Latest MADcorp documentation: https://docs.google.com/open?id=0B9k6jY73glEKNzIzMmQ1ODMtYmI1Zi00OWIwLWFjZWEtZDVkZTlkYTVjYzVk
(the "traschan edition" ones are the newer ones.)

Ok, so, one of the last things I'm not happy with in MADcorp is the monster rules. You can review the current ones in the core rulebook PDF in the "how to be the Ref" chapter. Like, I'm not happy with any of it.

Let's review my philosophy for monsters:
1. Monsters are fucked up.
2. Monsters are rare and for serious.
3. Monsters don't come in species. There's no such thing as a hydra. There might be such a thing as a horrible being with multiple regenerating heads named Hydra.
4. Fighting a monster isn't about trading shots with it until it falls down. In fact, 90% of the time that should get people killed. Figuring out how to get the odds on your side should be necessary; knowing is half the battle. But like all such brain challenges, it should either be a "problem" with an open-ended solution, or a "puzzle" with a more-or-less set and preplanned solution. If it's the latter, it should either be not-required (as in, players should have the option to leave this monster the fuck alone) or there should be a way to discover hints elsewhere in the situation, if you're paying attention.

These current rules don't really arrive at this. They really, really don't. The effects-first thing isn't working. I think I need to start "fiction-first," in the sense of thinking up what this monster is and does in purely fictional terms first, then figuring out the mechanical effects. It should end up looking more like a PC than the mess of rules that it looks like now -- it should be defined in terms of a few numerical rules (bonuses/penalties, frex), but mostly in terms of exceptions and unique functions, like how all the Employee Handbooks are done. So... how do you explain to the Ref how to do this? There needs to be a way to do it from scratch, and supplementing that with rolling charts is high on the list of priorities, in case you have to make up a monster in a hurry.

There are some things it's important to preserve. Classifications are important -- aberration, mechanical, undead, and phantom are all mechanically relevant to various PC classes. But the way these classifications are defined right now is ass.

I got nothing. Any thoughts?

Make a list of monsterous abilities/featuers, according to your various classifications, and let the GM frankenstien those abilities together any way he wants.

For your puzzle ways to defeat monsters, make sure you include defensive mechanisms that make them highly resistant to some forms of attack.  Also, make some features offer some relatively devasting or powerful moves, but come with a weakness as a package.  Like a video game - the hugely armoured thingy is basically impervious to all forms of damage except when it uses its power attack which momentarilly flares the armoured scales around its neck, exposing it to potential damage for a short time.

or something.

That's pretty much how it is right now. In practice, it doesn't work the way I want it to. I pretty much want to throw it out entirely and come up with something different. It's not even fun to make a monster, and it should be.

Maybe the list of abilities is just made up *of the wrong abilities*. I don't know; it's possible.

Hey Marshall!

I've played the game, so I know my suggestion might lead to a re-tinkering of the rest of the game, as well (which might be just as well for all I know, as when we played, we spent most of the time in withdrawn-from-fiction mechanical combat mode, which is fine in itself but not exactly what you want with all this, if I got it right).

So, how about something of this form...

Werehound of the Kennels - heals all wounds within minutes, except those made by silver.
When the monster attacks you, flip a coin to see if you die. If you don't die, flip it again to see if you're horribly maimed or mutilated.
Driven by bloodlust and instinct.
Faster than a car.
Shreds and tears through anything weaker than concrete.
Human by daylight.

It's mostly just a fictional description of what it does, but it pretty much tells you why you need to fear it and what you could do to escape with your life and maybe kill it. The first two bits are the most important, I think. All of the monsters should have a "unkillable: because! except for..." clause there somewhere. Being unkillable makes them scary (in conjunction with the general rule of them being able to kill everything). The "because" allows the Ref to make judgement calls when the players get creative. Frex, if we dump the Werehound into a pool of acid, it's done, similarly how an otherwise-immortal incorporeal phantom can be dealt with a ghostbustery suck gadget. The exception is there mostly as a mental failsafe - if there's a "it can be killed by X" on the sheet, the Ref can rest easy that the monster is 'theoretically defeatable', and can go plum crazy with the rest of the sheet.

It can also have motivations that give it direction. "It's a monster and it's here to kill you" only gives you options to kill it or run away, but if it wants something other than killing that *makes* it want to kill you, you have room for discovery and improvisation. A ghost of a crazed killer that wants to torture living people and a weredog who just really likes to eat raw flesh have different priorities, and will act differently given different stimulus... Meaning, if players can't beat a monster, they can at least learn how to manipulate it.

Random "this monster is cool and can do this" things are easy, any person who wants to Ref can probably sprout off twelve of these for any monster he thinks up in a manner of seconds.

