|
Leperflesh posted:Ah, I see. Yeah, still a binary result from a die, but basically a coinflip result using a d20 to get it. Eh, d20s actually have some more fundamental problems beyond just the binary pass/fail thing in the games that use it. I've toyed around with a bunch of homebrew dice systems before, and nearly every method always runs into either the "can't win" or "can't lose" problem where there's no point to even roll. A d20 is especially notable because it can actually have both of those problems, when the target number is 15 and you've either got a +14 (so you succeed on a nat 1) or a -6 (so you fail on a nat 20). The stop-gap to this in most of these systems is to just make natural 20s and natural 1s always succeed or fail, but that's just capping the odds of success or failure at a minimum of 5% no matter how much you or the target number have blown past each other, and it doesn't even get used for skill checks in those games. Even dice pool systems can run into the same problem; if each die can roll a maximum of 1 success, then even though you can always fail by rolling all blanks you won't be able to roll 6 successes out of 5 dice. World of Darkness treats difficulty as a penalty to your roll instead, but then they have to put in another stop-gap minimum "chance die roll" when your total would go below 1. The only seemingly-perfect solution I've seen to date has been the dice pools in Edge of the Empire/Genesys, where you roll positive dice for your stats and negative dice for the difficulty and they cancel each other out. That makes it so you can always succeed by rolling blanks on all the negative dice and always fail by rolling blanks on all the positive dice, but with a nice curvy probability distribution as the difference between your stats and the difficulty increases.
|
#
?
Mar 30, 2021 22:50
|
|
|
|
| # ? Apr 28, 2024 18:28 |
|
|
In Modiphius D20, the maximum Difficulty of a test is 5, so any character, even unskilled, can buy three dice and have a chance of success. It also has a "succeed at cost" option, requiring GM approval, in which the player and GM agree the test will automatically succeed, and then you roll dice to find out how many fewer ranks of success you actually generate: the difference piles on Complications/GM dice. The effect of the system really does mean that a player never has to risk a catastrophic loss on a single roll: but, there are consequences (used up resources, additional tokens for the GM which translate into increased threat later, additional complications which translate into immediate setbacks now, etc.) so players are disincentivised to go all-in on unimportant rolls. Not to go too hard into the system, but: there are other parts of the system that feed into this. Character attrributes range from a minimum of 6 to a maximum of 12 (or 14 with an optional rule): totally untrained in a skill, the base attribute is the TN, which you roll under. So, a character's absolute worst chances on any attempt at anything is 2 20-siders with 30% chance of generating a success on each die; and typical Difficulties are around D1 or D2, so a single success can get you past a lot of fairly normal run-of-the-mill checks. A player with that character who wants to go in harder can buy 3 more dice and have five 30% chances to generate successes. Skill ranks have two numbers, bought together during chargen and bought separately for XP afterward: Proficiency adds to your target number, while Focus is treated separately. So a character with an average attribute of (say) 11, and two ranks of Proficiency in some skill, is rolling against TNs of 13; if they also have 2 ranks of Focus, than in addition to generating a success on any roll of 13 or lower, any die result of 1 or 2 generates an another success. Having any ranks of Focus in a skill thus makes it possible to get more successes with fewer dice. Lastly, characters have Fortune points (generally 2 or 3 at chargen), and a player can spend a Fortune point before rolling to add a d20 and set it to an automatic 1. E.g., if they have any Focus in the skill at all, that die is an automatic 2 successes. There's more: equipment, cooperative attempts, environmental factors. But the system boils down to: tons of player agency, and every PC has a nonzero chance to succeed at anything they can roll dice for, with the chances of success always being on a probability curve rather than a flat array. e. On the "can't lose" side of the equation, every natural 20 generates a Complication, regardless of whether it's a success or not, so there's always a possibility of setbacks and problems coming from every die roll, no matter how skilled and competent the character. Plus giving the GM more tokens for the pool means more challenge coming soon. Leperflesh fucked around with this message at 23:09 on Mar 30, 2021 |
|
#
?
Mar 30, 2021 23:04
|
|
|
Vanguard Warden posted:Eh, d20s actually have some more fundamental problems beyond just the binary pass/fail thing in the games that use it. I've toyed around with a bunch of homebrew dice systems before, and nearly every method always runs into either the "can't win" or "can't lose" problem where there's no point to even roll. A d20 is especially notable because it can actually have both of those problems, when the target number is 15 and you've either got a +14 (so you succeed on a nat 1) or a -6 (so you fail on a nat 20). The stop-gap to this in most of these systems is to just make natural 20s and natural 1s always succeed or fail, but that's just capping the odds of success or failure at a minimum of 5% no matter how much you or the target number have blown past each other, and it doesn't even get used for skill checks in those games. *sometimes your target difficulty can be increased or reduced by traits and such but the difficulty for straight roll tasks caps out well before the absolute maximum roll of 10 e: gently caress you leperflesh. gently caress. You. e2: not all 2d20 systems do it identically. In STA every roll of a 1 counts as 2 successes. You have 6 focuses, but these have no numerical value attached, they just are. When you're rolling your TN is based on adding together an appropriate Attribute (7 to 12) and a Discipline (1 to 5), and if you have at least one appropriate focus your doubles range increases to equal the discipline. Your determination (fortune but different) can only be spent when you have an appropriate Value (like "Go big or go home" or "We come in peace (shoot to kill)"), but since a 1 is always two successes playing to your values with determination is always a guaranteed decent success. Splicer fucked around with this message at 23:45 on Mar 30, 2021 |
|
#
?
Mar 30, 2021 23:14
|
|
|
haha owned
|
|
#
?
