Thursday, February 5, 2026

Not Just Low Levels, All Levels Part 3: The Phases of Difficulty for Scaling Characters

In the sample campaign world that I haphazardly slapped together last time, we wanted many more adventure sites, more lairs, and some more supporting tent poles on a broader map. There's also a thoughtful placement of certain challenges, campaign world cohesiveness and interconnectivity to consider, but that's next time. While mixing in difficulty is the right and privilege of a Adventure Gaming DM, much of the bands of difficult should roughly align to the phases of long term campaign play. 

In short, a starting (phase one) party shouldn't have to travel across troll valley, red dragon ridge, or past a nest of Umberhulks and Mind Flayers to get to their destination of an orc hole. Just like in the dungeon, the top floor would be goblins and kobolds and displacer beasts and 7-headed hydras (WotC IP not withstanding) would be on the encounter table for level 5. 



Given a hex world of at least 34x56 (ideally more like 51x56) hexes of about 5-6 mile scale, you should be able to accommodate, through distance and depth, all of the bands of difficulty for all primary phases of play. For reference, that's 4 Judge Guild sized maps. 

The gist being expressed best in Keep on the Borderlands where the Caves of Chaos is the main hook, but trouble can be found with lizardmen or the hermit in the nearby area. Players will enjoy options of going for a distraction, even if just information gathering, rather than simply trying to make a 1.5 day trip to the chosen adventure. Some of the time, that might not be an option and getting lost in the woods can lead you to the Swamp Witches lair instead of the Bullywug infested Ziggurat.

Below, I am going to elaborate on the phases not only to illustrate what they look like in my game, but also to differentiate a long term Fantasy Adventure Game from the typical shorter form game (phase one to two).

What are the primary phases of combat difficulty?

Phase one - Mathematical difficulty of low levels


The game play of level 1-3 is at a place where, mathematically, it's the easiest to challenge a player. Resources like food, water, light, encumbrance, and ammunition are specifically important here, as well. Monster encounters almost always tax the group. For example, consider the below encounters as they are all mathematically draining from an attrition standpoint, even potentially deadly.

  • 6 Skeletons vs 5 level 1 characters 
  • 5 Orcs vs 5 level 1 characters
  • 10 Goblins vs 5 level 1 characters
  • 2 Giant Spiders vs 5 level 1 characters
  • 1 Owlbear vs 5 level 2 characters 
  • 3 Ogres vs 5 level 3 characters

In modern D&D, those costs get rested away. In Old School D&D (both CAG and OSR), resources matter and leaving the dungeon to recover is a real tough decision. Aside from the ability to successfully turn undead on the skeletons or use Sleep on the goblins, engaging in combat is mathematically likely to drain valuable resources. Those resources are key to survivability towards the next challenge. 

If the party is missing half of its fighters hit points and the sleep spell has been spent, then an encounter with an another 5 orcs, the encounter can go wrong in the way that coasting on fumes to get to one more gas station exit can. A few lucky hits and a fighter goes down, the other gets overwhelmed, people are trying to escape, and more people die. Even if they skate that encounter without real loss, a random encounter on the way out/back to town might just finish people off.

Make a dungeon with 24 rooms, make 10 of them on this challenge scale, toss treasure in 7 of them (some hidden), make 7 rooms virtually nothing but supporting flavor, and finally add like 5 traps and 5 locked doors. Boom. D&D dungeon. Mix in another axis of attrition such as carry weight and light limitations and you have a balancing act that the game performs for you. OSR modules are largely set in this phase and these are the reasons why they have such gravity to both write and the ease for DMs to run.

Hirelings are highly valued, a well placed charm/sleep spell can eliminate a room enabling a push that much further, and turn undead has a chance to do the same. But overall, most combat encounters drain the sand out of the top of the hourglass at alarming rate. Veterans know how to balance this while newer people tend to make deadly resource allocation errors.

Even the modules that are considered the best in the past decade or so follow this formula. Add in a dash of seasoning for flavor (i.e. giant goose dragon, a fresh take on orcs, a fairy tale vibe, an undead horde to end the world, etc) with some fundamentals of dungeon layout and you get enjoyment for the 3 sessions it takes to complete it. Given the level spread of about 1-5, you can comfortable get 20 something sessions into phase 2, more so if your sessions are short, players are slow, and rooms are basic. And for a lot of people that's perfect.

In terms of CAG style gaming, consider that a play group is generally going to only hit a few of these before their options go to more lucrative options. If your group beats Keep on the Borderlands (B2), The Moathouse (T1), and Saltmarsh (U1) then they aren't going to also want to do Against the Cult of the Reptile God (N1) but another group might want to. If you plan on making the world persistently played by multiple groups, offer enough options. If you start your game at level 3, you probably don't need many of these. 


Phase two - accelerating campaign momentum



Starts at about 5th level. Resources change. Torches are replaced by magical items and continual light spells. Food, Water, and Ammunition are still considerations but spells (create food and waterbags of holding, etc) and resources mitigate some the need for such laborious tracking. Item saving throws are used more as a higher amount of situations can make that potion, scroll, or magic arrows ignite into flame.

The dynamic of encounters is trivialized or made more potent by an action economy, monster abilities, and more powerful magic. 

