The Gaming Den

Thaluikhain wrote:Sure, you could do it that way, I was thinking of getting rid of all the wound rolls and armor saves and modifiers. And I was thinking of simple tables with 3 options on them rolling on a d3, with one always being "out of action", so they aren't too complicated.

Admittedly, I'm not sure if that's such a good idea, I just wanted to lighten the rules a lot, and that might have gone too far.

I updated my answer before I saw this comment

Yeah I think infinity skirmish used a single d20 roll for damaging but l can’t say for sure. There’s also combat advantage number CAN discussed on here time to time

Was wondering, are things that insta-kill PCs in a RPG a good idea? For example, pit traps, where you either fall in and die or don't fall in and aren't affected. Or a gaze weapon that turns you to stone or does nothing.

I'd always assumed they weren't, especially when you've got options for similar attacks depleting resources such as HP. Suddenly not having a character doesn't seem like fun, but resource management is something you can play. But I might have missed something.

Honestly... I don't think so. But I suppose it depends on what kind of resources the characters have. If they have limited but reliable resurrection, then maybe it's more interesting. In my experience, tho, at low levels, it's just an annoyance and the player either gets to sit out or has to make a new character.

It depends on what kind of game you're running. As has been discussed in other threads you can have a lot of fun with retro stupid games where PC corpses pile up like cordwood and creating your replacement character is handled by scribbling Ensign Bob or Hank the Barbarian on a sheet before marching right back into the breach.

Generally, I think Save or Die mechanics are actually pretty OK because they're genuinely fast and they're useful to have in place for NPCs even if you decide you want some sort of plot armor mechanic on hand to prevent PCs from dying before their player is satisfied with their arc.

Thaluikhain wrote:Was wondering, are things that insta-kill PCs in a RPG a good idea? For example, pit traps, where you either fall in and die or don't fall in and aren't affected. Or a gaze weapon that turns you to stone or does nothing.

I'd always assumed they weren't, especially when you've got options for similar attacks depleting resources such as HP. Suddenly not having a character doesn't seem like fun, but resource management is something you can play. But I might have missed something.

I feel most poeple like a good noble sacrifice or Pyrrhic victory, but "lol roll u died stupidly" is a niche for humor games

Like Tenra Bansho Zero has mechanics where you're down to 0 hp you're out f the scene but can RP how you show up later "I fell into the river" "I was taken prisoner then..." but then then to use 100% of your character's power you raise a death flag that means 0hp = death. That's a more specific example where story is tied to mechanics.

A pit trap is an encounter.

If a zombie appears out of the shadows, slashes someone with a rusty longsword and then collapses into dust, that's an encounter, too. But compared to a zombie that engages the party and is defeated by the party, it's a pretty bad encounter.

Traps work when the PCs interact with them in a meaningful way. That means noticing them before they work is critical. At the very least, put the trap on a timer. If the PCs hear a 'click' and they have 2 rounds before the trap activates they have a chance to do things (possibly to maneuver themselves away from, or, amusingly INTO the trap).

A trap that is a binary 'you lose 5 hit points' without any action by the PCs is dull. Instead, saying 'you notice that the door you're opening is connected to a pair of ropes; one of them taunt' with a perception check to notice the pair of heavy crossbows pointed at them is better for a couple of reasons.

Firstly, they can choose not to engage. They opt to set off the trap. They throw open the door, shield out, and hope to avoid or eat the damage and move on.

Secondly, if they choose to engage, they can apply their resources to the threat in what they think is an appropriate manner. A globe of invulnerability or trying to disable the trap.

If the time between the trap activating and the trap resolving doesn't include time for the PCs to take actions, it may ablate resources, but it is also a VERY SHORT waste of time.

Did any edition of DnD ever outline how much damage causes a character to bleed?

Like "oh I SPOT ROLLED some blood here..." is a big story convention

1 point will deliver blood agents like poison or diseases, so any amount seems sufficient to draw some blood.

One of the optional rules somewhere in 3.X (I want to say CompWar) is for duels, with various options including "First blood". That suggests a tenth of their maximum HP is the amount to go for (leading to one person on the old WotC boards making a character who would suffer Massive Damage at 1/10 and thus could arguably be killed by a paper cut because clearly a paper cut draws blood and thus MUST do 10% of your HP).