Mar 30, 2021 23:20
|
|
|
Leperflesh posted:An additional issue with the D&D/D20 style of play is the hard success/fail nature of those rolls, as Arivia pointed out: you miss your target number by 1 or by 10, doesn't matter, you failed. You made it by 1 or by 10, doesn't matter, you succeeded, with the exception being crits. That’s where I enjoy the PF2e method of critical successes and critical failures, where they happen if you beat or fail the roll by 10, respectively. It helps mitigate the “can’t succeed/can’t fail” situations too, because hey, while you might not fail you might still jot crit succeed and vice versa. As for systems where you roll multiple dice and see how many successes you get, I’ll have to check them out. The closest to one of those I’ve played is BitD, which I thought was okay but the overall system wasn’t what I was looking for
|
|
#
?
Mar 31, 2021 00:16
|
|
|
It feels like the only really reliable way to combat the inherent swingy-ness of d20 systems is through playing with the actual target number the players need to hit, with the main problem I've seen in implementation in games like D&D being that the game usually takes this in the opposite direction and only plays around with increasing the difficulty of rolls or not giving a sensible baseline for what an average target number should be. It also leads to a case where you're kind of playing around with the worst aspects of both percentage based systems (Where the die roll is completely unmodified while the target number is altered) and systems that use a bell curve (Where the target numbers are usually fixed and the rolls are modified)
|
|
#
?
Mar 31, 2021 00:26
|
|
|
Leperflesh posted:Still: I'd say that a poster asking about a "leaf" type rules question isn't necessarily assisted by a series of posts assaulting the trunk... OK, maybe some users need to hear "this game has a severe flaw further up that you should consider house-ruling" but some posters really just would like to know when they should apply a penalty, either by hearing other people's interpretations of the rules as they're written, or finding out that there's some evidence of what was actually intended, perhaps elsewhere in the rules ("the acrobatics skill section applies x penalty under y conditions, that's perhaps what was intended for athletics as well"). I think "assaulting" isn't what I'm trying to talk about here, because I agree with you that "game bad, play something else" is a useless thing to say. But it is useful to point out when there are things elemental to the forest that make someone's leaf question unanswerable from within the forest, because then the discussion can focus on something other than trying to find an answer that doesn't exist in the text. Eg, (in my opinion), it's impossible to get a satisfying answer on the leaf issue "what was the wording of this particular D&D spell intended to mean", because there's a tree-level issue of "D&D contains hundreds of differently-worded spells which do not follow a consistent framework even when they have similar effects to other spells or non-spell rules", and a forest-level cause of "it was done that way on purpose because when the game deviated from that format, an apparently large number of players complained that the game was no longer "really" D&D". In other words, it can be impossible to get a meaningful RAI from the text itself on a particular issue because of an issue further up the tree. When that happens, the discussion must therefore center around different ways people personally approached that issue rather than on attempting to discover some "real interpretation". Elector_Nerdlingen fucked around with this message at 01:38 on Mar 31, 2021 |
|
#
?
Mar 31, 2021 01:35
|
|
|
KingKalamari posted:It feels like the only really reliable way to combat the inherent swingy-ness of d20 systems is through playing with the actual target number the players need to hit, with the main problem I've seen in implementation in games like D&D being that the game usually takes this in the opposite direction and only plays around with increasing the difficulty of rolls or not giving a sensible baseline for what an average target number should be.  Changing the target number and changing the difficulty in a d20 system have exactly the same result. Whether you change the modifier or the target number doesn't affect anything, all else being equal. Nothing you're talking about has anything to do with how "swingy" a dice system feels. That comes down to consistency in succeeding on a task/class of tasks attempted regularly, which has nothing to do with whether you apply a modifier to the roll or to the target number. If you want to fix D&D feeling swingy, fix the part where the game has, since 1974, been built on the assumption that everyone should spend their first three levels having a 60% chance of failing at hitting a goblin, or at least, that a 60% chance of failing is a reasonable thing to feature in a standard combat situation.
|
|
#
?
Mar 31, 2021 04:09
|
|
|
LatwPIAT posted:
Sorry, I feel like I have something resembling a point but am not communicating it particularly well. What you mention is basically the meat of what I was trying to convey: The D20 system's approach to modifying difficulty is really weird because having both an adjustable target number that can be manipulated by the DM and modifications to the dice roll itself essentially do the same thing and having them both in the system together makes determining/recognizing the actual chance a roll is going to succeed more complicated than it needs to be. The swinginess of the die doesn't feel as noticeable to me in the D100-based systems I've played, but those still don't incorporate a bell curve in the rolls. The difference is that the D100 systems all work on a percentage roll-under system so the actual chance of success is way more clearly communicated to both the player and the GM before the roll is made. This not only makes it more clear to the player how likely they actually are to succeed, it mroe clearly communicates to the GM how difficult the action ends up being mechanically. D&D, by contrast, obfuscates probabilities behind two layers of manipulation, which only serves to make it less obvious to the GM how low the chance of success for a roll actually is. Again, I feel like I have a point somewhere in here that I'm struggling to communicate properly (Or at all).
|
|
#
?
Mar 31, 2021 04:42
|
|
|
KingKalamari posted:Again, I feel like I have a point somewhere in here that I'm struggling to communicate properly (Or at all). It sounds like what you're trying to say is that the way most d20 systems adjust both the die roll and the target number intentionally obfuscates the probability of success, that you think that this is bad design, and that a much clearer way to do it would be to adjust only the target number? Elector_Nerdlingen fucked around with this message at 05:00 on Mar 31, 2021 |
|
#
?
Mar 31, 2021 04:57
|
|
|
Elector_Nerdlingen posted:It sounds like what you're trying to say is that the way most d20 systems adjust both the die roll and the target number intentionally obfuscates the probability of success, that you think that this is bad design, and that a much clearer way to do it would be to adjust only the target number? Yeah, that sounds about right. Thank you!
|
|
#
?
