your own ship.
* Deciphering this ancient script is difficult ... unless you happen to speak the language.

Challenge is best thought of in aggregate rather than a set of fixed points which must each be challenging. You have a series of obstacles. Depending on the party, some of those will be challenges and others won't. The ones that aren't should generally be resolved quickly, which means that in terms of real-time spent playing, challenging situations will account for a larger portion than their absolute frequency would suggest. If you miscalculated and an entire adventure is resolved with very few difficulties? Not a problem IME - no player has complained about a single adventure being too easy, and often it's a welcome chance to show off. If all the campaign is too easy it could get dull, yes, but it's not hard to make the next part more difficult.

I was going to write a spoof of the "have to be challenged to be playing the game" stance, but plenty of posters have already pointed out the problems with this mindset, so I'll refrain. Instead, I'll just ask, what does the world-building look like when the clueless kid beating on your head with a rock not infrequently does better at brain surgery than the trained medical professional? It seems to me that, due to the world-building overhead, "everyone can participate has a chance" is decidedly not the playable/"best"/easy solution.

For aggregate difficulty… I mean, the same group made 3 parties. One was "average", a second struggled with most things, and the third was a group of BDHs who waded through almost everything like they were humans (and struggled to convince the citizens that we were… if not "the good guys", then at least "the lesser of two evils". So I think that there's a huge range of amount and frequency (and source) of challenge that can still make for a fun game. Other than that clarification, @icefractal, I agree with your excellent post. Kudos on pointing out that the "not a challenge" portions generally shouldn't consume significant table time!

[Cluedrew]

On D&D's Quality: D&D is kind of like Call of Duty. Call of Duty is a good computer game, no question about it. Is Call of Duty the best computer game? Does liking computer games mean you have like Call of Duty? No to both.

Instead, I'll just ask, what does the world-building look like when the clueless kid beating on your head with a rock not infrequently does better at brain surgery than the trained medical professional?

You know I don't care how good at brain surgery you are, hitting someone on the head with a rock is not going to fix brain damage. So I expect those two to do about as well as each other. Which leads into my question:

What d20 system (or any really) makes both of those viable character concepts and then puts no more support behind it than a skill modifier? Can a name a system that has "Medicine: Brain Surgery" as a skill and lets you attempt that with no modifier or other buys? How about equipment? Does giving the trained medical professional tools, an operating room and a support team do anything?

Which I think gets into the real problem here: I have my problems with the 1d20+stat vs. a target number resolution system. But even locking all that in (and not adding things like graded success, multi-part checks or rerolls) that is already ignoring so much of what can make a skill system good or bad. Like where do the modifiers come from? What rules determine when you make a role? Plus you have things like how you frame the results which aren't mechanical but do effect the feel of a system.

In other words: I think d20 variance is too high but good supporting rules structure can make or break the system overall.

[Telok]

In other words: I think d20 variance is too high but good supporting rules structure can make or break the system overall.

I think d20 variance is great for playing Paranoia. Honestly you could write almost any randomization system (heck, I did one that used beer, pretzels, and a bowl) a get it to work for almost anything with great supporting rules. Likewise you could take the best dicing method and turn it into "roll for shoes" with bad rules.

We have 40+ years of using d20s in the hobby, it can work just fine. Write something like the early d&d 4e skill challenges where %success went closer to 100% the higher "difficulty" went and it isn't the dice that are the issue.

[HidesHisEyes]

Sure.

But my point was that if you bother with as may rules as 5E has and all its rule based restrictions, you would expect the rules to give you proper answers instead of heavily relying on freeform.

There are many games with way better skill systems where i don't have to explain away how the performance of the fighter is not actually better than the bard ones even if the rolls say so. It is a weakness of 5E. And it remains a weakness even if i can paint it over with freeform and GM fiat.

Well that’s the thing: when I run 5E the way I described I’m not “painting over” anything. The process of setting the stakes and determining contextually what success and failure mean on this specific roll aren’t some extra thing you need to do. It’s just how the game works. You call for a check when something’s at stake, taking the whole state of the fiction into account to decide what’s at stake. That’s just playing the game. And the rules do give you proper answers: the ability check answers “does this go well or badly?” which is all it’s meant to answer, everything else comes from context.

