Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Model-Based Testing for Dungeons & Dragons

loskutoff.com

Post reply on HN
Print view View original

61–70 of 78 posts

Re: Model-Based Testing for Dungeons & Dragons

#61

Post by jghn »

bluefirebrand wrote: post #55
pigpop wrote: post #36

Earlier quoted context omitted.

I agree, tabletop RPGs should be 2/3 Role Playing and only 1/3 Game.

This is why I don't play TTRPGs much anymore. If I wanted to really get into character and roleplay my heart out I'd join an improv group... I liked the crunchy bits of TTRPGs like charting dungeons and running kingdoms and shit. Can't find many groups interested in that anymore

It's why I haven't tried to get back into TTRPGs. I suspect that even if I were to join in on an OSR group that the play style would still be closer to the modern improv group aesthetic.

Re: Model-Based Testing for Dungeons & Dragons

#62

Post by gavinray »

cammasmith wrote: post #19

One of my biggest issues with playing DND is that I never fully understood the rules. I'd play with people who had been playing for years, and they didn't explain things very well, and that made it hard to play. Hopefully, this will help with that.

Baldurs Gate 3 the video game taught me DnD, videogames where you can go at your own pace are a nice option.

Re: Model-Based Testing for Dungeons & Dragons

#63

Post by krapp »

pigpop wrote: post #35
krapp wrote: post #21

Earlier quoted context omitted.

You think there will still be writing jobs for human beings at all by then? AI will be so normalized across culture that any raw, unfiltered human expression will read as gross and unprofessional by most people.

Maybe for resume cover letters and LinkedIn posts but I haven't met anyone with half decent taste who prefers AI writing, even well prompted, to skillful human writing. I'm not a stranger to using AI for writing tasks by any means but it's only ever a starting point that gets heavily rewritten by both myself and the model.

>Maybe for resume cover letters and LinkedIn posts but I haven't met anyone with half decent taste who prefers AI writing, even well prompted, to skillful human writing.

That attitude is one, maybe two generations away from extinction. Taste is created by the market, which caters to the young. When enough people have been born into a world in which AI generated culture and communication is the norm, that is what will define what good taste is. People like you (and I) will just come off like old people yelling at clouds.

We can already see this happening at the fringes. People have relationships with AI, they prefer AIs to real people, they use AI as a primary source of truth, they consider AI generated art to be superior to human work, they trust AI more than people. People identify as AI. AI is filling an emotional, sociological and creative space that an increasingly alienating and hostile society denies to people, for better or worse. Generative AI has only been a thing in popular culture for four years or so and it has already completely transformed human society and human sociology.

Barring a complete collapse of the AI bubble, which seems existentially impossible at this point given how invested our economies and government are in it, that's just what normal is going to be in a decade or so.

Re: Model-Based Testing for Dungeons & Dragons

#64

Post by nextlevelwizard »

e28eta wrote: post #37
nextlevelwizard wrote: post #17

Earlier quoted context omitted.

Yeah, both players were either rogues or tabaxi (although feline swiftness isn’t dashing) This is also directly why I don’t like D&D. It is way too combat focused and video gamey. If your combat system is so complex that people find (or even feel that they need to find) “exploits” in it then your system probably sucks. So many class features are purely combat focused completely ignoring the actual roleplaying part of…

re: grapple leapfrog, it links to this question: https://rpg.stackexchange.com/q/136964 Maybe the AI used the accepted answer (with 4 votes vs the next with 39) and then mangled things from there? re: counter chaining, I think so. I spent some time watching Critical Role and iirc they liked to counterspell a counterspell.

First one is pretty easy. The player is just trying to do RAW instead of intention. Intent obviously is that while dragging every feet of actual movement costs two of your characters movement allowance, so dropping the burden doesn’t give you more movement.

Countering a counterspell feels like a waste since for one you have to have another caster with counter spell and now they are wasting their reaction plus a slot instead of just going another round. I guess there are situations where that makes sense, but somehow feels bad

Re: Model-Based Testing for Dungeons & Dragons

#65

Post by nextlevelwizard »

pigpop wrote: post #32
nextlevelwizard wrote: post #17

Earlier quoted context omitted.