Mar 31, 2021 05:02
|
|
|
Raising both the TN and the roll bonus may give players the sense that both the challenge has become harder, and, their character has become better at it. It does have the effect of making those challenges effectively impossible for characters who have not trained & leveled up a bunch in that die roll. But, the practical result is that the GM must make carefully-adjusted fine-tuned exact TNs just to challenge specific characters, and as the whole party gets to mid-high levels, each challenge applicable to just one skill has just one character who can do it with some chance of success, and if the GM forgets a bonus or someone casts a spell, the challenge vanishes and the attempt is once again trivialized. So like: your low-level rogue can climb walls OK, your high level rogue can probably still have a chance of failure at climbing a greased glass surface 10 stories tall, while the rest of the party watches, unless they stack more bonuses than expected. But this puts a lot of work on the GM, and the results may be dissatisfying. For sure I prefer systems these days where a character well advanced in their adventuring career is not 10 or 20 or even 30 times more embiggened than when they started (but only within their specialties). Increasingly I favor games where the rewards for the players come from achieving in-game goals (solve problems, grow as people, or even become wealthy and politically powerful) rather than chase ever-bigger numbers on their sheets. But, I can't deny the long-lasting and persistent appeal of the level-up, and expectations from players often stem from that experience - even novice roleplayers get it from countless video games. This speaks to Nerdlingen's point maybe; you can notice how the binary D20 pass/fail skill roll mechanism works (a tree) but peel back to the forest-level and you see another aspect of this mechanic: systems where you gain levels, a forest-level foundational decision that affects most everything else about the game.
|
|
#
?
Mar 31, 2021 05:24
|
|
|
Is the idea behind having beef with d20s that their implementation is poor in most games where a die roll is needed relative to other dice systems out there that use the same or different dice? Part of the philosophical approach to older versions of D&D was that combat was pretty lopsided. If you had a magic user, you had a loaded gun since they could hit something with 100% accuracy via Magic Missile or circumvent an even-odds encounter by casting Sleep. Mortality rate was very high and combats were more desperate at low level as a result, which generally meant instead of going toe to toe with monsters, you had to weight the fight grossly in your favor via tricks - ambushes and the like. The alternative is likely being overwhelmed, which meant getting into a fight in the first place meant something went wrong (this also was replicated outside of D&D in games like Shadowrun, where older editions played out like that scene from Inglorious Basterds where everybody just wasted each other in the bar after a protracted staredown). This is one of those discussions where the stuff about more clearly qualifying the beef might be useful. Since a d20 is just a way to determine a linear binary success/failure thing, then saying "d20 bad" could potentially be referring to this specific point, "binary success/failure on a d20 bad". The corollary "implementations of d20 and target numbers plus modifiers bad" rolls up to the binary success/failure thing. Other dice systems have nonbinary outcomes like margin of success (Powered by the Apocalypse), picking and choosing results (7th Sea, Legend of the 5 Rings), rolling multiple dice instead of a single curve (many, many systems), or using proprietary dice (Genesys), or no dice at all (Everway). "Too swingy" would perhaps (please clarify if I got the read of it wrong) imply that the d20 throw has too much influence on whatever it was that is being done where an element of chance is at play. You can get big gains, abysmal failures, or extended passages where rolling and failing or rolling and succeeding has a dramatic effect on the enjoyment and outcomes of the game. If this is the case, then judgement on when a d20 throw comes into play seems to come into effect, but instructive rules on these things are sparse in most game texts, with the best offering being "don't roll if it's boring". However, if something like combat demands a roll and if everybody constantly misses, it makes it boring... Anyway, digging back into the personal experiences bucket again, I have had the best success at my tables that use d20s by using "margin of success" and "omission or permission of rolls". That is, a roll is entitled based on the specific character, their proficiencies, and what they are reasonably capable of - a Knowledge Cleric in D&D 5e could be expected to know a thing or two about esoteric forbidden lore, but a Fighter who never set foot outside of their training hall may not. Based on how well the roll succeeds or fails based on a target number (which sometimes is optional), define the results accordingly. Part of the "fun" of the above is messing with the signal to noise ratio outside of combat. Some rolls which could be made in secret like Knowledge rolls are instead made by the players who can see the result, but since they know there are margins of success, they may receive information based on their roll, but have no clue if it's accurate or not. This feeds back into a roleplaying loop because if someone got Big Number, surely they know Everything about Topic X, and that might be true in that case. However, when they receive intel about Topic Y and roll very low, they might still know that information -- they just don't have a way to verify its true unless they do some experimentation. Whether or not the player and their character decide to act on that information is actually particularly interesting and engaging at the tables I run. I also don't know where this originally came from, but likely from "a bunch of people in the larger gaming circle I grew up with in the 90s and 00s" is very likely. It was further reinforced in other non-D&D systems like Dungeon World and the like, so as I ended up back in D&D land I just took it and ran with it with success. Anyway, if the idea behind "d20 - bad" is "binary results" plus "frequency of rolling" plus "modifiers which obfuscate the binary results but really it's all just smoke and mirrors", those are individual components that could theoretically be changed, if one so wished.
|
|
#
?
Mar 31, 2021 05:27
|
|
|
Leperflesh posted:This speaks to Nerdlingen's point maybe; you can notice how the binary D20 pass/fail skill roll mechanism works (a tree) but peel back to the forest-level and you see another aspect of this mechanic: systems where you gain levels, a forest-level foundational decision that affects most everything else about the game. I think it does, and it's probably a better example than the one I was trying to use. Elector_Nerdlingen fucked around with this message at 05:42 on Mar 31, 2021 |
|
#
?
Mar 31, 2021 05:37
|
|
|
aldantefax posted:Is the idea behind having beef with d20s that their implementation is poor in most games where a die roll is needed relative to other dice systems out there that use the same or different dice?
|
|
#
?
