The Gaming Den

It seems to me that much of the failing of D&D3e was in PC's not having the capability of handling many level appropriate encounters.

Much of the failings of D&D4e was in PC's having to grind round after round against level appropriate threats that had little power to actually hurt the PC's in turn.



As for fixing 3e's problems (or exposing them), the Same Game Challenge is a good measure. That revolves around pitting PC's/Parties against level appropriate encounters to gauge of useful they are to the party. It seems that The Den's Tomes and Frank's Races of War were built around creating characters who can overcome a Same Game Challenge.


But with that said, it seems that the majority of "this is my RPG project, check it" always begin with PC classes, with monsters coming up last, if they even do appear. So with that being said, wouldn't it be good practice to first get down what an Orc does, what a Giant Scorpion can do, and the toughness of a Dragon, then create PC's which can overcome such challenges together?

Pretty much you begin with the Same Game Challenge tiered by level, then you create the PC's to overcome them.

Well first you need an action resolution system, then you need challenges, and then you need PCs. I'd say it's roughly that order. Sections of the PCs may be part of your writeup for action resolution (resource management, skills, action declaration, etc.), so there are definitely parts of the PC end that you can be productively working on before you get into the monsters. And many of the monster abilities are going to be PC abilities as well, which means you can get a two-for-one there.

But yeah, I think the constant consideration about whether a Barbarian should have +3 attack or +4 in the absence of minotaurs for them to be attacking is rather pointless and leads to poor decisions. This sort of methodology is what leads us to 20 level Monk classes that give all kinds of weird abilities every level but never actually get the ability to contribute meaningfully in a single level appropriate challenge at any level at all.

-Username17

Would hashing out the design spaces for PCs and opponents fall under designing the action resolution system? It might be a common sense fallacy, but it seems equally prudent to design what the PCs can do or what you want them to do, and then design challenges around that. You get the opponent/PC ability two for one and hash out the action resolution parts. To avoid mathematical dickery, don't hash out specific numbers until the action resolution system is complete (things like to hit and damage math are included in action resolution, right?).

Mask_De_H wrote:Would hashing out the design spaces for PCs and opponents fall under designing the action resolution system? It might be a common sense fallacy, but it seems equally prudent to design what the PCs can do or what you want them to do, and then design challenges around that. You get the opponent/PC ability two for one and hash out the action resolution parts. To avoid mathematical dickery, don't hash out specific numbers until the action resolution system is complete (things like to hit and damage math are included in action resolution, right?).

I don't really think it's helpful to start from the actions characters do and work to the challenges the characters are faced with. Once you've committed yourself to Angel Summoner and BMX Bandit, no challenges you can devise will provide real challenge to both.

Not until you decide what level the PCs are going to be confronted by ghosts can you say with certainty what level the PCs are going to need ghost busting attacks. Not until you determine what level the PCs are going to be confronted by flying archers can you say with certainty what level "guy who punches things" completely ceases to be a valid character concept.

-Username17

The best ideas seem obvious in retrospect.

And this is one of them.

FrankTrollman wrote:

Mask_De_H wrote:Would hashing out the design spaces for PCs and opponents fall under designing the action resolution system? It might be a common sense fallacy, but it seems equally prudent to design what the PCs can do or what you want them to do, and then design challenges around that. You get the opponent/PC ability two for one and hash out the action resolution parts. To avoid mathematical dickery, don't hash out specific numbers until the action resolution system is complete (things like to hit and damage math are included in action resolution, right?).

I don't really think it's helpful to start from the actions characters do and work to the challenges the characters are faced with. Once you've committed yourself to Angel Summoner and BMX Bandit, no challenges you can devise will provide real challenge to both.

Not until you decide what level the PCs are going to be confronted by ghosts can you say with certainty what level the PCs are going to need ghost busting attacks. Not until you determine what level the PCs are going to be confronted by flying archers can you say with certainty what level "guy who punches things" completely ceases to be a valid character concept.
-Username17

Makes sense, but I would posit that this would be done in coming up with design goals before creating challenges or even the resolution system. They influence what action resolution system you want to use, and color the perspective you see your resource allocation minigame(s). Design goals would provide a framework to prevent making a game where Angel Summoner and BMX Bandit are assumed to be on the same page.

Although in order to do that, you...need to know the challenges they are going to face, or at least have a framework that will inform you of what challenges they'd face.

Lago PARANOIA wrote:The best ideas seem obvious in retrospect.

And this is one of them.

Diggin' your new avatar, Lago.

