This is about designing your own adventures and encounters with an admittedly confusing set of benchmarks that probably lead both novice and even experienced DMs to incorrect assumptions. A group of 5th level characters should never be pitted against a level 10 monster, right?
CR (challenge rating for those of you who are new, like me, to the wide world of modern RPGs) was something that I could NOT get my head around when running Pathfinder. I mean, I READ all about it; about how you stacked up all the baddies in your encounter and then applied maths. But it honestly didn't seem (in actual play over the 8 months I used it) to reliably foreshadow how the encounter would roll.
In concept, I liked the idea. But in practice, it felt like the value of CR was undermined by the fact that there are just too many variables in a game as complicated as D&D.
But hey, it's not like AD&D got it right either. "Monster Level" in AD&D was meant to more or less serve the same purpose but I would term it an unreliable filter. Relying on Monster Level to populate your dungeon is uh...not the way to go.
The disparity, or perhaps I should say the haphazard way, in which Monster Level was assigned in AD&D can be experienced viscerally, for example, when you cast Monster Summoning I and prepare to roll on the list of available monsters (DMG p222). Your stomach immediately wrenches and you want to fudge your roll for a Manes Demon (which has 3 attacks per round, spell immunities and requires +1 or better weapons to-hit) vs uh....yeah...that giant rat you just rolled.
But hey, we're not here to discuss what doesn't work----nor are we here to modify the Monster Summoning tables in the DMG. We're here for solutions to one of the toughest parts of being a DM (probably for ANY edition): how do you design encounters at the right level of challenge?
Design principles are something I'm familiar with, having worked with and tested systems now for nearly 20 years. This does not mean I have all the answers. It just means that I'm not totally making shit up.
Some folks would approach the design problem by trying to hammer out a new CR-type formula. Something like: Here's the formula, plug in monsters and treasure, voila!
Voila, is usually the ILLUSION of something done easily by a master who's done it a million times. And formulas are great for calculating the cubic feet your fireball is going to occupy in a hallway. But when it comes to making an encounter FEEL like a cohesive part of the adventure AND well-balanced besides...mmm, that's a tough nut to crack. In fact, you might spend your entire life testing, validating, updating and defending your formula instead of actually making adventures and playing D&D.
So then how do I do it? In lieu of brittle formulas, I use Design Principles.
Principle the First:
Odds of Hitting & Damage per Hit
AC is the first thing I look at. What is the AC of this monster? I then glance at my screen and see what a FIGHTER of n-level needs to hit that AC in melee combat. Next I look at the reverse. What does this monster need to hit the average AC of n-level FIGHTER in melee combat. Damage per Hit [or] Damage per Round (generally more applicable to monsters with a variety of attacks) is the third part of this principle.
If you look at fighter and cleric THAC0 and do the math you will quickly draw some conclusions. If you are making a 1st level dungeon, the Fighter has a base THAC0 of 20! That means every time he swings (which is only once per round) he has only a 20% chance of hitting the 3 hp goblin that you decided to put in plate armor...because you thought it would be challenging and fun.
(this, BTW, is another reason why "Adj to hit AC by weapon type" should be in your game and on your character sheets...but that's a different discussion).
Bottom line is, 20% sucks, and unless the fighter has poured all his starter money into armor, the goblin will probably have a better chance of dealing damage.
Even if the odds are the same for PC and goblin, 20% means your combat round might go directly to that horrible place we've all been, right? Roll for initiative. "I miss." "Uh, the monster missed too." Roll for initiative. "I miss again." "Same." Roll for initiative...
I've seen that go one for 10+ rounds. No lie. This is not a fault of the system. This is poor design and poor use of the system by the DM that designed the encounter. The designer has failed to understand and respect probability.
Probability is the first thing you MUST understand when designing ANYTHING for a game founded on dice and it cannot be summed up easily through HD (or any other singular part of the monster stat block).
Still, AC+HD+DMG output IS a good place to start.
Ideally, you want PCs (IMO) to have, on average, a 50% chance of hitting the monster you are throwing at them and sometimes more. If HD of the beast are inflated, its AC should generally be worse...so that every time the fighter swings she feels like she is making progress.
50% generally works for adequately sized parties because if two or three characters are swinging every round, one of them (at least) usually connects. A solo monster, by contrast, may get 3 attacks per round (which it needs vs an entire party) but its HP should be in the range where 2 to 4 solid hits kill it.
And these are the most basic points of encounter design. Monster damage is the threat the PCs "see". Their minds are generally so focused on the amount of damage the monster is doing that they underestimate their odds of winning and often perceive the encounter to be far more dangerous than it actually is.
You have to also reverse (but tweak) this line of reasoning for the monster. Monsters are cheaters because they are often facing superior numbers. Solo monsters should have at least a 60% chance of hitting the AC of a fighter in melee with it and the damage/effects it inflicts should be able to kill that fighter in 1 to 3 rounds (because cooperation is the keystone of AD&D: ergo if the party is not working together, the fighter needs to die). If monster odds of hitting are reduced, it should deal more damage...potentially killing or KO-ing a PC with one hit. AD&D combat generally only lasts 1-3 rounds. Big battles can stretch out at the mid-levels, but tend to shrink again at high level due to greater crowd control and auto-win conditions that proc.