Mar 31, 2021 11:44
|
|
|
LatwPIAT posted:
This right here outlines one of my greatest design peeves. Never, ever, loving ever build 'player's turn is wasted' into your system as an expectation. Make every action matter. High miss chances just mean combat drags on for hours, and doesn't really open up any interesting possible results. It just means some of your players are going to be getting increasingly frustrated and/or bored doing something that is effectively 'not playing the game' and doesn't advance play.
|
|
#
?
Mar 31, 2021 15:10
|
|
|
Systems need to account for the fact of streakiness in any truly random distribution, as well. If a single failed roll is disappointing, a system that assumes you rarely make five bad rolls in a row, is screwing something up... because you are going to make several bad rolls in a row, often. That's just how random distributions work. (And of course, frequently make several good rolls in a row, too.) How many of us have had entire D&D combats of multiple rounds where all our characters did was miss their attacks repeatedly? That may be "realistic" in some sense, but for some of us, it becomes tiresome and can drag down an entire play session. I've also heard of new players entirely turned off of RPGs because their first couple of game sessions consisted of a sequence of failures and setbacks without a heroic sense of accomplishment or ultimate victory arriving soon enough to entice them to stick to the game. Systems in which players either have an option (perhaps at the cost of resources) to declaratively succeed irrespective of a dice roll; or, receive some benefit or interesting outcome even when they miss a target number on a roll, are a significant improvement in feelgoods at the table. If "balance" is desired, these systems can also either empower the GM to sometimes cause antagonists to succeed by declaration; or, sometimes have some penalty or setback occur even when a PC succeeds. I think these mechanisms can be well or poorly implemented, of course; and the style of gameplay they produce may or may not appeal to everyone. It can be a challenge for a GM to come up with an endless sequence of Complications that remain novel and interesting, off the cuff, through a long adventure or campaign, for example. Players may hit more analysis paralysis if they have an option to adjust probabilities of success on every roll via expenditure of an available but limited resource - and if that resource is shared with the party, play can break down into extended interplayer debates about optimal resource usage. For some, "just roll a D20 and you hit or you miss" is a desirable, straightforward mechanic that requires less decision-making or on-demand creativity.
|
|
#
?
Mar 31, 2021 20:27
|
|
|
KingKalamari posted:Sorry, I feel like I have something resembling a point but am not communicating it particularly well. What you mention is basically the meat of what I was trying to convey: The D20 system's approach to modifying difficulty is really weird because having both an adjustable target number that can be manipulated by the DM and modifications to the dice roll itself essentially do the same thing and having them both in the system together makes determining/recognizing the actual chance a roll is going to succeed more complicated than it needs to be. admittedly, not nearly as difficult as the L5R exploding dice system, where you end up with multi-humped probability curves
|
|
#
?
Mar 31, 2021 21:33
|
|
|
Leperflesh posted:Systems need to account for the fact of streakiness in any truly random distribution, as well. If a single failed roll is disappointing, a system that assumes you rarely make five bad rolls in a row, is screwing something up... because you are going to make several bad rolls in a row, often. That's just how random distributions work. (And of course, frequently make several good rolls in a row, too.) Yeah, to use an anecdote I quite literally fell asleep in my pathfinder game recently as I am playing the fighter, and just had a session where I rolled every save possible horribly, so effectively did not exist for a 2 hour combat while the casters in the party worked out how to handle the opponent. If I were a new player, I would've needed some talking around to come back. I'm not, so we had a good discussion afterward about why the monster was statted like that, and the GM was frustrated as well because we're playing a pre-gen adventure and he did not expect it to be able to just effectively delete anyone with a poor Will save. Tunicate posted:admittedly, not nearly as difficult as the L5R exploding dice system, where you end up with multi-humped probability curves You want a real pain in the rear end try to stat Shadowrun 3e, where there's a stack of modifiers on everything, dice pools are usually 10+ on d6's that can explode, and target numbers can go to 12.
|
|
#
?
Mar 31, 2021 21:39
|
|
|
I'll echo what others have said and go further - the problem of uninteresting failure and wasted turns is not something that's solved by switching from a flat distribution to a normal distribution. The odds might be less swingy, but repeatedly failing on 3D6 or 2D20 or whatever dice pool system you're using does not feel any better than failing on a single D20. You can push the probabilities around to make certain outcomes more likely, but ultimately the way to fix the issue is
|
|
#
?
Apr 1, 2021 00:57
|
|
|
mellonbread posted:Make a failed roll more interesting than just "you fail and nothing happens" - lots of newer games explore this, but even rules like FFG Star Wars that hypothetically have very intricate and varied possible outcomes can still spit out a result of "you miss the blaster shot and nothing happens" if the die results that generate the more interesting outcomes cancel out, or just don't appear Notably, FFG Star Wars is the only system I've played where "nothing happens" was such a rare outcome that it actually somehow became interesting the few times that it did. The successes and failures are spread across different faces of the dice from the advantage and threat, so it's pretty often that you're either succeeding with threat or failing with advantage unless there's a big gulf between the ability and difficulty of the check. I think I've only ever gotten like 2 or 3 rolls with a result of "nothing happens" over half a dozen full campaigns, to the point where it was actually pretty unique and fun to narrate how I failed to jump from the ledge to the ship's open boarding ramp, but somehow in a way that didn't harm or disadvantage me in any way.
|
|
#
?