Various "I win" buttons than are dependant on the setting the monster is in are also a possibility. "Human by daylight" seems to make the monster weaksauce, yeah? But I'm the Ref and I say - in this dungeon, there's no sunlight. Ever. So to press the "I win" button, the players get a logistical exercise of getting light into a dungeon, exploding its roof, stuff like that. Either way, fun.

So basically I think it all boils down to giving it a reasoned physical description (what does Hitler-brain kill-bot really *want*? How does Hitler-brain kill-bot do the things he does? What's he made from and how does *that* work?).

I think that if you give the monster stats, it'll be dealt with by out-statting it, and a player is far better motivated to out-stat a monster than a Ref is to do the converse. Go descriptive, and they'll be forced to deal with the monster in the fiction.

Marshall wrote:
That's pretty much how it is right now. In practice, it doesn't work the way I want it to. I pretty much want to throw it out entirely and come up with something different. It's not even fun to make a monster, and it should be.

Maybe the list of abilities is just made up *of the wrong abilities*. I don't know; it's possible.

Daumantas,
I'm really liking that approach. I like the idea that the monsters are so Other that most rules don't apply to them. (I probably wouldn't fuck with the way they deal damage, though. The When Animals Attack table is nasty enough, let alone the other special damage tables.)

Regarding divorced-from-fiction tactical combat, I think that's just a timeframe thing. Fighting without trying to game the situation isn't  viable over a longer timeframe, because it's too risky; any deadly weapon has a 5% chance of inflicting a fatal blow on an unmodified damage roll, and while you can Gamble it (in the sense used in the Step on Up essay) and hope that you roll that 5% before the other guy, sooner or later you're going to catch one between the eyes. That's the idea with all the lethality and maiming in the rules: to require (eventually) that the players start making sure that things are slanted firmly in their favor before engaging in violent conflict, whenever possible. Otherwise they're just gonna keep losing dudes, and probably get frustrated with all the random, meaningless deaths.

Stefoid (Steffen? Stephen? I can't remember),
I don't know that I'd consider the game crunchy. The idea is to be concrete, so that reference & judgment with regard to the fictional situation can be made clearly and in clear terms. There are very, very few things in the rules that produce an effect without first requiring reference to the situation. The skills and stuff that the character classes get are all about what situations they thrive in, and what they can do to re-contextualize problems.

Except the monster rules as they stand: those are crunchy as fuck and all ate up with numbers and balances and extra hit points. I hate it. Especially the extra hit point thing: if the system as-is requires you to put extra stats on a monster to make it threatening, then something's wrong. That's exactly the kind of boring-ass weak-sauce escalation of challenge that I've railed against for years. As Daumantas points out, it's much better to simply say that you don't hurt it at all unless you do it the right way.
(By the way, the way to handle that in play is the player describes how his dude attacks it and the Ref says sorry, that doesn't work, and the player says oh fuck. With no rolls involved, because that's the rule for resolving impossible tasks.)

Hi Marshall,

I've achieved exactly what you're looking for with demons in Delve.  I'd like to say I have a perfect system to offer you, but my success has been based partly on how I GM, and the work to translate that into rules is still ongoing.  So I'll just touch on a few bases and see what you think.

1.  Have a secret logic behind all monsters.  Provide an appropriately alien set of motives, methods, and perspectives that define all monsters.  Example: demons seek to destroy organization and hijack communication, they do this with some intelligent planning but also compulsive taking advantage of every opportunity, and they see the world in a way that views physical matter as the enemy and anything chaotic or magical as a friend.

2. Each individual monster's capabilities provide a specific way of doing the thing all monsters do.  Example: this demon eats memories and regurgitates them into inanimate objects.

3. Randomly generate one way to beat the monster.  Pick a tarot card or something.  Example: there's a cage on this card... so the situation includes a cage designed to hold this monster.  So the players need to learn of, locate, and figure out how to use the cage, and then chase the monster into it.  Hey, that could be today's adventure!

4. Take the specific weakness implied by the previous step and generalize.  Example: if this monster can be trapped in the special cage, perhaps it can be trapped in other ways as well.  So perhaps this creature is impossible to damage, but it not infinitely strong, so can be pushed around.  And maybe it's not too bright, so can be lured into places.

5. Take your understanding of all monsters and this monster into the game, prepared to ad-lib based on it.

6. Here's the part that is specific to Delve.  I doubt this is an option for MADcorp, but I offer it just to show how I make the previous monster stuff work in play, in hopes that it might give you some ideas for how to integrate it with MADcorp.

Remain loyal to the pre-established stuff, and refuse to solve anything for the players; but, within those constraints, try to help them solve the puzzle of your monster.  Feed them clues, hints, physical evidence, info from previous monster-fighters, etc. until they come up with an approach to the monster that isn't suicide and that you can use to give them even more useful info.  And if they come up with a plan which, based on your knowledge of all monsters and this monster, seems like it would work, then it does.

