D&D General Story Now, Skilled Play, and Elephants

If you are not familiar with charop, there are acres of examples on various forums. Google Hexblade or Sorlock 5e. I hope you will see that what you are asking is unnecessary. People have made and played optimised characters in 3rd ed and now 5th ed for decades.

When I was running 3e a lot in the 2000s this was pretty massive. It certainly varied by group. Some groups frowned on things being overly optimized. But I had a number of players who loved optimization (that seemed to be one of their big attractions to the game and where they had much of their fun). So I ran several campaigns for groups that were mostly optimizers and some of the things you could pull off in that system with class dipping, taking the right feats and getting the right magic item (which was one of the reasons wish lists were so popular then) was stunning. I actually found that the CR EL system fell apart completely with such groups and I as the GM had to develop system mastery in order to keep up with my players. I wasn't naturally inclined to system mastery or optimization, but I just literally couldn't throw them a decent challenge if I didn't optimize my threats. Actually pretty thankful for that experience because it taught me to appreciate the style of play and improved my overall ability to understand RPG systems. And it was fun once you knew how to do it
 

log in or register to remove this ad

Do you roll dice behind the screen?

Generally speaking I don't roll behind a screen (I don't use a screen), and would only call for secret rolls in cases where players wouldn't know something (like if they make a divination roll checking for favorable omens, I find the player shouldn't know if the roll was successful or not; that there needs to be some room for doubt; or rolling to see if they detect someone putting poison in their drink). For something like rolling for a random guard encounter, or rolling randomly for details about the guard, unless the latter involves rolling on a table the players can see, I would just roll in the open but not let them see the table. The reason I like to roll in the open usually is to build trust. It also keeps you honest as a GM. And I like 'let the dice fall where they may' (which I think works best with out in the open rolls)
 

@clearstream

In focusing in on feelings I feel you substantially undervalue the integrity of the play space. That leaves room for blatant manipulation of play (as long as the players do not suspect it) which regardless of player feelings is not skilled play in any meaningful way in my opinion. For play to be skillful I think the GM needs to be meaningfully constrained by player actions and not just appear to be constrained.
Ah, I am sorry to have given that impression. That is not it at all. I very much believe everyone works toward what they call skillful. However -
  1. for technical reasons, I don't believe that they know for a demonstrable fact that what they call skillful is formally skillful
  2. I don't believe that what they exclude from what they call skillful, is necessarily not-skillful
To try to explain 1. With the Elo two players who have never met have a prediction as to which of them will win. With RPG, if we took a player from a group and put them in some random other group (I mean really random, chosen from among all groups on Earth) I don't think we will be in position to sync them up the same way: they might have been seen as skillful in their home group, and seem unskillful in this random group.

Referring to charop, say the group they are dropped into are huge charop'ers, and their DM has shifted encounter difficulty to deal with this uber-munchkin-ness. They might flounder with their unoptimised characters and be felt by the group to be a low skill player. Of course they could over time accept what the group calls skill and adapt themselves to it...

EDIT Also, I don't believe this poses a problem for the debate. We need only declare up front our mode of play (or whatever establishes what we are calling skillful) and then within the terms of that mode we can have a coherent discussion. Where we seeing fractious arguments is across modes. X is skillful! Oh no it isn't. Oh yes it is. Etc.
 

Ah, I am sorry to have given that impression. That is not it at all. I very much believe everyone works toward what they call skillful. However -
  1. for technical reasons, I don't believe that they know for a demonstrable fact that what they call skillful is formally skillful
  2. I don't believe that what they exclude from what they call skillful, is necessarily not-skillful
To try to explain 1. With the Elo two players who have never met have a prediction as to which of them will win. With RPG, if we took a player from a group and put them in some random other group (I mean really random, chosen from among all groups on Earth) I don't think we will be in position to sync them up the same way: they might have been seen as skillful in their home group, and seem unskillful in this random group.