Apr 1, 2021 01:16
|
|
|
I thought Iron Heroes did an interesting job by adding more dimensionality to the system where d20 throws were common since it was a 3.5 offshoot. To memory, it gave classes resources to spend for combat benefits, but I have no idea if those things supplemented or circumvented d20 rolls. Part of the appeal of d20 systems has been simplicity even if to a lot of folks it has been overdone. You can see when there's a big number, and many systems that revolve around it depend on making the big number. I don't think that's necessarily to everybody's taste, of course, but it has served many tables. To think on the clip from the Bard about Heroquest, paraphrased: "...The dice add a certain drama to the game. Will I leap heroically into battle, or trip over a flagstone? (Immediately after a lovely roll) ...The best thing about Heroquest is ignoring the dice! Clearly intended for the younger crowd more accustomed to Chutes and Ladders..." Anyway I think d20s like the D&D brand have a whole lot of baggage associated with them but are not inherently bad dice because all dice are just things to make random numbers on. You want more even results, you add more dice to curve your results. You want kooky poo poo, add special things that happen on specific dice results (I long entertained adding house rules for special triggers that happened in things that do happen in the most common scenarios like fights like "missing AC by 1" as a "Just Defended" perk and narrow hits etc. but it was all a bit unwieldly to use in normal play). I like the mechanics of 'cut to the quick' for combat and just eschew to-hit rolls entirely, as a tangent, and just have everybody roll damage. Very fundamental shift in the actual game feel but makes everybody feel useful because nobody is useless due to a whiff - but on the same token, if you're in a fight, nobody is walking away uninjured.
|
|
#
?
Apr 1, 2021 01:29
|
|
|
mellonbread posted:Make a failed roll more interesting than just "you fail and nothing happens" - lots of newer games explore this, but even rules like FFG Star Wars that hypothetically have very intricate and varied possible outcomes can still spit out a result of "you miss the blaster shot and nothing happens" if the die results that generate the more interesting outcomes cancel out, or just don't appear Without getting too semantic about what "nothing" is, the effect of a missed roll is that the monster does not take 1d6 damage, which has the knock-on effect of making the monster have one more turn to kill the PCs in. (Moreover, the 1970s D&D model this all builds off had a miss mean another turn during which you roll on the Wandering Monster table, potentially summoning more monsters and changing the nature of the combat. Missing in combat in old D&D was a race against the clock.) While easy to ignore, "not killing my target so they can attempt to harm me" is something that happens. And in the interests of making this not entirely playing semantic games, I've found that the "you fail and nothing happens" approach works decently well in high-lethality games. Missing the monster in Call of Cthulhu is likely, but it can also be terrifying, because you know the monster is going to have a 30% chance of just straight-up eating you on its next turn. Whereas where I see people most often complain about nothing happening on a failed roll (in combat, skill checks to pick the lock or whatever has a lot of very different concerns) is 3.0+ D&D, where combat is a lengthy process. This, I believe, is because in D&D the outcome is, broadly, already known. You know that over the next ten turns you're going to hit on average four times and on average deal 8d6 damage and take 3d6 in return. Missing, even hitting, can be thoroughly unsatisfying because it doesn't really change the outcome. You know that unless something really unlikely happens, you're going to walk out of this with over half your hit points, and if you don't, you'll just have to consume some of your effectively infinite resources. Rolling for each individual miss (and hit) does not lead to any interesting outcomes at all.
|
|
#
?
Apr 1, 2021 01:48
|
|
|
There's also problems of encounter design demonstrated really strongly in D&D. The higher level you get, the more likely that you're fighting a single big monster because nothing else can survive your damage output more than a round. This means both that you're more likely to have that 'and nothing happened' moment and that you're outperforming the opposition 4:1 or better in action economy.
|
|
#
?
Apr 1, 2021 01:56
|
|
|
Liquid Communism posted:Yeah, to use an anecdote I quite literally fell asleep in my pathfinder game recently as I am playing the fighter, and just had a session where I rolled every save possible horribly, so effectively did not exist for a 2 hour combat while the casters in the party worked out how to handle the opponent. Similarly, I was involved in a D&D game where I spent the climactic battle failing saving throws. Literally just failing saving throws. I did not get to take a turn beyond saying "I failed my saving throw". And the response I got afterwards was "that happens sometimes, it's just bad luck" and a refusal to entertain the idea that I didn't get to play D&D that night because I "got to roll dice".
|
|
#
?
Apr 1, 2021 02:15
|
|
|
LatwPIAT posted:(Moreover, the 1970s D&D model this all builds off had a miss mean another turn during which you roll on the Wandering Monster table, potentially summoning more monsters and changing the nature of the combat. Missing in combat in old D&D was a race against the clock.) Games that riff on the old rules will roll for an encounter during or after combat due to the amount of noise it makes, or attach a modifier to the next encounter roll.
|
|
#
?
Apr 1, 2021 03:58
|
|
|
mellonbread posted:Sort of? OD&D says you check for wandering monsters every turn, and "There are ten rounds of combat per turn." The distinctions between rounds and turns were not always super clear in the pre-Basic rules, but I'm pretty sure you didn't check for wandering monsters every combat round. Very possible! OD&D is confusing as gently caress. In any case, the longer you have to fight the monster the higher the chance its friends arrive.
|
|
#
?
Apr 1, 2021 04:39
|
|
|
mellonbread posted:Sort of? OD&D says you check for wandering monsters every turn, and "There are ten rounds of combat per turn." The distinctions between rounds and turns were not always super clear in the pre-Basic rules, but I'm pretty sure you didn't check for wandering monsters every combat round. Turns were actually directly codified in Book III: The Underworld & Wilderness Adventures of the original edition of D&D as comprising 10 minutes, which I believe remained consistent through Basic and AD&D. It's often overlooked today, but the early editions actually had a lot of specific rules and procedures for dungeon-crawling that covered stuff like how long light sources last, how much characters could do in a turn and how much of the dungeon they could explore at once.
|
|
#
?
Apr 1, 2021 04:48
|
|
|
I have wondered about this. The usual OSR is 90% of that comes down to “old” D&D being a heist game, and any “rest” or other resource recovery meant end of heist, with unrecovered treasure probably lost or at least a harder trip for less reward next time. This was then broken by the introduction of story based modules which could not tolerate that kind of ongoing loss. But that seems not to ring true if you look at things in practice. I’m not aware that, say, Keep on the Borderlands could deal the PCs just saying “well, any further expeditions will not break even, where’s the next keep?” and leaving the Tomb of Horrors without killing Acererak, no matter how much loot the PCs got, was such a failure condition that an actual trap tried to trick the PCs into it.
|
|
#
?