This is extremely fun and easy for me to do as GM, but I designed it, so no shocker there.  As for supporting others to do it, I continue to test out combos of rules and presentations.  Can't really offer you anything concrete yet, alas.

If I remember right, the "evil plot" approach you were using for Hex Rangers (fka Witch Trails) successfully produced quality puzzle-monsters.  Maybe you could distill that stuff into something tighter, where the monster's story is played out in its lair or on its body, rather than throughout the whole town?

Perhaps you pitch to the group to describe what their characters have heard of, of the monster (which is to say, to make up these things, but doing so through a characters perspective). All the PC's describe what they've heard of it. You note all of this, then modify it somewhat, determining what is truth in the description, what is false, what wasn't mentioned at all (ie, stuff you make up and add on). Obviously this is a huge departure from preperation before play.

What are 'monsters', anyway? Just there to make something to do? Does the monster get a preplanned solution on how to beat (and beat presumably means kill) the PC's? How long term is gameplay - a real conflict is long term gameplay Vs genuine capacity for PCs (even all of them) to die. Often the latter gets axed to enable the former - but when the PC's can't die but the monster can...it almost makes the PC's the monster and so monster creation loses its savor as it's more like creating a lamb for the slaughter.

Gameplay is long term. In MADcorp, long-term victory conditions are fixed, but in the as-yet untitled sandboxy world-gone-weird game (built on MADcorp's system), players set their own victory conditions (retire to my own private island, become king of Kansas, whatever). PCs get cacked all the time, generally in random, gruesome, and meaningless ways, if the players aren't careful. But you play multiple characters, generally with one as your 'main' whose goals you actually focus on.

Players choose what risks to face and which ones to avoid; learning to choose well in this regard is a central part of the challenge. Therefore none of the challenges are necessarily 'fair' or 'balanced.' If you think the Sears Dragon will eat all of your dudes, then stay the fuck away from the Sears Tower.

For this type of player determinism to work, the choices have to be real. The options have to be independent of what the players want, and magician's choices and palette swaps are cheating on the part of the Ref. Which is one reason why random generation procedures are also needed.

David, I'm really liking that approach. Thinking of monsters as a mini-situation all by themselves (or not-mini, depending on the monster) is a good approach. You guessed right about #6; clues have to be handled differently in this game. Either you have to experiment, or discover something on your own, or you get lots of clues that are conflicting and have to suss out which are true.

I love experiments!  I think the most important thing there is feedback.  Whatever you do, whether it works or not, should tell you something.  Another key is the ability to be wrong but live to try again.

Videogame boss fights are good models here.  You have to avoid Bowser until he comes out of his shell and then jump up into his belly when he's in midair to hurt him.  You have to kill the WoW demon tree's beehives so they stop regenerating any damage you do to the tree, even though the hives take forever and are the least threatening to the characters of all the enemy targets in the scene.

Ye average videogame lets you fail, learn, and try again via multiple lives.  Ye average tabletop RPG needs another method, such as multiple escapes.  "Get near death, run away, heal" has been the interlude in most of my learning-and-experimenting monster battles.  Despite multiple characters, I get the impression that the experiment-and-learn process in MADcorp is probably a better fit for multiple tries per character, right?

It can be tough to trot out some intimidating monstrosity, let the players try something stupid/ineffectual, and respond by not killing them.  I try to get around this by hitting them somewhere that hurts other than death.  I do better with baffling creepy mind-sucking metal-eaters than big strong deadly beasts.

One trick I use to help inform the players to perform useful experiments is to piggyback on their existing knowledge.  If they learn a little bit of my demon metaphysics, the next demon they fight will probably be designed with their knowledge as its weak point.  Educated guesses = fun, shots in the dark = suck.

Pardon the flood of generalizations.  Once you whittle down your approach a little more, maybe I can contribute something more specific.  Are there any particular technical/mechanical/system constraints here?  Anything you're aiming for a monster encounter to do in those terms?

Yeah, feedback, definitely. "It doesn't work" isn't enough information. Let's say the monster is the Nemian Lion, with its impenetrable hide. We don't call it the Nemian Lion, of course, but as Ref I know that's what it really is. So, when somebody attacks it with an edged weapon, we skip the HIT roll and I go straight to describing how the blade just glances away from its hide with a clang (and probably have them roll for Wear & Tear). If the players know their mythology, then they'll figure out what the lion is, and remember that blunt force trauma can hurt it -- and if someone attacks with a blunt weapon, we'll roll to HIT and damage as normal. Other than that, there would be some clues somewhere -- tavern gossip, whatever. Barring even that, they can try other modes of attack. Some of these will require judgment calls on the Ref's part -- will dynamite hurt it? My call would be yes.