But yeah, a single point is enough to at least break the skin and get a droplet.

Are generic random encounter tables useful? As well as having level appropriate encounters, the monster demographics are going to be quite different depending on which spooky forest you are walking through.

Coming up with custom tables for everywhere seems a bit of work, but you could say this forest has got wolves, this one goblins, this one bandits and tweak from there a little. Though, taking much of the random out of random encounter.

I don't think completely generic random-encounters-by-terrain-type tables are useful since they tend to get really same-y if a party spends a lot of time in similar areas, but it's possible to get a lot of variety without hand-crafting tables for every little region.

When I come up with random encounter tables for a large-ish sandbox, I pick a handful of terrains and/or encounter sites to be "anchors" and draw up tables for those--one table for the Eastern Mountains, one for the Spooky Forest, one for the outskirts of the Town of Generica, etc.--to get the full thought-out thematic treatment, plus one really generic one for the sandbox as a whole.

Then other areas just get encounter tables that are pointers to the anchor tables. The part of the High Road right outside Generica is just 90% "roll on Town of Generic table" and 10% "roll on Whole Sandbox table," the portion near the forest is 60% Spooky Forest/30% Town of Generica/10% Whole Sandbox, the portion near the mountain pass is 80% Eastern Mountains/20% Whole Sandbox, and so forth. That makes different areas noticeably distinct if the party gets to know those areas well ("We're near Three Tree Fork, once we pass that we'll start seeing fewer fey and more bandits"), and you can handle similar terrain features by having multiple similar anchor tables and using different percentages for different regions--say, make a Spooky Faerie Forest table and a Spooky Evil Forest table and have three different spooky forests with 70/30, 50/50, and 30/70 mixes.

And while in my particular case I use hexmaps and have coded up a script to spit out appropriate percentages based on distance between hexes for a smooth but granular progression as you move from anchor to anchor, looking at a sketched-out map and eyeballing works just fine.

Hex environments/'encounter tables' influencing neighboring hexes make the world feel more connected.

The Forest has a mated wyvern's nest, they're going to wander into the Plains and Mountains for food.

The Hills have ancient hidden Wight burrows, if characters of similar worldview and class to the wights sleep in adjacent hexes they may receive a vision.

The Village has an underground gay hangout, incognito gays from the other towns and cities may appear there if you know where to look.

Are there good guidelines for running a playtest for an RPG?

The main issue with such guidelines existing is that 'RPG' is a very broad term. What set of procedures would adequately test both SR4 and Munchausen?

I think the most generally applicable advice is to make sure you test destructively. There's a story I half-remember where a wargame designer was astonished when a novice player tried something stupid (charging his forces uphill into a defended position?) and it worked. Because the more experienced players just assumed such a thing would never work, they'd never tried it, and it had become an undetected bug. So try seemingly stupid things to make sure they're actually stupid and not secret win. And bring in outside people who don't share your assumptions and blind spots.

The version of "wargame broken by what should be a stupid tactic" I remember was Doug Lenat testing his machine learning system Eurisko on Traveller's Trillion Credit Squadron tournament in the 80s, and winning the whole thing by just bringing a flotilla of defenceless, nigh immobile gun platforms. Which was laughed at by experienced players until they saw the resulting win streak and he walked away with the trophy.

And then did it again next year when they introduced a "fleet mobility" rule to stop his pet robot's strategy, which Eurisko worked around by just scuttling any damaged ship and still won. So they threatened to cancel any tournament that had a Eurisko fleet in it, giving Lenat a title of Grand Admiral in consolation.

(Some details there taken from the Traveller wiki. Because of course that's a thing.)

- If you have a single mechanism to test, it's better to play many short scenes with different characters. The scenes should have context and be something that could come up in the game not an "ideal" situation that will never happen.

- Test both with premade character and player generated characters. For premade characters, have both "standard" characters and extreme characters.

- Keep assumptions at a minimum. It's tempting to say "nobody will ever do that" or "that's not the spirit of the game" but unless this is enforced by the game, you will have people trying to do it.

- Don't try to update the rules on the fly, keep things as written (unless it doesn't work at all).

- Take a lot of notes, take notes as soon as you notice an issue. Don't wait until the end of the game.