Apr 1, 2021 06:12
|
|
|
LatwPIAT posted:Missing, even hitting, can be thoroughly unsatisfying because it doesn't really change the outcome. It’s particularly annoying because the abstract nature of D&D's hp means that the first 10 hit points don’t actually mean much anything in the fiction, either. I just "hit" them and they got "damaged". If I get hit for ten points, I guess that hurts a bit but I’m not going to actually change anything about how I’m playing, it’s just combat tax. It takes fairly good narrative chops to make those rolls interesting, and that ends up taking even more time. Siivola fucked around with this message at 10:02 on Apr 1, 2021 |
|
#
?
Apr 1, 2021 09:08
|
|
|
So, what I'm seeing now is more of an interesting focus on combat. Makes sense given the focus for many games. But what about noncombat? Sometimes a skill roll is crucial and sometimes it isn't. In a typical melee with three to five PCs against some bad guys, yes, the players probably know the outcome is a foregone conclusion - they're going to win - and the distance between the worst-case and best-case scenarios are perhaps a matter of some spent resources and a certain amount of attrition. But success/failure at skill checks can alter the course of the adventure. We want to get past this door, but we failed to: now we must re-route. We wanted to convince the guildmaster to look the other way for a few hours, but he's refused: now we need another plan. We wanted to sail through this storm, but our skill rolls weren't quite good enough: we're off course. What now? Of course not all skill rolls have as much at stake; I failed my Climb roll and fell, and took some attrition to my damage stat, and now we try again. I haggled poorly and had to pay a couple extra coins for this bag of candles. And not all combats are trivial: we defeat the lich and win the day, or the lich defeats us and, I dunno, we all die, or maybe we flee, but the kingdom remains under his thrall? It feels like to me in the typical structure of an RPG adventure, though, that we tend to have more "interesting pivots" tied to skills than we do to combats. Is that a design problem? Perhaps nonpivotal fights are just a trope of the Dungeon Crawl genre that we can enjoy (maybe more tactical-oriented games have combats that play out like puzzles we can enjoy engaging with regardless of the stakes) or not enjoy ("I basic melee attack" for five rounds, took 14 damage, quaffed a potion, and we're done). I think maybe the everpresent gorilla in the room of D&D tends to warp the perceptions of what is or isn't a problem "in RPGs" - I admit that while I've played around a dozen different RPGs by this point in my life, the only games I've ever managed an extended series of adventures in is D&D, and not even that especially frequently. But I don't recall having strings of uninteresting combats with little at stake in any other RPG, even those with similar basic tropes like Warhammer Fantasy RPG. When I play a superheroes game, it seems like every combat has an outcome that is meaningful to the story... maybe we only defeated the Legion of Evil's lowest-tier mooks, but that still means we got access to their security system, or stopped them from destroying the water treatment plant, or captured a criminal who gives up the goods on who he's working for really. Was my Punch Man's role in the fight (Punching a lot) a bit tedious? Maybe, but the experience is enhanced by the value of the victory to the team and the story anyway. Maybe the takeaway here, not that our job in this thread is to "fix D&D," is to find a way to design D&D and other dungeon crawl-type games' adventures so that every combat has something at stake for the party beyond just resource expenditures, a few gold coins, and advancement into the next 10' corridor?
|
|
#
?
Apr 1, 2021 17:23
|
|
|
They can absolutely happen in other RPGs since qualifying the stakes of a given singular roll or scenario in which multiple rolls are made are generally the province of referees at the table combined with any written text they are using. However, stakes generally in D&D get lower because of the decreased treat of dying the closer you get to the modern era, so the compensation is if everybody is knocked out, which generally means an overtuned combat in favor of the DM. On other systems, stakes are generally higher because the lethality portion in combat is still there, but not always. Games like Legends of the Wulin are actually somewhat difficult to take out notable characters, but it is usually just a few solid hits in combat. The reason why the combat happens is more of interest than the combat itself often though as part of genre conventions (wuxia and related genres being much more about interpersonal relationships and the dramatic poles therein). However, that makes for rather uninteresting rules play, so people structure many things around it to make that part of the game fun, since most people actually do find combat fun in the game systems which prominently feature them. D&D has always been a combat engine first and a story vehicle second owing to its roots, but an attempt to return to it placing combat at the top of the crop in 4e had the backlash that made 5e a "return to form". There are some RPG systems out there with fairly cut and dry combat. LANCER's narrative combat, forked off Shadow of the Demon Lord, is a good indicator, and its narrative structure is lightweight but has the basics required to judge when a roll should be made. Hillfolk (and Dramasystem RPG) resolves combat by its rather simple procedural resolution system as well and makes no particular attempt to make it any more complex than it needs to be. If I were to go ahead and say that every roll should be meaningful in a D&D game, then judging the meaning and the call for a roll should have much better instruction to DMs in the games text (as an instructive guide or as an actual rule in the rules engine). However, I suspect that even with the things placed in the DMG for 5e, which is primarily advice on worldbuilding and tweaking things, the actual support structure to help newbie DMs decide when to and when not to make roles is too divisive to take a meaningful stand on.
|
|
#
?