Referring to charop, say the group they are dropped into are huge charop'ers, and their DM has shifted encounter difficulty to deal with this uber-munchkin-ness. They might flounder with their unoptimised characters and be felt by the group to be a low skill player. Of course they could over time accept what the group calls skill and adapt themselves to it...

EDIT Also, I don't believe this poses a problem for the debate. We need only declare up front our mode of play (or whatever establishes what we are calling skillful) and then within the terms of that mode we can have a coherent discussion. Where we seeing fractious arguments is across modes. X is skillful! Oh no it isn't. Oh yes it is. Etc.

Skill is obviously contextual, even within games played with same system (and play priorities which I would argue are part of system design). That's what happens when you have a living breathing human being making judgement calls as part of the process of play. That does not mean the skill was not real. I mean the same could be broadly said for NFL football. The way refs call the games and decisions coaches make on play calls have a huge impact on how a given game will play out. That does not mean there is no skill being expressed.
 

Skill is obviously contextual, even within games played with same system (and play priorities which I would argue are part of system design). That's what happens when you have a living breathing human being making judgement calls as part of the process of play. That does not mean the skill was not real. I mean the same could be broadly said for NFL football. The way refs call the games and decisions coaches make on play calls have a huge impact on how a given game will play out. That does not mean there is no skill being expressed.
Yeah, I'm only seeing half of this conversation, but it floors me that the argument is still being made that because things are messy, you can't tell anything at all.
 

Skill is obviously contextual, even within games played with same system (and play priorities which I would argue are part of system design). That's what happens when you have a living breathing human being making judgement calls as part of the process of play. That does not mean the skill was not real. I mean the same could be broadly said for NFL football. The way refs call the games and decisions coaches make on play calls have a huge impact on how a given game will play out. That does not mean there is no skill being expressed.
That is true, and I would draw your attention to the features of NFL
  1. A single highly regulated mode of play (take a look at this rulebook 2020 NFL Rulebook | NFL Football Operations)
  2. Comprehensive data is collected on every match and player (take a look at this dataset NFL 2021 passing stats - Players | NFL.com)
So the skill-construct for NFL is well defined and we can make predictive statements about what the factors of skill are and where players are on those. We can drop a random player into a random team and - so long as they have a decent dataset - predict the likely performance of that player in that team (caveated on group-effects upon performance).

Here is a thought - say we thought of RPG as more like football games overall. NFL, Aussie Rules, Rugby, League and Soccer. They're all kind of similar - similar sized teams handle balls in regulated ways to get them through goalposts - but they're also quite different. Rugby favours a running game so quicker players are considered more skillful. I know little about NFL, but it seems to have more defined roles and in some roles being great at running won't help you. In Soccer, you can't pick up the ball and you aren't allowed to block tackle or ruck.

Now as I said, compare RPG to that. How many distinct "regulations" are there for RPG: hundreds of different rulesets. Have we ever written out the "laws" in the exacting way that Fide have the laws of Chess (https://www.fide.com/FIDE/handbook/LawsOfChess.pdf)? As an aside, I am making an assumption here that sport "laws" are more like written "principles or techniques" than game mechanics - sports mechanics are the laws of human neurophysiology or anatomy, and physics! Crucially, where is our rigorous dataset across thousands of players?

EDIT On rereading your post I feel we are close on our views. I am adding that without the sort of data NFL collects, we have to be wary that our testaments to skill may speak more to what we value seeing in play - what we prefer and find satisfying - and the overall skill of our group + difficulty of tasks our DM presents us.

EDIT EDIT And I can see how that caveat can feel undermining. Please take it that I am advocating humility and caution. Say ingenuity feels skillful to you, then we can both call that skillful and talk about forms we have seen it take, and learn from that. If we then try and club another player over the head with our belief in our true skillfulness... that is what I am against.
 