Back on topic, I do believe that more D&D-clones need to make just as much of a focus on monsters and enemy NPCs as they do on PC options. Monsters and challenges are a huge part of the game, and PC classes should be designed with these things in mind.

Reminds me of Lamentations of the Flame Princess. It got some flak for not including a collection of monsters (unlike other retroclones). There were building guidelines, but it couldn't be GMed right out of the box.

It's chicken-and-egg problem, once it comes down to the actual numbers.

That is, if 1st level PCs should be in the neighborhood of +5 on "stuff", therefore if you want 1st level PCs to hit this monster 50% of the time, it should be have an AC of 15. Turning *that* around - that is, writing up a bunch of 1st level monsters, seeing that you've given them an average AC of 15, and concluding from that the 1st level PC should be rolling ~+5s?

So my intent is thusly:
* Figure out approximately what +s PCs of a given level are getting.
* Make some monsters that drop in the correct number of swings.
* Balance the classes against those.

DrPraetor wrote:It's chicken-and-egg problem, once it comes down to the actual numbers.

The actual numbers are the easy part though. For all Mearls' whining about how scaling attack bonus, armor class, hit points, damage output, and damage reduction across thirty levels is "too hard", it's basically just repeated division and multiplication. Given enough time and paper you could hash out any sort of design goal numbers with some nerdy fourth graders. Sure, the work would go faster if you explained logs to them, but it could still be done. Expected numbers of battle turns are ratios, hit rates are ratios, hit points to final expected damage are ratios. Multiply that shit together and get the numbers you supposedly want.

Of far more interest is the GTFO abilities. Because ultimately it is those, and not the lack of scaling numbers that ensure that Fighters can't have nice things. When you look at the challenges you have done up and notice that half of them are effectively immune to swords, you know that your 4e-style Fighter will never be good. And while you could claim that challenges and character abilities are also a chicken-egg problem, they really aren't. You tool up agents with the skills and equipment they need to accomplish the mission. You need to decide on what the missions are going to be before you decide on what training and equipment your agents are going to need.

So it's actually pretty simple. Challenges first. Then decide from the challenges what the PCs are going to need to be able to do. So basically the Tome plan, but you'd hopefully make the monsters tighter and less lolrandom in their abilities while you were at it. This has the advantage that you're not going to ever commit yourself to Angel Summoner and BMX Bandit, because at any level where the monsters are "dire sheep", angel summoner would have to get scaled back, and the moment "manticore" becomes a thing you're going to have to give BMX Bandit something to do about flying enemies.

-Username17

When building my heartbreaker I set a few benchmarks every pc should be able to do. By level 4 you need to be able to deal with flying archers. By level 7 you need to be able to deal with insta-death abilities. That sort of thing. (It's 10 level long, with unofficial implied support for infinate). Then when writing my class outline I plugged placeholders into each class at those levels, +\- 1 for various classes so I could ensure I remembered to include them. Ended up with super diablo leaping barbarians and angel winged paladins. I also hashed out the earliest various abilities could come online, even if everyone doesn't get them.

Then I hashed out what monster groups and progressions should look like without writing specific details. Next I wrote out specifics of classes and lastly specifics of monsters. ( need to revise monsters because I half assed it)

Point of post - is that a decent method or am I doomed from the start?

I've known for a while now that I should be doing it this way, but I'm not really sure how. Most of my ideas start with who I want to be, not what I want to fight, you know? Actually a lot of the awesomeness montages in my head are scenes of people showing off in practice and shit, not fighting in life and death battles at all. How do you recommend re-orienting away from "thinking about what it would be cool to be able to do" toward thinking about challenges?

Lago PARANOIA wrote:The best ideas seem obvious in retrospect.

And this is one of them.

Thanks Luigi, the salute is appropriate.

Orion wrote:IHow do you recommend re-orienting away from "thinking about what it would be cool to be able to do" toward thinking about challenges?

Krusk wrote: Point of post - is that a decent method or am I doomed from the start?

Yeah, having the theory is nice but putting it in practice is the next step.


Figuring out how strong your giant scorpion should be (how giant IS it?)
Figuring out how magical your unicorns are
Figuring out how many orcs is equivalent to fighting an ogre and so on.

Then apply your conflict resolution system to them with different number settings to get an idea of how they fit.

Think of yourself as the BBEG on an Awesome Budget: My greatest lieutenants will be doing X, while my regional governors will be doing X-1, while their lieutenants will be doing X-2, and then the captains under them are at X-3, etc., either because 1) that's all I can afford, or 2) all evil challenges want to kill the ones on top of them and take their spot, so I order them this way to create stability in my organization.