To your first paragraph: I’ll point out that, outside of combat, the rules of 5E are pretty damn simple. I mean, I’m only really talking about ability checks here. As I said, none of this applies to combat because in combat there are all these other rules and structures that determine the stakes for you. But you are right that once you start playing in this way you feel less need for all those complex rules. 5E isn’t my favourite game and I wasn’t defending it per se - just explaining why I don’t think the “d20 variance” is actually a problem.

[Hytheter]

Honestly you could write almost any randomization system (heck, I did one that used beer, pretzels, and a bowl) a get it to work for almost anything with great supporting rules.

I believe that's what they call designing a TTRPG. :P

You're right, though. The RNG is less important than the mechanics it plugs into. You can get a long way just by adjusting modifiers up and down, let alone all the rules and interactions that go on top.

[HidesHisEyes]

I believe that's what they call designing a TTRPG. :P

You're right, though. The RNG is less important than the mechanics it plugs into. You can get a long way just by adjusting modifiers up and down, let alone all the rules and interactions that go on top.

Yeah definitely. I think structures are more important than specific mechanics, in fact.

[kyoryu]

I believe that's what they call designing a TTRPG. :P

You're right, though. The RNG is less important than the mechanics it plugs into. You can get a long way just by adjusting modifiers up and down, let alone all the rules and interactions that go on top.

See my post way earlier in the thread about how to reduce the worst bits of this :) And most of that isn't even with adjusting modifiers, just in changing how skill checks are framed.

[HidesHisEyes]

See my post way earlier in the thread about how to reduce the worst bits of this :) And most of that isn't even with adjusting modifiers, just in changing how skill checks are framed.

Just went back and read it, and yeah, exactly that.

A good simple concrete example is to track time in some way. Use a timeline, or countdown clocks, or random encounter checks so you can fall back on “it takes up time” as a default meaningful consequence.

[martixy]

That said, I think it’s actually impossible to have an RPG that doesn’t involve some amount of GM (and player) fiat, free form narration, judgement calls and “whims”. The medium relies on a shared understanding of a fictional scenario, no matter how rigorous the rules. In a game like 3.5 the rules for climbing a cliff might exactingly account for all the different factors and potential outcomes. But to engage those rules in the first place the group has to agree, in a qualitative and free form way, that someone in the imaginary world is attempting to climb a cliff. The 5E style I described is doing the exact same thing, just the free form part extends a bit further and the role of the rules is smaller.

Allow me to introduce you to something called a "Single player computer RPG". No GM fiat, no free-form anything, no judgements, no whims. At best one player's headcanon.

What d20 system (or any really) makes both of those viable character concepts and then puts no more support behind it than a skill modifier? Can a name a system that has "Medicine: Brain Surgery" as a skill and lets you attempt that with no modifier or other buys? How about equipment? Does giving the trained medical professional tools, an operating room and a support team do anything?

Put the schmuck in scrubs in the operating room with all the support staff and equipment (and for the sake of argument accept the absurd situation he's allowed to perform the operation) and tell me how high the chances of success are. It seems to me you're moving the goalposts here.

The point is, calling someone a specialist, when someone lucky can achieve the same success is a weakness of the system.

Lets approach this from the other end. Answer the following question:

At what point do we consider someone a specialist? What part of the performance of a specialist should be luck? 50%? 10%? 1%?

[HidesHisEyes]

Allow me to introduce you to something called a "Single player computer RPG". No GM fiat, no free-form anything, no judgements, no whims. At best one player's headcanon.

Yeah when I said RPG I meant TTRPG, not video games. I was talking about medium specificity and that’s a different medium of course.

[martixy]

Yeah when I said RPG I meant TTRPG, not video games. I was talking about medium specificity and that’s a different medium of course.

Why exclude video games though?

[kyoryu]

Why exclude video games though?

Because this is a TTRPG-focused forum, in a TTRPG-focused section of the forum, so that's the understood context unless specified otherwise?

[martixy]

Because this is a TTRPG-focused forum, in a TTRPG-focused section of the forum, so that's the understood context unless specified otherwise?