Apr 1, 2021 17:40
|
|
|
I think half of what people mean by "nothing happens" is "nothing progresses". In a 3.x+ D&D fight, as was said earlier, winning is mostly inevitable. A miss doesn't mean things get interesting, it just means the fight HP counter doesn't go down this turn. You did not progress the fight. Your "turn" was "wasted". This also applies to a lot of non-fail-forward skill checks: I failed to unlock the door so no progress is made. I failed to convince the guard so he stonewalls us. I failed to jump the ravine so I'm stuck on the wall and need to make another roll to climb up. Since there's no progress it feels like "nothing happened". Hitting the orc but taking 1d6 damage yourself for doing so, that's progress. Making the jump and losing an item, that's progress. Failing to unlock the door when you know there's an alternative route, that's progress. The other half is when you fail at something you, narratively and/or emotionally, feel you "should" have succeeded at. A character is 1d6 HP away from going down in a bad way and two or three players work in concert to formulate a multi-player plan of action... well you made the plan. Two people gave up their turns to get there. You've "earned" the hit. If you miss... well, that's bullshit. All that effort and "nothing happened". Or you have exactly one spear and huck it at the nearly dead orc... well, it would be real cinematic if you hit with this desperate throw and saved your friend's life. You'd give up a bunch of in character resources to make that happen! But instead you just miss and the cleric goes down. So yeah something happened from your miss, but realistically "nothing happened". Even just rolling a 2 when you needed a 3 to hit can feel like an unearned miss. And yeah I don't think I need to elaborate on when and how this can occur outside combat. Neither of these cases is necessarily solved by making "something" happen when that "something" is worse than if you hadn't tried at all. For the former the bad something would need to be some form of progression; "You missed AND you dropped your sword lol" isn't really progression toward the end of the fight in 3.x+ D&D, it's just made getting to the end more tedious and expensive. For the latter the bad something would have to be in addition to or an alternate way of succeeding at the thing you really wanted to/feel like you earned doing; again "You failed AND made things worse lol" won't really scratch your itch. A good fail forward system solves both these issues but it's far from the only solution, and definitely not an always best panacea. Splicer fucked around with this message at 18:30 on Apr 1, 2021 |
|
#
?
Apr 1, 2021 18:27
|
|
|
Siivola posted:This is a really good point! I’ve been playing a character who mainly deals damage in fights, and I've noticed the first couple of hits in particular feel completely pointless. Regardless of how hard I hit the monster, someone else probably needs to finish it off. My only contribution is whether that takes two or three attack actions. Yes, I really appreciate the set up/take down paradigm in use in games like Fellowship and Voidheart Symphony, where enemies have one hit point and sometimes a checkbox that says "set up?" (The big ones have multiple individual features, each with a checkbox and one hit point.) Every success in combat is a setup or a takedown. Leperflesh posted:So, what I'm seeing now is more of an interesting focus on combat. Makes sense given the focus for many games. But what about noncombat? Sometimes a skill roll is crucial and sometimes it isn't. In a typical melee with three to five PCs against some bad guys, yes, the players probably know the outcome is a foregone conclusion - they're going to win - and the distance between the worst-case and best-case scenarios are perhaps a matter of some spent resources and a certain amount of attrition. It's interesting to see you say that - isn't there much less capability to spend resources to improve the circumstances or results of skill rolls, as compared to combat rolls?
|
|
#
?
Apr 1, 2021 19:08
|
|
|
Splicer posted:I think half of what people mean by "nothing happens" is "nothing progresses". In a 3.x+ D&D fight, as was said earlier, winning is mostly inevitable. A miss doesn't mean things get interesting, it just means the fight HP counter doesn't go down this turn. You did not progress the fight. Your "turn" was "wasted". This also applies to a lot of non-fail-forward skill checks: I failed to unlock the door so no progress is made. I failed to convince the guard so he stonewalls us. I failed to jump the ravine so I'm stuck on the wall and need to make another roll to climb up. Since there's no progress it feels like "nothing happened". Hitting the orc but taking 1d6 damage yourself for doing so, that's progress. Making the jump and losing an item, that's progress. Failing to unlock the door when you know there's an alternative route, that's progress. Oddly enough, I think some of these problems in D&D specifically have been exacerbated by the loss of time-keeping and wandering monster mechanics. Failing a lockpicking check with those mechanics still in place becomes a failure of consequence because you've not only failed to pick the lock, but you've wasted 10 minutes of in-game time trying to do so and have given the GM another opportunity to roll for wandering monsters. That's of course not to say that earlier editions of D&D were completely free of boring or inconsequential failure (Because they still had a lot of that) just that as some of the earlier mechanics have fallen by the wayside, it's only exacerbated the problem.
|
|
#
?
Apr 1, 2021 19:28
|
|
|
Glazius posted:It's interesting to see you say that - isn't there much less capability to spend resources to improve the circumstances or results of skill rolls, as compared to combat rolls? That depends a lot on the system. In D&D, there's piles and piles of magic spells that matter; in other systems, there may be cooperation mechanisms, resources you can use (toolkits, expendables, potions, NPC allies) or minigames (hacking in cyberpunk games for example), etc. And in many games, roleplaying in a noncombat encounter grants a bonus or may even bypass a die roll that would otherwise just be a skill roll. Splicer posted:The other half is when you fail at something you, narratively and/or emotionally, feel you "should" have succeeded at. An approach previously mentioned is player empowerment to control when and whether a risk of this is going to be there or not. E.g., if as a player I made a conscious decision not to spend a limited "I auto-succeed" resource, and then I roll and fail, even if the odds were heavily in my favor, well: I understood the risk and made a choice. To me this is a fundamental aspect of playing a game: as a player, I am empowered to make decisions, in which I have some but not absolute predictive power as to the outcome, or for which there is no clear-cut "always best" choice; and then I find out the results, and move on to the next decision. Whether the uncertainty of outcome is due to random generation, a shuffled deck, or action by another player or the referee, what "feels good" to me is when I can engage with the system and explore the options and not feel as though any of these feelbad things happened:
That last item is iffy but what I mean by it is like, trap choices. Someone says "pick a feat, they're all cool or flavorful" and I pick one that sounds good, but then because I wasn't a rules expert, I didn't understand that actually this feat does almost nothing or is way worse than many other feats I could have taken. This is a bit different from choosing at random, in that maybe I was misled by the flavor text for the feat, or whatever. I made a choice that wasn't fully-informed, through no obvious fault of my own. There's a wide range of choice types that don't fall into those four feelbads, and many of them include the possibility of a "failure" outcome. In poker, I have pocket aces in position and am shortstacked; the right choice, statistically, is to shove all-in. My opponent draws to a flush and beats me. Did I make the wrong choice? No. Did I fail? Yes. If I'm at all good at poker, I'm aware that this is just how variance works, and should be satisfied with my play. But in a roleplaying game, we may not always get thousands more chances for our calculatedly-correct choices to smooth out a curve. That's where I'm referring to pivotal moments - most attack rolls aren't so important that a miss is career-ruining, but as in the spear example, this time, I didn't save my cleric buddy even though I made a decision based on heavily favorable statistics, and it doesn't matter that I'd have made that throw 85% of the time, it feels bad. This doesn't fit into any of the four categories I mentioned above, so I'm adding a new one that doesn't apply to a game like poker, but does apply to a roleplaying game:
IMO this is where a game should give players the option to not have to leave it to chance, as I led with. In the previous Modiphius 2d20 example: a player can spend resources to auto-succeed on dice, and can even negotiate with the GM for a "success with consequences" skill check (in this system, combat is also skill checks) wherein success at the attempt is locked-in, and you roll dice only to discover whether and how many Complications are generated. The resources are limited, and the GM doesn't have to agree to let you use that option, so the practical effect for play is that you use this only rarely... like, when it really matters to you. There's surely other games with similar options to mitigate the feelbads of making an apparently correct choice, just having luck go against you, and winding up with an outcome that isn't satisfying for you or for the table.