I gotta think about what mechanical constraints I have to keep in mind. I know for sure that I need to be able to classify monsters by type -- undead, mechanical, abomination, etc. -- because those terms matter for certain rules. I also need to be able to classify them by threat level, to have a way to determine how much THREAT it costs to bring a monster into play. And I have to be able to generate them randomly. I'll have to think of what else.

David wrote:
Ye average videogame lets you fail, learn, and try again via multiple lives.  Ye average tabletop RPG needs another method, such as multiple escapes.  "Get near death, run away, heal" has been the interlude in most of my learning-and-experimenting monster battles.  Despite multiple characters, I get the impression that the experiment-and-learn process in MADcorp is probably a better fit for multiple tries per character, right?

I am seriously interested in getting this sorted out, but I've got some ideas on other stuff first: I noticed your using one of my pet peeves; "margin of success". You can replace that with just reading the number on the dice as your success level and saying "sorry, that's too awesome for that guy to do" when they roll a 15 with a "hit 5" guy. Then busting over your score by 5 can always implicitly be being cocky in some way.

Downside of this? People will immediately realise that their guy is never going to be awesome straight off the bat. But more importantly this mucks up the way success interacts with the initiative system; at the mo you resolve actions from highest successful roll, meaning that crap actions by skilled people generally go first. I can see how this could be pretty awesome, although you could change it to doing them in ascending order "to finish with a bang"/"because maybe you were rushing it" and tiebreaking by stat size.

Anyway onto monsters:

It seems like the things your rules need to assure about monsters is that they are learnable; that they have signs to their weird behaviour. Secondly players have to be able to implement that knowledge to survive it. So you need to block off everything that puts them "out of reach" or unavoidable.

How to do this?

• Well maybe start off with a creature type, and a rough starting role for it (eg pit lurker/dead end of doom, slow juggernaught pursuer, gatekeeper, long term harasser, custom enforcer(=act the right way and it will ignore you), straggler muncher,  etc), the game dark souls is great for getting you into the mood for making these.
• Then you could have categories of questions to ask about it, that lead to suggesting powers, then you could compare those powers to a list of checks made from playtesting, to avoid making it too powerful etc.
• Then you could go from powers to signs of those powers, and build the environment and appearance of the creature from there. Also try and imagine it "in repose", which is the rule Ray Harryhausen used and Guillermo del Torro uses for making monsters seem more physically believable.
• Then you'd work backwards from any other distinctive features of the creature, removing red herrings by adjusting or adding powers to fit the appearance you've imagined.

That way you'd build creatures that are obvious, readable, that don't telegraph powers they don't have, and do show signs of the ones they do have. Hopefully that would help.. The roles would be there just to start you off making them something that players can orient too, even if they don't know precisely how to solve them. They could easily blur from the role over the course of development, but that could give you a starting point at least.

Your dungeon is not composed of encounter balance, but does have an intuitive sort of clock, that presumably players will get used to dealing with: They'll start to get a feel for it. So what you probably need for the threat side of things is a judgement system, so that after making a monster, a ref can reasonably accurately assign it the right threat rating, not for balance reasons, (because making it a coherent encounter is handled separately by the stuff above) but so that it fits the rhythm of the dungeon that your threat mechanic produces. Also I think judgements at the end is the only way you can produce a coherent picture; powers don't add arithmetically to each other, so having three powers of supposed threat 1 might lead to a very dangerous creature if they all work together, or to a not very dangerous one if they contradict each other. Actually producing threat values from specific components would be much heavier maths!

Perhaps you could look at what distinguishes the monster from the critters, what are the differences in how people are supposed to deal with them?

The bit then that players will need help with, but that I'm not sure how you can assure with the rules:
Is it focused/distinctive as a concept? Does it behave differently and unexpectedly? Is it unnerving or darkly funny?

Maybe just a really solid set of examples?

Hey Josh,
The 'margin' *is* just the number on the die, at least until hit by blocking. And that other thing has me confused. I think you're confused about how modifiers work? Maybe? They're not like d20, they're basically re-rolls.

Now, as for this stuff about monsters, this is gold. I gotta spend some time thinking it over, but I think I'm on the verge of a methodology.

Ah for some unknown reason I thought you were subtracting the stat from the roll and comparing it to a table, which does have an interesting interaction with "highest roll goes first". Which is what I thought you were doing! Disregard the first bit.

Marshall wrote: The standard operating procedure for MADcorp is to first observe if possible, then try something (ideally in a way that puts interns at risks instead of your main dudes), and if it doesn't work then make a choice between gettin' out while the gettin's good, or trying another possibility.

Marshall wrote: I know for sure that I need to be able to classify monsters by type -- undead, mechanical, abomination, etc. -- because those terms matter for certain rules. I also need to be able to classify them by threat level, to have a way to determine how much THREAT it costs to bring a monster into play. And I have to be able to generate them randomly. I'll have to think of what else.