This is not a good reason. Including them illustrates that you can have RPGs entirely devoid of GM-fiat.

I agree with HidesHisEyes on the stakes. It's a neat solution to a problem. The point I'm trying to make is not that the solution is bad. It's that the problem exists.

[MoiMagnus]

Allow me to introduce you to something called a "Single player computer RPG". No GM fiat, no free-form anything, no judgements, no whims. At best one player's headcanon.

That's also the kind of games that says "no you can't do that" every other idea you have that seems reasonable (like jumping/climbing over some obstacles but not others, or not being able to say relevant things to NPCs that could really use those info).

Especially if you're doing a single playthrough the the game, there is a fair share of fiat, judgements and whims, like if you had a GM. Except here it is the developers instead of the GM. Sure, you will eventually get a feeling of how the game will react (assuming the devs keep it consistent over time), but that's the same if you play with the same GM for a long time.

If the dev decided that something would happen at point X, it will happen, no matter how much sense it makes, how consistent it is with what was before, etc. IRL, you can point some obvious mistakes to your GM and they can correct them. You can give feedbacks. In video games, you are powerless against the whims of the devs.

The main difference is that the devs and beta-testers of a game have put thousands of hours of prep before you start playing to map all the different path they will allow you to take and try to maintain consistent rulings for each path. While a GM will be much more limited in prep time, and will be expected to not say "no you can't" each time the players try an action that was not planned at this moment of the scenario and rather rely on improvisation.

[noob]

Why exclude video games though?

In videogame rpgs it is common to put an emphasis over some aspect at the cost of most others such as an emphasis about a loop of killing and using the equipment of those you killed to kill more.
This in turn created a culture where people say that they are not "real roleplaying games"

[Lord Raziere]

In videogame rpgs it is common to put an emphasis over some aspect at the cost of most others such as an emphasis about a loop of killing and using the equipment of those you killed to kill more.
This in turn created a culture where people say that they are not "real roleplaying games"

This is funny to me, since DnD started that loop before any of those games existed, and I'm pretty sure you can trace back all the videogames that have the kill/loot cycle to DnD murderhobo games as their origin in some manner. so people saying they're not "real roleplaying games" while in turn playing DnD is the original pot calling all the kettles made in its image black.

[noob]

This is funny to me, since DnD started that loop before any of those games existed, and I'm pretty sure you can trace back all the videogames that have the kill/loot cycle to DnD murderhobo games as their origin in some manner. so people saying they're not "real roleplaying games" while in turn playing DnD is the original pot calling all the kettles made in its image black.

I do not think that this loop is a bad thing in itself: it makes pretty fun games.
What I do not like is the tedious farming aspect some videogame rpgs have: I do not want to do 5000 times the same dungeon: I am no longer playing a fun role that makes any sense I am now a worker working at the most boring killing factory.
One videogame rpg I would have liked to try was based on that loop but did not have farming: when you killed something it was dead forever and you could not just restart and see the thing alive again and kill it again.(I believe the name was something like van hellsing I forgot the real name)

[HidesHisEyes]

Why exclude video games though?

Well, because it’s a different medium. It just happens to have the same name. For me, it’s part of the definition of a tabletop RPG that it relies on a shared understanding of a fictional space. In a video game the developers create that space and as player you have to accept it. That’s a very different process and not related to the point I was making.

This is not a good reason. Including them illustrates that you can have RPGs entirely devoid of GM-fiat.

I agree with HidesHisEyes on the stakes. It's a neat solution to a problem. The point I'm trying to make is not that the solution is bad. It's that the problem exists.

How do video games fit in to this? I’m confused.

[martixy]

Well, because it’s a different medium. It just happens to have the same name. For me, it’s part of the definition of a tabletop RPG that it relies on a shared understanding of a fictional space. In a video game the developers create that space and as player you have to accept it. That’s a very different process and not related to the point I was making.

What is your point exactly?

How do video games fit in to this? I’m confused.

Oh, those 2 statements were mostly unrelated. Sorry for the confusion.