  • 5 Ghouls and a Ghast vs 5 level 5 characters (either ambushed and shredded or turned and hand waved)
  • 8 Gnolls vs 5 level 5 characters (party positioning is likely to trivialize the encounter)
  • 4 Gargoyles vs 5 level 5 characters (12 attacks vs a party with 5-8 actions per turn)
  • 3 Wights or 2 wraiths vs 5 Level 5 characters (Risk of someone getting level drained, though not particularly difficult otherwise)
  • 2 Trolls vs 5 level 5 characters (action economy becomes more level, trolls deal significant damage and have 3 rolls each to do so, combined with a sturdy hp pool and regeneration, there will be attrition loss)

The bestiary of monsters start to widen as more of the Monster Manual(s) are able to be used. There are 350 monsters available in just the core AD&D Monster Manual and they aren't all for decoration. Relatively normal 'on-level' creatures (like a ghast or a troll) all pose a risk, surprise can be deadly, rewards become steeper tempting journeys to extend just one more room. With steeper rewards come the chance at more magic items that add another layer of complex decisions as well as increase power. 

Limited use items such as potions, scrolls, and wands ask the question if this encounter is worth it. Contextually good items become choices for setting up encounters (The crypt head is guarded by Caryatid Columns, so do you draw your Flametongue for the undead likely to be in the crypt or your Sword +2? Or neither because they could break? 

Encounters, such as those listed above, could be a handwave away. 5 ghouls and a Ghast surprise the party and paralyze two members or positioning of a Paladin. What if it was an ambush while the party was dealing with a patch of mold? What if the Cleric just turned all the ghouls round 1?

These encounters, for all their danger, hit that lower part of the U-shaped difficulty curve. The worlds factions become more center stage beyond vague familiarity. The group has been involved with many of them on some level at this point. This is the stride of a CAG D&D group and continues through the next two phases.

Phase three - full speed ahead



  • Lycanthrope (good action economy, difficult lycanthropy risk)
  • Hill Giant (high damage, high hit point pool, strong ranged attack)
  • Ice Toad (punishing melee range ability, knowledge check)
  • Hydra (great action economy, interesting foe)
  • Mummy (fear brings high risk, mummy rot is potentially costly)

Phase three is the back swing of the mid-levels, a continuation of phase two. It's where cohesion in play becomes more advanced between the players while the players become more fluid in playing to the options that their characters provide. This could be anything by using class skills to spells prepared, to which magic items are available to use.

More conditional effects, a wider exposure to the breadth of the Monster Manual(s), and a realization of how the action economy plays out on a longer timeline. Specifically, in the context of a sandbox campaign, this is where the freedom of travel, adventure options, and selection of tasks become more decision points. The world expands, travel expands further, and delves become deeper.

Phase four - the base of high level


  • Dragon (A deadly encounter in the open, a risk for subdual, or a carefully chosen battleground)
  • Vampire (Double level drain, a perpetually escaping intelligent enemy)
  • Vrocks (Innate abilities, magic resistance, potential gating of allies)
  • Umberhulks (Confusion and Brawn)
  • Drow Party/Other Mid-high level party (Steel against steel, bolt against arrow, spell against spell)

At the very top of mid level play, sliding into the real of higher level play, we have the infamous creatures that helped make the genre famous. 

Additionally, this phase largely introduces the concept of Magic Resistance, which can often take party members by surprise the first time. Dispel magic becomes doubly useful and more now than ever, the GM plays the enemies on a level deeper than tactically to make gameplay more challenging.

Take any dragon, vampire, or devil/demon and put them in a white room with 5-7 party members and watch whatever it is get shredded mathematically. A claw, claw, bite routine or a telekinesis isn't going to over come a slow spell and 10 attacks flurrying in the first round. No, the dragon would try to use its flight to an advantage, a demon would care to leverage a situation or incapacitate or gate in help rather than forcefully engage on it's own, and a vampire wouldn't just take gas-form to it's coffin 17 feet away. No, these are true villians, they plot, they scheme, and they prepare. It's another layer to the game to require the same of the adventurers to fairly get a leg up on a powerful opponent.

Lastly, these kind of monsters would be often be systemic in a bigger campaign organism. A lich, a dark god, or Orcus himself. By this point, these figures have been present indirectly the whole campaign and can provide the motivating force for the following high level play.

Phase five

  • Demons and Devils (Not only serving to be an in-world threat, with each tier being more powerful and influential than the last, but also the roster of a destination of high-level play.
  • Lich (The set up of the lair, the minions, the deadliness of fighting a powerful magic-user)
  • Mind Flayer (In groups, in lairs, attack on a different axis than a traditional party is used to)
  • More Dragons (More elaborate lairs, top end of the age category, with potential minions)
  • The horrors from the planes beyond existence

This is where High-level play starts to take on different forms.

High level adventure gaming is a broad array. It's not just inflating numbers to scale with higher level play, though that can be some of what makes up play. It's also big outer plane adventures, possibly in strange environments that affect movement, casting, magic items, or are even just plain dangerous without taking precautions. It's also moving through armies on a domain scale, leading to massive wars or even territorial skirmishes. It's high risk deadly adventures that leans less into the numbers of the characters, it's so many things in addition to domain/research. It's the tricky/trappy Tomb of Horror type dungeons.  It can border or cross over into PC->NPC progression. 

It's all of these things that the campaign has earned the right to by unwrapping each of these bands slowly to provide the options for high level play. The impressions over the past 15 years of the OSR seem to be like 'yeah man, do politics and domain management' or 'yeah let me just roll up 14 purple worms on the random encounter table' which is kind of like just eating the frosting off the whole cake.

While high level is varied, the world persists. Lower phase adventuring is still taking place in a world shaped by those who came before them. Have the cake, with the frosting, with a coke, and the ice cream on the side. High level diabetes can set in, though. Make sure to balance it with a healthy diet of low to mid range adventures.

In Part 4, I talk about a more cohesive world. Fleshing out more hex entries, interconnectivity, random encounter tables, and a functional world without the players.

Not Just Low Levels, All Levels Part 3: The Phases of Difficulty for Scaling Characters

In the sample campaign world that I haphazardly slapped together last time, we wanted many more adventure sites, more lairs, and some more s...