Instead of imagining what characters will be doing, imagine yourself as a potential BBEG plotting world/planar/universal domination.

Krusk wrote:When building my heartbreaker I set a few benchmarks every pc should be able to do. By level 4 you need to be able to deal with flying archers. By level 7 you need to be able to deal with insta-death abilities. That sort of thing. (It's 10 level long, with unofficial implied support for infinate). Then when writing my class outline I plugged placeholders into each class at those levels, +\- 1 for various classes so I could ensure I remembered to include them. Ended up with super diablo leaping barbarians and angel winged paladins. I also hashed out the earliest various abilities could come online, even if everyone doesn't get them.

Then I hashed out what monster groups and progressions should look like without writing specific details. Next I wrote out specifics of classes and lastly specifics of monsters. ( need to revise monsters because I half assed it)

Point of post - is that a decent method or am I doomed from the start?

Well, as long as you're willing to go back and change things iteratively, you could eventually finish like that. But here's the problem: imagine that you decided to set some basic limits (flying archers by level 5, incorporeal enemies at level 6, instant death powers at level 7). And you make characters that can deal with those things. But now you make up a Same Game Test for the first ten levels of simple combat encounters:

10KF SGT wrote:

• Battles of Level 1
• 20 Giant Spiders/Giant Rats/Snakes in a pit
• 5 Orc or Elf Warriors
• 2 Gnolls
• 10 Goblin Thugs
• 10 Zombies/Skeletons
• 5 Pixies
• 2 Worgs
• 1 Ogre
• 2 Fire Imps
• 5 Enfields

• Battles of Level 2
• A single Amphitere on the wing
• 20 giant bats
• 10 claw demons
• Swarm of bees.
• One huge Scorpion
• 10 hungry Ghouls
• 20 Vampire Thralls
• 2 Dryads
• 5 Lizardfolk soldiers
• One Giant Crab

• Battles of Level 3
• A Gargoyle comes to life.
• 2 Bears.
• 2 Alphyns
• One Ergentyne
• Five Mamuna
• A Werewolf
• 10 Tengu bandits
• 20 Brownies
• 5 Wererats
• 20 Plague Zombies

• Battles of Level 4
• A Minotaur in a confined space.
• 5 Wights
• Five Hippogryphs in flight
• Two Cockatrices
• Five Calopus
• A Vampire
• A plague of Locusts
• 20 Fang Demons
• 2 Lamias
• A Mandragora

• Battles of Level 5
• 2 Cave Bears
• 5 Harpy Archers
• 20 Red Caps
• 5 Mummies
• 2 Gamelyons
• A Cerberus Hound
• A Hydra
• 5 Griffins in flight
• A Manticore in flight
• A Troll.

• Battles of Level 6
• One Wraith in a confined space
• A horde of shadows (20) come to life and attack.
• Two Basilisks
• 5 Hill Giants
• A Hellwasp Swarm
• 5 Fu Dogs
• 2 Land Sharks
• 2 Succubi
• 5 Byakhee
• A Wyvern

• Battles of Level 7
• A spectral wizard.
• 2 Gorgons
• 5 Medusa Archers
• 5 Stone Giants
• 2 Galla
• A Golem in a confined space
• 20 Will-o-Wisps
• 2 Rakshasa
• 2 Nymphs
• One Tatzlwurm

• Battles of Level 8
• One Chimera
• One Beholder
• 10 Opinicus
• 5 Nightmares
• 5 Storm Demons
• One Naga in a confined space.
• One Vampire Lord
• 2 Erinyes
• 2 Kirin
• 20 Troglodytes in the tunnels

• Battles of Level 9
• 10 Salamanders
• One Mummy Pharaoh in a confined space.
• A Phoenix
• 2 Cthonians in the tunnels
• 2 Mind Flayers
• 1 Dragon Turtle
• 2 Asura
• 1 Death Cloud
• 5 Frost Giants
• 1 Geryon

• Battles of Level 10
• A proper Dragon
• 5 Fire Giants
• One Balrog-like Demon
• One Lich
• 20 Chaos Beasts
• 10 Serpent Fiends
• 2 Catoblepas
• 10 Elder Things
• One Kraken
• 5 Nightmare Beasts

OK, so we note that things work out that after level 6 or so the PCs can just go ahead and be flying archers themselves, because encounters that can be beaten automatically by levitation kiting are gone from the second half of the list. That's good. But we also notice that we have a bunch of other breakpoints:

• At level 3, monsters appear which require special weaponry to kill (like, but not not exclusively the Werewolf). We also have diseases (Werewolves again, but also plague zombies).
• While actual "death" isn't being handed out until the Gorgon's death breath at level 7, we're still dealing with petrification at level 6 (basilisk), paralysis at level 4 (cockatrice), and various charms and dominates in between. While not "dead", the character is still at the very least removed from battle by these effects.
• The fucking swarm of bees is in there at level 2. Sure, it's a not-terribly impressive real-world threat and probably belongs at first or second level, but it's essentially immune to be being killed with a sword, meaning that warrior types have to be able to do something meaningful with fire and smoke at level two.