Although not entirely... in video game design there is a term "ludonarrative dissonance" - when the gameplay conflicts with the story being told to the player. 5e suffers from this effect. Being fixed by GM fiat does not erase its existence.

TBH I wish people would answer the question I posed in that post instead of focusing on my off-hand remark.

[Cluedrew]

Likewise you could take the best dicing method and turn it into "roll for shoes" with bad rules.

Why you beating on Roll for Shoes? It may not have the deepest mechanics (it fits on an index card) but I played it, I had fun and the mechanics contributed to that fun far more than they detracted from it. That makes it a good system in my book.

It seems to me you're moving the goalposts here. [...] At what point do we consider someone a specialist? What part of the performance of a specialist should be luck? 50%? 10%? 1%?

I was very purposeful ignoring the original goal posts and aiming for another set and did not intend to trick anyone. Because that was the point, I think people were asking the wrong question. "Is it better for a specialist to have a +11 or +22 base modifier" is not a useful question to ask in isolation (even if we know its 1d20 plus stat).

In fact, how am I even supposed to answer your other questions? I'm not sure how to define specialist beyond "is specialised" but even taking the extreme of someone really good at one thing only, the second question is still hard. Is this in a single action or over a scene? Are they doing something routine or something they were called in for because they were the only one who even has a chance. Context is important. Also I'm not sure I could put exact numbers to it either. I once had homebrew system, crunched the numbers to get the success rates where I wanted and one play test in I realized that those numbers were obviously not what I wanted.

So, sorry I can't answer the question, but I honestly don't see how. I can tell you I want a system where characters can contribute to scenes, people can operate outside their strengths but it is also really obvious when a character is doing something they are good at. And that's not even the full list, just some of the high priority ones related to a the topic. Boiling all of that done to "+22 or +11" doesn't mean anything to me.

[Telok]

Why you beating on Roll for Shoes? It may not have the deepest mechanics (it fits on an index card) but I played it, I had fun and the mechanics contributed to that fun far more than they detracted from it. That makes it a good system in my book.

Oh I'm not bashing Roll for Shoes, its got a good system for it. You just probably shouldn't use the system for something less light-hearted and improv.

Its a bit of an extreme comparison but D&D 4e and 5e non-combat felt a lot like Roll for Shoes without the fun, under a couple different DMs. In combat you're one-hit-killing ogres and dropping fireballs that roast a dozen goblins. Leave combat and its all "walk up a moderate gravel slope: dc 11" and "offer a bribe to the guy who just asked for it: dc 13" or "spot unhidden hanging net trap in middle of well lit room: dc 17".

[HidesHisEyes]

What is your point exactly?



Oh, those 2 statements were mostly unrelated. Sorry for the confusion.

Although not entirely... in video game design there is a term "ludonarrative dissonance" - when the gameplay conflicts with the story being told to the player. 5e suffers from this effect. Being fixed by GM fiat does not erase its existence.

Ah, I see where you’re coming from. My point was that fiat isn’t a fix for something, it is an inherent part of TABLETOP roleplaying games. Not just fiat but conversation, collaborative imagination and subjective agreements. This is definitional imo: if it doesn’t have these aspects then it’s not a TTRPG. And I don’t consider it a problem whatsoever.

I was saying that games like 3.5 or Pathfinder try to limit the subjective aspect of the medium by making the rules more deterministic, in pursuit of consistency, predictability and so on. And I was arguing that 5E (outside of combat) takes a different approach where the mechanics are much less deterministic and subjective agreements about the fiction play a bigger role.

To be clear, both approaches are valid. My point was that if you take the more subjective approach (as I think 5E wants you to) then the problem of “the numbers are too small, they get drowned out by the d20 variance” isn’t a problem. It’s not that you solve the problem, it simply isn’t a problem in the first place.

[HidesHisEyes]

Oh I'm not bashing Roll for Shoes, its got a good system for it. You just probably shouldn't use the system for something less light-hearted and improv.

Its a bit of an extreme comparison but D&D 4e and 5e non-combat felt a lot like Roll for Shoes without the fun, under a couple different DMs. In combat you're one-hit-killing ogres and dropping fireballs that roast a dozen goblins. Leave combat and its all "walk up a moderate gravel slope: dc 11" and "offer a bribe to the guy who just asked for it: dc 13" or "spot unhidden hanging net trap in middle of well lit room: dc 17".