I'm going to ballpark for you here.
Decent Single Monster Encounter:
Party Odds to hit Monster: 50%
Party Hits to kill Monster: 2-4
Monster Odds to hit Party: 60%
Monster Rounds to kill PC: 1-3
Monster Hits to KO PC: 1-2
At face value, the outcome looks uncertain but the list does not account for players making crazy plans. And that is the real player advantage. You want monsters to destroy both the reckless AND the unimaginative.
Well, maybe, Anthony...but anyway, how do I even know what AC the monster is going to be trying to hit when I make my adventure? Fighter AC can vary wildly. It's a good question. AC 0 is quickly achievable and a solid benchmark even for low levels. Players are going to buy plate as soon as they can afford it, and a shield to go with. Fighters are as likely to boost DEX as they are CON, so if you assume even a +1 from DEX, you are already at AC 1 for a rich 1st level fighter.
Assuming a fantasy setting, your fighters will, through gear, likely achieve AC -3 even at low level and at medium to high level, AC will be in the -5 to -8 range.
Anthony, I'm appalled. NEGATIVE EIGHT? You are a sham and your campaign must certainly be a travesty of monty haul madness!
I don't know what I'm being compared to, so uh...I suppose sure! It must be! But to your point, when I attack your fighter (who has -8 AC) with 4 Aerial Servants, each with a THAC0 of 7, it means each of them has a 30% chance of hitting. Every hit is going to do 8-32 dmg (but PROBABLY 18-22 [because math]). And your 9th level fighter has 72 hp thanks to CON.
So: if you get unlucky, in any way, I'm going to wreck your -8 AC fighter and I don't need any special abilities or defenses to do it.
Thankfully, your fighter is not alone, and through maneuvering and tactics, I think you have a great chance at success.
Now, regardless of whether a fighter in your campaign ever achieves an AC of -8, the example still serves as a good benchmark for probability. When damage output is very high, the odds-of-hitting-slider can be moved down. Likewise, if the Monster only gets ONE swing per round (vs the party's 8+ attacks) a single hit from the monster should likely KO or possibly even kill outright anyone it hits.
Think of it this way: when you proclaim that you only need a 15 to hit the fighter and the fighter knows the four Aerial Servants can end him in as little as 3 hits, he's going to be on the edge of his seat. Nay, he's going to be terrified.
Terror is the goal of combat because it is foundational to drama.
What is likely to happen, of course, is that One of the servants is obliterated by Magic-Users, One of the servants is ganged by two fighters (one under Strength spell, the other wearing Gauntlets of Ogre Power). Meanwhile the Cleic cannot be touched by the summoned creatures thanks to Protection from Evil and moves to restore Fighter 1's HP.
Then again, it could go horribly wrong.
The probabilities of hitting and the average damage of each hit are only part of what goes into designing an encounter, but they are a great place to start and they are a great way to start scaling encounter difficulty too.
Principle the Second:
Understand the Arena
Published modules from TSR back in the early 80's, especially Gygaxian ones, are by and large, very well "tuned" and you can read them and run them to note how they play out and how they benchmark. Typical Gygaxian encounters deal a tremendous amount of dmg but also have only decent odds of landing hits. Gygaxian encounters very often use the environment to the monster's advantage in some clever way, through use of secret doors, supporting traps, or bonuses to surprise etc.
Gygax also often punishes those who charge in first and stop to think later. He presents the players a description where a charge is invited and then places an illusion-covered pit between the players and the monster (who appears unprotected and reliant on missile attacks).
When you design your encounter, account for the arena and allow it to bestow bonuses on your monsters unless the PCs take precautions. Likewise, don't be afraid bestow bonuses on the PCs if they are clever. Allow them to mop up quickly because, as I've said in other posts your dungeon should automatically balance out such quick victories through diversity of design.
More important, however are the tools in the party. Spells like Sleep, Hold Person, Heat Metal...are going to shift balance and you must have a sound understanding of the spells available to your group. If you allow Stoneskin, for example, modules published by TSR pre-Unearthed Arcana, might get steamrolled thanks to the entire party being immune to physical blows on round one.
You have to account for this. Introduce a trap that emits a cone of fire instead of a spear.
When you combine knowledge of the arena with the first principal (basic odds of hitting and dmg per round) you arrive at a place where encounter design becomes intuitive rather than formulaic and (hopefully) at a place where each encounter has room to breathe and adjust to PC tactics.
The BEST thing you can do after cobbling together your first encounter is to place tokens on a mat and roll dice. PLAY the encounter out with a quickly made party of NPC adventurers and see what happens. Then do it again. You may be surprised how quickly the PCs demolish what you assumed was too-potent an adversary. Or, as in the example of the plate mail clad goblin above, how what you thought would be a fun tweak, turns into a nightmare.
Bottom line is that there is no formula for making great encounters. Just design principles and a little bit of testing. To help you in the wilds of probability, you may find this incredibly useful.
And that's it. You don't need to buy a book on how to make adventures. It's just basic probability combined with diversity. That said, it's always nice to buy someone's book.
and happy gaming.