Yeah, both players were either rogues or tabaxi (although feline swiftness isn’t dashing) This is also directly why I don’t like D&D. It is way too combat focused and video gamey. If your combat system is so complex that people find (or even feel that they need to find) “exploits” in it then your system probably sucks. So many class features are purely combat focused completely ignoring the actual roleplaying part of…

From my limited experience, many players and DMs seem to get things backwards in exactly the way you're describing. They take the rulebook as the starting point or the "controls" for the game and since combat is the most detailed they tend to focus on that to the exclusion of other parts of the game. I've always viewed the rules as a way of settling disputes or uncertainty instead, so you start from the role playing…

When most of the games rules are about a thing that thing becomes the focal point. 5e also assumes pretty high amount of combat encounters per day to keep all classes in balance, if you are having less then it will make some classes just bad picks which can feel bad

Personally I don’t like it when people don’t play by the rules of the game we have decided to play together, so definitely things should work as the rules say and then ambiguous things are sorted with GMs world’s logic as “rulings”.

If you start by ignoring the rules and only consulting them when there is a dispute then I want to play another game with less rules to begin with

Re: Model-Based Testing for Dungeons & Dragons

#66

Post by nextlevelwizard »

antisthenes wrote: post #40
nextlevelwizard wrote: post #17

Earlier quoted context omitted.

Yeah, both players were either rogues or tabaxi (although feline swiftness isn’t dashing) This is also directly why I don’t like D&D. It is way too combat focused and video gamey. If your combat system is so complex that people find (or even feel that they need to find) “exploits” in it then your system probably sucks. So many class features are purely combat focused completely ignoring the actual roleplaying part of…

> If your combat system is so complex that people find (or even feel that they need to find) “exploits” in it then your system probably sucks. Couple of things. 1. People will try to find exploits in just about any system. That's kind of part of the fun. 2. If the difficulty curve sucks in a particular D&D campaign - that's the DM's fault, not the system's. Plenty of tools at DM's disposal to make campaigns less comb…

Eh, I don’t find it fun because if you can break the combat then you either decide not to play “optimally” or GM has to purposefully create situations to fuck with you specially which is just antagonistic

I don’t know how you go to difficulty curve

Re: Model-Based Testing for Dungeons & Dragons

#67

Post by pigpop »

ghurtado wrote: post #39
pigpop wrote: post #35

Earlier quoted context omitted.

Maybe for resume cover letters and LinkedIn posts but I haven't met anyone with half decent taste who prefers AI writing, even well prompted, to skillful human writing. I'm not a stranger to using AI for writing tasks by any means but it's only ever a starting point that gets heavily rewritten by both myself and the model.

It's not even just for writers either. If I was currently hiring, not using AI would be the cheapest, fastest way to impress me. I'm not kidding when I say that typos are not too far from becoming a sign of higher intelligence. Or at least better taste than most. I'm surprised tunable intentional "human" mistakes are not a core feature of LLMs. Maybe it's actually hard for them?

It's not hard to get them to copy a style, you just have to provide examples and they will happily produce similar text including grammatical and spelling mistakes. The trouble is with the composition and novelty. Most of the big models have had all of the interesting parts hidden behind a wall of RLHF. Local models are better since you can use ones that are not indoctrinated as a "helpful assistant" and also control the system prompt, temperature and see the top K alternate tokens which let you steer them in interesting ways.

Re: Model-Based Testing for Dungeons & Dragons

#68

Post by pigpop »

krapp wrote: post #63
pigpop wrote: post #35

Earlier quoted context omitted.

Maybe for resume cover letters and LinkedIn posts but I haven't met anyone with half decent taste who prefers AI writing, even well prompted, to skillful human writing. I'm not a stranger to using AI for writing tasks by any means but it's only ever a starting point that gets heavily rewritten by both myself and the model.