Last edited:

So I don't think of RPGs broadly like the NFL. Not even like football. More like on the same level as sports or board games. Skills don't transfer from game to game. Mostly the only thing that does is your ability to reason about the fictional space. I think of Apocalypse World or Moldvay B/X (or any game that has a completed design with objectives and play processes detailed) like I do the NFL. Skills will generally transfer pretty well from game to game, although you might benefit from understanding how the ref calls stuff.

Mostly I do not think there's a great way to talk about this stuff in a broad way outside of the context of specific games.
 

Sure. I've not experienced this, because I've only played a handful of 3E sessions, and no PF. But it's widely reported, and given that I've done this in AD&D (with clerics and wizards) and 3E only powers up these character types compared to other build options, I believe those reports.

I just don't think Pun Pun is the evidence for it. And I don't think it provides any support for a generalisation to other RPGs that connect PC building to player skill.
Pun Pun in absolute isolation, perhaps not. But Pun Pun in context? Yeah, I kind of think it does demonstrate a (degenerate) instance of a big part of 3e-era skilled play for a variety of reasons:
  • Optimization in 3e, with few exceptions, is almost purely self-focused. Buffing one ally < buffing yourself; in-combat healing < dealing damage; teamwork with the potential for payoff < personal effort with direct payoff. I once lamented that 3e optimization was an exercise in each individual player maximizing their personal contribution to the exclusion of party play...and had several expert 3e optimizers chime in with agreement. Pun Pun demonstrates a (degenerate) side of this through personal nigh-infinite numbers loops that do nothing to benefit allies.
  • Expert knowledge of several highly disparate sections of the rules is, at least according to some of 3e's designers, an intentional expectation of "learning to properly play 3e." Pun Pun in particular requires a diverse array of things: Divine Minions, Savage Species options, prestige classes, and Monster Manual entries, just to name a few. This is a well-known aspect of 3e skilled play: the degree to which you have working knowledge of the numerous options available directly correlates to the power you can wield if you so choose.
  • The expectation, in 3e, that the rules will be followed more-or-less as written unless there's a good justification for doing so. Though I think people often play up how dramatic the difference is between prior editions and 3e, there is something to the notion that, when people played 3e, they expected that the rules as written (particularly the player-facing ones) were in some sense part of the social contract, and thus could not be changed without a reasonable rationale first.
Yes, Pun Pun is degenerate. But it is degenerate in order to highlight some of these characteristics that appear in 3e skilled play, much as the Tomb of Horrors is a degenerate demonstration of certain characteristics of older-school play (incredibly high lethality, "puzzle" monsters, difficult-to-foresee but very dangerous traps, the struggle between player-made SOPs and DMs wanting to induce difficult choices, etc.) By throwing these characteristics into sharp relief, without any of the typical mitigating elements present, both things communicate something about the context in which they appeared, even though neither truly represents its associated style.
 

Dismissing Pun Pun as one degenerate case fails to think through what is going on in 3rd and 5th edition charop. System mastery in those two games allows construction of mechanically stronger characters. The intended game balance as laid out in the DMG, e.g. as guided by CR, entails that those characters will defeat encounters with ease that ought to be matched for their level. The tweak is for the DM to adjust encounters although often what happens is that the optimised character joins a group of unoptimised characters so that route isn't open.

<snip>

It is old news that more skill in the area of charop in 3rd and 5th edition lead to needing less skill during play (given a group is following the written principles of the game).
When you refer to CR and to defeating encounters with ease are you referring to combat encounters? What about (say) Tomb of Horrors, which (notoriously) contains very little combat?

More generally, it may be that very little 3E or 5e play involves engaging the fiction as opposed to playing a miniature skirmish game. But once we get to those episodes of 3E and 5e play that do involve playing the fiction, what role does character build play? And - more importantly given your claim - how does that affect the skill of play in the moment? My experience in playing AD&D is that many (not all, but many) players who are better at building mechanically effective characters (which in AD&D mostly concerns spell load-out) are also better at engaging the fiction when that is necessary, including by "creative" spellcasting. This experience of mine doesn't support the claim that the display of skill in one arena of play makes it unnecessary to, or is at odds with, displaying skill in another arena of play.