|
|
#
?
Apr 1, 2021 20:13
|
|
|
RE: Skill tests outside combat The RPG I've played the most of is probably Delta Green. Delta Green is a game about investigation, and whether it's fun or not depends very strongly on how the scenario and the person running the game handle skill checks. If you demand die rolls for everything and the result of a failed roll is just blocking off an avenue of approach, the result is utter misery. You spent an hour investigating and planning? gently caress you, you flubbed the skill roll. Back to square one, rear end in a top hat. You spend the session beating yourself against walls and feel like you wasted your time. If the Handler is more conservative about asking for skill tests, lets players succeed automatically at certain things without rolling ("with 50% in Bureaucracy, you find the file easily"), and uses failed rolls to raise the stakes and increase the consequences of finding clues, the result is much closer to the Delta Green experience as advertised. "The characters must pay a price for success and choose whether getting information is worth the sacrifice" is not only more fun than "the characters are incompetent and gently caress up constantly", it also fits the game's themes better. I understand that too much failing forward can also be unsatisfying for the players - if you always get what you need, it starts to feel as though your decisions don't matter, either in character creation or gameplay. The flip side is that if someone dips serious investment into a really obscure skill, and it actually comes up, I err on the side of just giving it to them rather than giving them the 30% miss chance on something that may never come up again. Gumshoe, Delta Green and the latest ed of Call of Cthulhu have all basically converged on the idea that critical information and the ability to progress the core of the scenario should not be locked behind skill tests. The degree to which this is actually implemented varies strongly across the modules and across Keepers. Whether to gate something off entirely on failure, give partial information/success, give success at a cost, or just not require a roll at all is something these games try to teach, but also has to be learned through experience.
|
|
#
?
Apr 1, 2021 20:36
|
|
|
Splicer posted:I think half of what people mean by "nothing happens" is "nothing progresses". In a 3.x+ D&D fight, as was said earlier, winning is mostly inevitable. A miss doesn't mean things get interesting, it just means the fight HP counter doesn't go down this turn. You did not progress the fight. Your "turn" was "wasted". This also applies to a lot of non-fail-forward skill checks: I failed to unlock the door so no progress is made. I failed to convince the guard so he stonewalls us. I failed to jump the ravine so I'm stuck on the wall and need to make another roll to climb up. Since there's no progress it feels like "nothing happened". Hitting the orc but taking 1d6 damage yourself for doing so, that's progress. Making the jump and losing an item, that's progress. Failing to unlock the door when you know there's an alternative route, that's progress. I agree with this. Fail forward is such a change that the first time you encounter if after being used to 'welp, either someone manages to make the roll or the adventure just kinda stalls here' is eye-opening as a GM. I am a huge proponent of the idea that player characters should have a domain in which they are good and can be confident of success. There should be a thing that a character can do that they will succeed at barring truly extraordinary circumstances. I'm a professional in my field, for example, and there isn't a 5% chance every time I do a routine task that I'm going to critically fail at it. It may cost character resources, or create a narrative consequence, but it's important to let players have that thing they can be confident their character can do. After all, Luke Skywalker fumbling a pilot check and running into the Death Star's wall is a terrible story, but having the 'group' sacrifice resources to let him take the shot is great.
|
|
#
?
Apr 1, 2021 21:35
|
|
|
Liquid Communism posted:I agree with this. Fail forward is such a change that the first time you encounter if after being used to 'welp, either someone manages to make the roll or the adventure just kinda stalls here' is eye-opening as a GM. It seems "fail forward" has had two meanings at different times. Sometimes it means "if you fail the roll something else happens". Sometimes, it means "if you fail the roll, you succeed but something goes wrong later on" (eg, "if you roll low to pick the lock, you do open the door but the guards get a bonus to detect you because they heard something weird"). I think the last one was taken from Sorcerer at some point?
|
|
#
?
Apr 2, 2021 00:55
|
|
|
|
| # ? Apr 28, 2024 18:28 |
|
|
Matt Campbell talks about it in his rant on failure states. Basically the logic is never letting a failed roll derail the game. He does a great explanation of what it looks like using a scene from Raiders of the Lost Ark as an illustration, check that out.
|
|
#
?
Apr 2, 2021 01:33
|
|