- If you have your game ready and want to do full tests, you should just give the book to a group that never played any earlier iteration and let them play a game without you. They'll be able to play without any assumption and to play the rules as written, not the rules you think you have written.

Blade wrote:- If you have your game ready and want to do full tests, you should just give the book to a group that never played any earlier iteration and let them play a game without you. They'll be able to play without any assumption and to play the rules as written, not the rules you think you have written.

It's also good to hand the book to a fresh group and sit in on a session (not answering any questions or trying to guide the group in any way, just observing and taking notes), if you have the time and a cooperative group, because a lot of things might not make it to a playtest after-action report.

Maybe they run into an issue but don't note it down and forget about it by the time they're writing up their report, maybe the GM fills in a gap in the rules that they thought was left to GM discretion but you intended to work some other way, maybe what the GM uses as a miniboss encounter was something you intended to be easier but turned out difficult due to something unforeseen, maybe the session goes swimmingly but you noticed some things you want to change that the group wouldn't have considered a problem and reported back to you.

Omegonthesane wrote:The version of "wargame broken by what should be a stupid tactic" I remember was Doug Lenat testing his machine learning system Eurisko on Traveller's Trillion Credit Squadron tournament in the 80s, and winning the whole thing by just bringing a flotilla of defenceless, nigh immobile gun platforms. Which was laughed at by experienced players until they saw the resulting win streak and he walked away with the trophy.

And then did it again next year when they introduced a "fleet mobility" rule to stop his pet robot's strategy, which Eurisko worked around by just scuttling any damaged ship and still won. So they threatened to cancel any tournament that had a Eurisko fleet in it, giving Lenat a title of Grand Admiral in consolation.

(Some details there taken from the Traveller wiki. Because of course that's a thing.)

Thanks for sharing the story. I looked into this and found some really fascinating articles going into more detail on both the details of how he won TCS and more about Eurisko and its descendants. Links for the curious:

Traveller Wiki
"Eurisko, The Computer With A Mind Of Its Own"
"How David Beats Goliath", New Yorker
"One Genius' Lonely Crusade to Teach a Computer Common Sense", Wired

Not a question I want answered, if that has even been the purpose of this thread for years, but this:

JE Sawyer (Fallout New Vegas, Pillars of Eternity etc) does Q&A on his tumblr, mostly on the topic of game design. In a recent post he was talking about the challenges (in both CRPG and TTRPG) presented by not having any NPCs be Essential.

In doing so he referred to what we might call "murderhoboing" as "reckless shootin' and hollerin' " and that is what I am going to call it from now on.

3.5 D&D question: if a dragon and a fiend have a kid together, is the child a half-fiend dragon or a half-dragon fiend?

The rules provide a clear-cut answer for metallic dragons (the half-fiend template cannot be applied to a Good creature), but applying either template to the other creature seems game-mechanically valid for the chromatic dragons.

rampaging-poet wrote:3.5 D&D question: if a dragon and a fiend have a kid together, is the child a half-fiend dragon or a half-dragon fiend?

Flip a coin.

I would say that since the half-fiend template is more restrictive and also

d20 SRD wrote: Half-dragon creatures are always more formidable than others of their kind that do not have dragon blood

then these critters would be half-fiend dragons, since the fiendish template doesn't seem as dominant.

The half-fiend template being a strict alignment restriction is both fucking dumb and has no bearing on whether metallic dragons can have half-fiend children. They just can't have half-fiend children who are Good instead of Neutral or Evil at the moment of their statting.

I was also under the impression that D&D redefined the "Always" in "Always Chaotic Evil" to mean like 99% probability, but don't care enough to look it up after a quick Google failed.

Both require a base creature. You can apply both templates to a creature that exists.

If a fiend and a dragon have a child, there is nothing to say whether it is a Nalfeshee or a Dragon or a goblin.

To avoid being ridiculous, it would be easiest to take the dragon as the 'base creature' and just apply the half-fiend template. You could take a dragon and add the half-dragon template, too, but most of it doesn't really help (but you would get +4 Natural Armor compared to a full-blooded dragon).

There are a lot of oddities with templates like 'half-celestial' is straight up better than 'celestial'.

Are we the grognards now?