>Maybe for resume cover letters and LinkedIn posts but I haven't met anyone with half decent taste who prefers AI writing, even well prompted, to skillful human writing. That attitude is one, maybe two generations away from extinction. Taste is created by the market, which caters to the young. When enough people have been born into a world in which AI generated culture and communication is the norm, that is what will…

There's taste and then there's taste

Popular taste is guaranteed to be awful since it is driven by economics and fads. That's the type you point out as created by the market and catering to the young. It's a disposable product of consumption used to sell shoes and overpriced paintings.

I don't disagree that it will permeate everything, it already does. It'll just be written by an AI instead of people being paid to find the next style to cop. I don't think it will extinguish human writing, you'll just have AI writing that you feed to official or public channels and then real writing that goes in private or pseudonymous channels. Using AI writing among friends or an in group will still be a faux pas and cringe because it will have become the norm to be rebelled against.

Re: Model-Based Testing for Dungeons & Dragons

#69

Post by pigpop »

nextlevelwizard wrote: post #65
pigpop wrote: post #32

Earlier quoted context omitted.

From my limited experience, many players and DMs seem to get things backwards in exactly the way you're describing. They take the rulebook as the starting point or the "controls" for the game and since combat is the most detailed they tend to focus on that to the exclusion of other parts of the game. I've always viewed the rules as a way of settling disputes or uncertainty instead, so you start from the role playing…

When most of the games rules are about a thing that thing becomes the focal point. 5e also assumes pretty high amount of combat encounters per day to keep all classes in balance, if you are having less then it will make some classes just bad picks which can feel bad Personally I don’t like it when people don’t play by the rules of the game we have decided to play together, so definitely things should work as the rule…

I guess it depends if you want the game to be a grind for the next level or if you want real interactive fiction. Different people like different things.

Re: Model-Based Testing for Dungeons & Dragons

#70

Post by cthalupa »

jghn wrote: post #60
gglitch wrote: post #41

Earlier quoted context omitted.

There's a community and play-style called OSR or "old school renaissance," that recreates versions of the earliest editions of D&D, and encourage a style of play that's heavily oriented around few rules and the DM making quick decisions/rulings on the spot, rather than lots of rules and lots of time spent mining the rulebooks. In fact, the expression is "rulings over rules." This might appeal to you.

There's a dichotomy here that I have always found amusing. To me, the older style of play felt crunchier, despite there being less of a rule focused. The most common style of play back then was more of a dungeon crawl, closer to "roll playing", low fantasy, usually lower level, murder hobos were very common, and all of that. Whereas today's game is far more complicated rules-wise by most measures yet it tends to be m…

Today's game can be just as much roll-playing, it highly depends on the group. One of the things that drove me back to B/X and ADD was the sheer number of min-max players and rise(!) of murder hobos in 5e vs. even 3e/3.5/pf1

Most of the early old-school stuff was way too deadly for players to be murder hobos or try to solve everything with combat - if you went into Caverns of Thracia at level 2 as a murder hobo you're just going to die over and over and over again. It'll be endless TPKs. Right now I'm two years into DM'ing an Arden Vuul campaign, running a mix of OSE (Streamlined B/X) and OSRIC (ADD 1e) rules, and it's really only the past 6 months or so that my players have felt comfortable engaging in regular combat - before then they might have spent a whole session or two trying to stack up every advantage they could because they never wanted to be in a fair fight.

And from my experience with a whole lot of OSR play over the past 6-7 years is that this is the sort of feel most OSR players are after. They're not wanting to play late ADD 2e, Dragonlance era, where the shift to the more heroic play started happening - they want to have to think and outsmart things. Faction interaction was also huge in the more sandbox environments, and that was where most of the roleplaying occurred then, and occurs now in the games I run. The players RP a bit with each other, but not as much.

Modern D&D is a kitchen sink approach that tries to solve every possible playstyle, and that makes it popular and reasonably good at most anything people want to do with it. But I don't know that there's any facet of it that it does as well as other systems.

Post reply on HN