Each time you have asked for examples, you have quibbled the examples. And often not on the terms that the example was requested for. For example with the force example, the quibble was that it felt like it was teaching. We were not debating teaching, we were debating if force could be used in any way positively.
Your example in Burning Wheel is a different case.
Re force: You had posted (#285) that
A GM could certainly use force unskillfully which would very likely mean it was dissatisfying to players and limited their opportunities to play skillfully. On the other hand, a GM could use force skillfully to elevate the challenges and open up opportunities for players to play skillfully.

<snip>

I am saying that any DM decision is open to being skillful or unskillful, elevating or deflating to player skill.
This appeared to me to be a claim that a GM, by using force (eg by disregarding dice rolls or manipulating the hitherto unrevealed fiction) in a moment of play could enhance the degree of skill in that same moment of play.

But then you gave an example of using force to blunt the consequences of a player's declared and mostly-resolved action so that they might (i) learn about consequences and (ii) do better next time. Which is what I described as teaching.

Do you have an example that actually cashes out the claim you made that I have just quoted? Ie of the use of force in the moment of play to enhance the skillfulness of play in that moment? The only answer I can construct from your posts is that you treat all GM narration as the use of force. But perhaps I haven't drawn all the inferences that your posts support, or perhaps I have misunderstood your claim?

Re Pun Pun, you posted (#333) that
For RPGs with strong character mechanics, such as 5e, we have the peculiar problem that the more skill a player has in one area of the game, the less they will likely need in another.
Now I don't know exactly what you mean by strong character mechanics but I assume that 4e D&D and Burning Wheel would both count as examples, whereas presumably Prince Valiant and Cthulhu Dark wouldn't. I don't know about The Green Knight or Dungeon World, though I would think at least the latter might be on the "strong" side. It's not remotely obvious to me why Burning Wheel would be a "different case" from RPGs with strong character mechanics. It's presumably an example of such an RPG; that's why I mentioned it.

And my experience with BW, and also with 4e D&D and DW, make me doubt your claim (of course likely blunts the force of the generalisation, but if you're confining it to 3E and 5e D&D than why not just say so?).

Hence I asked (#372) "Are these conjectures based on actual play experience? Are you able to give examples that illustrate what you have in mind?" And you replied with Pun Pun, which is not an actual play example and doesn't (as far as I can tell) illustrate any general tendency for skill in RPGs with strong character mechanics to be zero sum either in when it is called for or when it is demonstrated. And even confined to 3E D&D, the fact of Pun Pun doesn't show that there is not the need to demonstrate skill in leveraging the fiction when other builds are being used.

So the reason I quibble with your examples is that they don't seem to address the concrete claims I take you to be making. This is further complicated, for me, because those concrete claims are mixed with other claims that I struggle to understand.

Perhaps I am reading your posts too literally? I don't know. In reply to my post (# 401) that mentioned "RPGing in which playing with skill is a (or the) major preoccupation of the game?" you replied:
I like this approach. I would modify to say that it is impossible that playing with skill can be what is demonstrably happening for reasons you note (self-reporting), but that it is very possible to be preoccupied with playing in the way we call skillful (us being the group, and the wider community for whom the concepts and reports resonate).

Where some posts have most failed for me has been in claiming the high ground. I feel forced to resist your original framing not out of sheer pettiness, but because I read posts that seem to take it that if they are "playing with skill" then they know what skill is and is not, and can go on to dismiss what other groups call skillful.
Your "modification" rests on an apparent misreading of what I posted. I posted about RPGing in which playing with skill is a major preoccupation of the game. This would be a RPG in which (in this respect) play more closely resembles (say) chess than it does (say) snakes-&-ladders. I didn't say, nor conjecture, nor posit, anything about whether or not anyone is actually playing with skill (demonstrably or otherwise): I'm referring to what play demands of players, not what they may or may not offer up in response to that demand.