That’s what Kyoryu and I have been getting at: those examples are bad GMing. Unless there’s some other pressure or danger in the fictional context, none of those things are meaningful challenges for a character and they don’t call for a roll. I agree what you’re describing sucks, but it’s not the game’s fault.

[Telok]

I agree what you’re describing sucks, but it’s not the game’s fault.

I would phrase it as "it's not the fault of the dice mechanics of the game when those mechanics are misused".

I consider the game to be the whole published bundle that people hold in their hands, read, and run from the books. Therefore I consider places where the game hands you a problem with no solution or bad advice to be the game's fault.

Now, how much of an issue these faults are is variable and subjective. Variable based on the skills of the groups, the style of the game, etc., etc. And subjective based on some people not noticing, not caring, having previous habits that avoid an issue, etc., etc.

I, personally, have not had fun in D&D non-combat encounters since the end of v3.5 up to the plague shutdown, if the d20 was in play. Because the d20 high-or-low result often made anything I said, anything I did, anything on the chatacter sheet, irrelevant. Plus of course the too frequent rolling and the 'normal' dc 15. To me that indicates problems in, if not in the mechanics, then the presentation and advice coming from the game materials to those DMs. Of course, I also despised StarFinder space combat. I like sci-fi games and have a selection of space combat models across them, each with its own strengths and weaknesses. The StarFinder space combat boardgame had several immedately apparent design flaws and that grated on my nerves. The DM and I sat down and hashed out agreements on which things we would avoid to not totally fubar the whole thing. Definite issues.

[Vahnavoi]

Computer games don't actually lack fiat, arbitration etc.. They just take place at a different point of a game's life cycle. Whoever makes a single player computer game has to make all the same kind of decisions as a tabletop game master while they are designing and programming the game. When and where ever they fail at making the game complete enough, the game program will crash, halt, terminate or be caught in a loop up untill the program is returned to them and they make a new ruling and program it into a game patch. There isn't a fundamental difference between that and a game master on the tabletop making rulings, the process just has greater latency.

[HidesHisEyes]

I consider the game to be the whole published bundle that people hold in their hands, read, and run from the books. Therefore I consider places where the game hands you a problem with no solution or bad advice to be the game's fault.

Yeah that’s fair. The fact that the game’s text doesn’t explain these concepts is a problem with the game. And the majority of my experiences with 5E have been plagued by such problems, including as a GM because it took me several years of playing and running the game to figure it out. To be honest I’m not sure how much of a conscious choice the design philosophy of 5E even was, on the designers’ part.

Computer games don't actually lack fiat, arbitration etc.. They just take place at a different point of a game's life cycle. Whoever makes a single player computer game has to make all the same kind of decisions as a tabletop game master while they are designing and programming the game. When and where ever they fail at making the game complete enough, the game program will crash, halt, terminate or be caught in a loop up untill the program is returned to them and they make a new ruling and program it into a game patch. There isn't a fundamental difference between that and a game master on the tabletop making rulings, the process just has greater latency.

Well I can see what you’re saying on a purely technical level, that it’s only a quantitative difference in latency. But I still think that on the level of the actual experience of play, it is fundamentally different. In a video game, the experience of play is that the game world is there (whatever version of it) and I’m interacting with it entirely on its terms. In a TTRPG the game world only exists through my interaction with the other players. We have to agree that something is the case in the fictional scenario before we can engage the mechanics at all.

[MoiMagnus]

Well I can see what you’re saying on a purely technical level, that it’s only a quantitative difference in latency. But I still think that on the level of the actual experience of play, it is fundamentally different. In a video game, the experience of play is that the game world is there (whatever version of it) and I’m interacting with it entirely on its terms. In a TTRPG the game world only exists through my interaction with the other players. We have to agree that something is the case in the fictional scenario before we can engage the mechanics at all.

If you're playing a single player TTRPG (which some peoples do), the game world exists entirely within the mind of your GM. And with a GM dictatorial enough, you're interacting with it entirely on his term too.