So noticing that sort of thing, you're probably going to have to go back and adjust things. And that's just the combats. There's two other important considerations: challenges and missions. A challenge is something like: what if there's a locked door or a magical glyph or a river of lava or a damaged bridge between you and the goal? The mission itself is whatever the goal actually is. And these are much harder to assign levels to and much more important to do so.

Let's take the mission of "go to the bowels of the Dungeon of Doom and retrieve the sword of Clan MacGuffin". Sounds reasonable enough, right? I mean, you could put something like that at pretty much any level, depending on what the Dungeon and the Sword do. But in D&D, that "mission" expires at ninth level. Literally at ninth level the Wizard knows scry and teleport, and the party can accomplish the entire mission in downtime without really dealing with any part of the Dungeon of Doom except maybe the traps or guardians that are literally in the room with the sword in it at the very end. It's not much of a "mission" at that point, it's really more of a challenge. So if you've put anything even remotely like that at level 9 or 10, the whole Scry & Teleport setup has to be jettisoned or moved up.

Or let's consider the mission "go to the bottom of the sea and stop the Sahuagin from sacrificing princess Plot Device to the hungry maw and unleashing the Kraken". If you intend to actually release the Kraken at some point, that mission presumably goes to level 10, but if you don't, or the players can reasonably expect to call on sea elf allies to fight the Kraken with them, then that mission could be placed at any level. Thing is: you have to be able to actually go to the bottom of the sea and be able to effectively fight there (meaning that you have to be able to defeat enemies after your bows and fire have been effectively removed) at whatever level it is placed. And that's going to require a much bigger set of tweaks to character capabilities than the individual battles.

-Username17

The problem in 3e comes down to them building the monsters at the same time as they built the characters. The Fighters and Wizards just have the same powers at the same places they had them in AD&D (only the Wizards suddenly have the same XP chart and even more spells, and the Fighters don't ever get Weapon Specialisation, let alone Grand Mastery, and longbows suddenly don't work). They monsters just got the same bonus tracks as the PCs.

The testing was for each resulting monster group's EL vs Tordek, Mialee, Jozan, and Lidda on railroad builds; with them still using standard 2nd edition tactics of "Fireball the Giants", "protect the Fighter at all costs", "Polymorph is a utility spell", and "no one uses Haste". Giants have funny stats because they were Huge until the last minute, because 2nd ed. Hill Giants are 16' tall. It's AD&D in drag.

Hell, CR itself was a last-minute change, they were really testing EL directly for each group size of monsters, so a Mind-Flayer might be Solitary EL 8, Pair EL 10, Inquisition EL 12, and Cult (EL 13), only individually playtested and adjusted for each grouping size. That's why the organisation line isn't (1-5), but instead (1, 2, 3-5).


Where was I? Right, so 3e CR in part doesn't work because they weren't even testing CR, nor how real characters would turn out after the optimisation boards got going. Everything was just converted over and assigned an EL vs the iconic 4, then that was converted to CR as the late feedback changes all piled on and ... it works surprisingly well considering.


4e? Characters get +1/2 level (class) and +1/5 level (item) and +1/7 level (stat). Monsters get +level. Powers are supposed to always get the missing +1/6 level to attack and defence, by dint of non-stacking buffs and de-buffs, but then the monsters can do all that too. Obviously that won't work, you don't even need to test it, but that's what they published. They seemed to do all the in-house playtesting on what rate of power use "felt right" or something, and active DM fudging to go by the skill challenges. So just like they're doing for Next. No wonder Monte left.

Did I mention Solos? The basic math of 4x HP and +2 defences means you're going to run out of encounter powers and be beating away on a massive slab (that's long run out of it's own interesting powers and can't do much harm now) with the most boring part of your repertoire. It can't be otherwise. And all the highest level stuff is Solos and Elites, so, I have no idea how that got through.