If you are denying that I can tell whether or not I, as a RPGer, am or am not concerned with skill, I simply disagree. I can tell when I am having to engage in the sort of skilled play that Gygax advocated (attention to gear and spell loadouts; careful manipulation of the fiction especially in respect of architecture; etc); likewise I can tell when I am playing a game that lets me ignore that sort of thing and just focus on inhabiting my character. This is just the same as I can tell the difference between playing bridge or five hundred, between chess and snakes-&-ladders, etc. And I didn't need Elo to invent his model of chess rankings to work out that chess is a skilled play game whereas snakes-&-ladders is not!

For similar reasons, I don't agree with your sceptical claim here:
Skill can't be known to loom large, only that the play satisfies the group and makes them feel like they exercised skill.
Of course people can overestimate (or underestimate) how well they played a game, but I think most of the time they can tell whether or not they were being called upon to exercise skill.

What I've been attempting to do is go beyond what amounts to statements about what is found satisfying to a group.
Whose statements do you have in mind? And which ones?

For instance, when I say that the play of classic D&D is fundamentally different from the play of Burning Wheel - with my evidence base being the game texts, the commentary of the designers found both in and beyond those texts, my own play experience, other's play experience, and the broader cultures of play associated with each - I don't take myself to be making a statement about what is found satisfying to a group.

One of the clearest skill constructs we have in gaming is the Elo. This has been emulated by Microsoft's TrueSkill and in other games such as League of Legends MMR.
I don't know the Elo other than what I've just read on Wikipedia. It seems to be a way of generating a ranking based on comparisons of performances in two-player competitive games. I don't quite see what bearing this sort of measure has on RPGing. Even classic tournament play doesn't seem to me to generate the sorts of pairwise comparisons that would be needed to implement a ranking of this sort.

It also doesn't tell us what skill consists in. And it takes as a premise that playing the game well requires skill. Quoting from Wikipedia:

Elo's central assumption was that the chess performance of each player in each game is a normally distributed random variable. Although a player might perform significantly better or worse from one game to the next, Elo assumed that the mean value of the performances of any given player changes only slowly over time. Elo thought of a player's true skill as the mean of that player's performance random variable.​

It's not so much the mathematical demonstration, as the concrete construct and reducing of confounds around task modelling, task difficulty, and small group self-perceptions. When posters in this thread and others like it speak about skill, I feel at best they can be speaking about what their small group has experienced (in this context, less than a thousand is small).

<snip>

I find that other posters are speaking ambiguously and making it impossible to discuss real differences. However, I am happy to revert to discussing experiences and preferences. One can learn from sharing those. One cannot make claims to be saying anything concrete about skill, however. Well, that is a little harsh: one can make guesses and speculate as to factors (e.g. ingenuity) and say what factors one values, but one cannot say that one self-reported example was any more skillful than any other - even in cases where the other sounds on surface less skillful
The only basis on which Elo's assumptions about chess rests is the widespread intuition that chess is a game of skill, in which the more skilful player will typically beat the less skilful one, and that any given victory is probably a demonstration of greater skill on that occasion. (Unless you treat him as offering a stipulative definition of skill. But there are many reasons to think that he isn't; and chess players engage in all the sorts of activities one associates with skill-based pastimes: they practice, they get coaching, they study the successes and failures of other players, etc.)

In the context of RPGing, if a single person A asserts that on this occasion of RPG play, skill was demanded of them, and on this other occasion it was not, why is that not a reliable intuition? Particularly if they can articulate their experience by reference to received markers of skill in RPGing - having regard to win/loss conditions, having regard to how that fiction might arise and what factors leading to or away from it the player sought to manipulate, etc.

I don't understand on what other basis you intend to make claims about the role of skill in RPGing. What conjectures are you hoping to establish, or disprove? What tasks are you intending to model? I am still having real trouble making sense of these elements of your posts.