Like, it's all deeply formulaic, there's no adjusting anything because that was never part of the plan. Maybe they broke Solos with a last minute +2 everywhere to stop players trashing them with saved-up daily power dumps, who knows. It seems like anything in the history of D&D that broke their combat formula just got cut, and they never even tested what was left. Just tried to make all the new powers and monsters into special snowflakes that hopefully reminded people of D&D, and bloated the white-space to fill out their empty-ass books.



tldr; Monsters first or PCs first doesn't even touch on it. The game they tested basically threw out all the results near the end, and the game they didn't test had no intention of changing anything anyway, because they already had their formulas.

<snip> Has all my stuff been double posting? Pfft.

There is a *mild* challenge in the math - which may or may not be more important if you allow any sort of multiclassing - in that you need to figure out the spread in defenses, in order to place an appropriate price for diversity.

At 5th level, Karg smash - Karg has a mighty sword, and we were paying attention to GTFO issues from Frank's list so he also has a "heartseeking zen arrow" which is very unpleasant if you decide not to close with him. These both work against physical defense, let's say his bonus is +10 for the stuff he does.

At 5th level, Lychee Himechan zaps. She has some zaps that do elemental damage, and she has some zaps which blind or confuse people instead, but either way they work against evasion defense, her bonus is maybe a little bigger (+11) but her defenses are generally not as good as Karg's.

At 5th level, Glowdeep Starfaggot both swings a glowing sword around *and* stuns people with his dreamy eyes. His glowing sword does some extra magic damage but still works against physical defense, but his smoldering look works against will defense. How big should his bonuses be? It depends on how chunky the monsters are.

In D&D 3rd this is a huge but highly-variable advantage - the Troll (CR 5) is 21/14/13, while the Achaierai (also CR 5) is 17/16/17. Against things like the Troll, your average TN would be 16 regardless, but the average "pick easier of 2" is only 13, worth an average of +3 (which is a lot). Against the Achiaerai he gets no bonus at all.

Of course if some monsters are really chunky in this way, it might as well be a GTFO ability. But unless you make some simplifying assumption like "every character attacks vs the same defense no matter what they do", it's going to be important to get this settled before writing up the characters, as well as the more conventional GTFOs.

Yeah, one of the things I understand least about 4e is how it mandates that every player character have one or two glaring weaknesses in their "NADs" while monsters don't. If the fact that monster of the week #3 has a weak will defense makes him die easy, that's no big deal. But if a player character dies easily because of a weak Fort defense, the entire game is over. It's one of the many design decisions of 4e that I don't agree with or understand the logic behind.

In other news, just as you need to write out not only your monster fights of each supported level, you also really do need to fill in your missions and challenges for each level. You can start with the template:

10KF SGT wrote:

• Challenges of Level
• Your path is blocked by a
• The trap that is vexing you is a
• You want their help, but they want a
• The information you want is known by a
• The target is obscured by a
• The treasure is behind a
• The door is sealed by a
• To undue the curse you need a
• The clue is in a
• You're trying to track a

• Missions of Level
• Destroy the
• Find the
• Explore the
• Rescue the
• Slay the
• Defeat the
• Solve the mystery of the
• Secure the aid of the
• Defend yourselves from the
• Travel to the

And fill that out for each level. This is the part where a lot of proposed characters are revealed to be basically BMX Bandit to be honest. I mean, if your challenge involves the door being sealed by a time distortion or the way being blocked by the vastness of space, what the fucking fuck is a "swordsman" supposed to do?

-Username17

Youmu Konpaku would say "cut through the time dilation" and "cut through space", but common sense (and BMX Bandit) isn't really a thing in Touhou.

Mask_De_H wrote:Youmu Konpaku would say "cut through the time dilation" and "cut through space", but common sense (and BMX Bandit) isn't really a thing in Touhou.

You can't let yourself be held back by common sense in Gensokyo!

Mask_De_H wrote:Youmu Konpaku would say "cut through the time dilation" and "cut through space", but common sense (and BMX Bandit) isn't really a thing in Touhou.

She wouldn't even need to draw her blade, because she runs fast enough to cause time dilation herself. The Gardener of Hakugyokurou class would probably have a handful of weird time manipulation movement powers, a permanent minion in the form of a Ghost Half, and a bunch of Giant Frog sword abilities that work better the more you use them in a fight.

Someone needs to write an Elothar for this.

Often theres fetch quest for the swordsman to get the time cutting sword, or to destroy the shield generator on endor so the deathstar can be attacked.

This is extremely similar to Test-Driven development in programming. So, yeah this line of thinking seems solid.