I have played and/or DMd - Dogs in the Vineyard, Burning Wheel, RuneQuest, D&D (all editions), HeroQuest, Earthdawn, Shadowrun, Bushido, Empire of the Petal Throne, Land of the Rising Sun, Chivalry and Sorcery, Ars Magica, Traveller, Aftermath, Paranoia, GURPS, Savage Worlds, a diceless game of my own invention (influential in my small part of the world), Amber, Rifts, Champions, two or three games of friends' invention. We've probably one-session'd some others that I don't recall.
Thanks for this list. Does this mean that you haven't played or GMed D&D?

I've bolded two of the games you mentioned, because Ron Edwards has made the following remark about them:

From Maelstrom (Hubris Games, 1994, author is Christian Aldridge):

Literal vs. Conceptual
A good way to run the Hubris Engine is to use "scene ideas" to convey the scene, instead of literalisms. ... focus on the intent behind the scene and not on how big or how far things might be. If the difficulty of the task at hand (such as jumping across a chasm in a cave) is explained in terms of difficulty, it doesn't matter how far across the actual chasm spans. In a movie, for instance, the camera zooms or pans to emphasize the danger or emotional reaction to the scene, and in so doing it manipulates the real distance of a chasm to suit the mood or "feel" of the moment. It is then no longer about how far across the character has to jump, but how hard the feat is for the character. ... If the players enjoy the challenge of figuring out how high and far someone can jump, they should be allowed the pleasure of doing so - as long as it doesn't interfere with the narrative flow and enjoyment of the game.

The scene should be presented therefore in terms relative to the character's abilities ... Players who want to climb onto your coffee table and jump across your living room to prove that their character could jump over the chasm have probably missed the whole point of the story.​

. . .
I can think of no better text to explain the vast difference between playing the games RuneQuest and HeroQuest.​

There is a lot going on both in the passage quoted from Maelstrom Storytelling and in the short remark that Edwards himself makes. But one consequence of the differences is that, in RQ, it makes sense - once the difficulty of the jump across the chasm is established by the GM - to try and identify advantages that will give a bonus on the relevant check and hence decrease the chance of failing. Conversely, in HeroQuest (and especially HeroQuest Revised) the player is not expected to try and manipulate the difficulties of tasks, which are set by the GM essentially by reference to pacing considerations. That's not to say one doesn't leverage the fiction, but to other ends - like establishing what is at stake and what the consequences - not in win/lose terms, but in purely fictional terms - will be. (Maelstrom itself uses various mechanical devices within its scene resolution system to handle this.)

Those differences in how the game plays have fundamental implications, in my view, for how one thinks about them from the perspective of "skilled play". For instance, I can just about conceive running Tomb of Horrors using RQ rather than AD&D as the basic mechanical framework (and I have in fact toyed with this using Rolemaster). But the idea of running ToH using HeroQuest seems to me like it makes no sense at all.

I wonder if you agree?
 

If you are not familiar with charop, there are acres of examples on various forums. Google Hexblade or Sorlock 5e. I hope you will see that what you are asking is unnecessary. People have made and played optimised characters in 3rd ed and now 5th ed for decades.
I'm pretty familiar with those examples. I didn't ask to be presented with them. I stated what would count as evidence for a certain claim:
Pun Pun isn't evidence in favour of this. Showing me the build specs and likely play for a cleric at (say) 1st, 3rd, 5th, 7th and 10th levels would be.

Because my point was about what would count as evidence, I didn't need to discuss whether or not the evidence was actually available. My reason for not expressing a strong view on the second point - ie the actual availability of the evidence - is that ENworld has many posters who deny that, in 3E or 5e D&D, it is possible to build characters with widely differing effectiveness, and I don't really care to get into an argument with them on the point. (I'm happy to make the claim about AD&D, as I've experienced it directly and hence am not relying on the reports of others as I would be were I to get into a debate about the 3E or 5e examples.)
 

Remove ads

Top