In my Running Dragons blog, I briefly mentioned the danger of an imbalance in the “action economy” during a boss encounter. A lot players coming into D&D might have their expectations for boss fights shaped by video games, where one extremely strong enemy takes on the party despite a numerical disadvantage. For your bosses in D&D, this is suicide.
Even the game’s own mechanics for making this more available fall short. “Legendary actions” (special moves the boss monsters may take at the end of other creature’s turns) only go so far: a lone dragon with only the legendary actions listed in the official statblock will be destroyed by an appropriate level party without a problem. Hell, the two dragon encounters I’ve run in my current game both featured clusters of minions, and they were a tier above the players – at level 8, they fought an adult black dragon (a CR 14 monster), and just a few weeks ago, at level 14, they battled an ancient black dragon (a CR 21 monster). The party slayed both dragons with only one casualty between the two encounters (during the latter, and easily reversed with a Revivify).
Part of that is because of some of the shortcomings that exist in the Challenge Rating (CR) system that we’ll get into in a moment. My table also consists of a lot of bona fide gamers that work hard to conserve resources and adequately prepare. They knew well ahead of time what kind of dragon lair they were walking into, and both times obtained some tools to mitigate the damage it could deal with its breath attack. All this to say that my advice here will not be good advice for every table: it could be that where I see failure in challenge ratings, they’ll be perfectly workable for your table. A crew of careful, calculating players, however, might want for a bit more difficulty when they roll for initiative.
Here’s some tips to give them that.
Reexamining Challenge Rating
The first things to take to heart when trying to make a more challenging encounter for D&D is to take CR less seriously. It can be very useful for determining whether a creature is an accessible foe for your players at their level, but it doesn’t mean it can hold its own without allies. In the Dungeon Master’s Guide example of this, they mention a Rakshasa (CR 13) being something that might prove more difficult than you think against a party with its limited magic immunity trait, but with its low pool of hit points, any well-balanced group will annihilate one of these fiends no problem. I think the Rakshasa works best as a late tier 2 antagonist, for a party of 8th to 11th level, despite the spellcasters at this level having no ability whatsoever to damage the creature without some kind of physical weapon.
And! With minions included in an encounter with a Rakshasa, your spellcaster players will have a valuable task to undertake once they know of its magic immunity while the more martial characters deal with the fiend himself.
Perhaps the worst problem the challenge rating values suffer is how hard they crash the second your players obtain combat-oriented magic items. That fighter finding a +1 shield? Your barbarian getting a +1 greataxe? Your wizard finding a wand of the war mage? Banded accuracy is thrown out of a window.
Your fighter is suddenly much, much more difficult for CR-appropriate creatures to strike. Your barbarian and wizard are both now landing their attacks much more often than the game was balanced around. Consider this: the scaling component included in the player characters’ power, their proficiency bonus, scales one point every four levels. An item granting them a +1 bonus accelerates them that much further ahead.
Now, I’m not advocating for holding these items back from the party. Magic items are a lot of fun to have – especially homebrew items that are on the cusp of breaking the game. It’s just another factor that contributes to CR faltering in the mid-to-late stages of the game.
Ultimately, I think bosses can comfortably sit a tier above a party playing strategically in combat and provide a healthy challenge. There’s a few cases that can cause that to fall short – spells that will pancake the player characters or attacks that deal an amount of damage they won’t be able to play around, but I think it’s easier to adjust those outliers than to try and scale a numerically-appropriate monster up to boss viability. I suppose that brings up another question …
How Powerful Should the Minions Be?
I generally include a creature or two very close to my boss’s listed CR as their lieutenants. In my ancient dragon encounter, I included a Blue Abishai (a CR 17 creature) as the primary lieutenant, with a homebrew dragonborn fighter opponent that I set at CR 11. The dragon and the abishai both began the battle away from the room’s entrance, hidden in the darkness of the cave, so the fighter and a few guard drakes (CR 2 creatures) were present to oppose the party while the big monsters got into position.
The party ended up successfully locking the fighter lieutenant down with a Banishment spell, removing them from the battle until well after everything else had been handled. The abishai used its Greater Invisibility to fight unseen, but the players managed to break his concentration and used a Stunning Strike to keep him grounded long enough for his elimination. Even the additional bonuses I gave my dragon – an antimagic darkness zone lair action, an ability to use an action and legendary actions to heal if it was at its hoard, an immunity to movement speed reduction when using its Wing Attack legendary action – couldn’t make this monster powerful enough to battle the party alone.
Create an array of additional enemies for your boss with a variety in their challenge ratings. Give them some chaff, weak monsters that can be eliminated with well-placed area-of-effect abilities. Give them an ally that’s dangerous, but wouldn’t be a problem without the heavier hitters in the room aiding them. And give them a powerful lieutenant almost as dangerous as the boss themselves to force a division of the party’s attention.
Building Complex Encounters
Another option for adding difficulty to an encounter is to ensure the goal isn’t just reduce all the enemies’ HP pools to zero. If every battle in the game runs that way, it can get stale regardless of the challenge you’re building. One of my most successful encounters in this campaign was earlier on (I believe they were around level 6, it was well over a year ago now). They had infiltrated the compound of an extremist group of zealots bent on using an ancient magic ritual to call radiant fury down on a village the party sought to defend. In that battle, I included a cleric NPC enemy as the boss, and calculated her to be about CR 9, gave her a martial ally based on an adjusted Champion statblock (down to what I thought to be CR 7), and a handful of CR 2 swordsmen and several priests. These priests, however, were first-and-foremost working on the ritual. The party needed to split their focus on interrupting the magic and defeating the dangerous enemies in the battle. Chapter 3 of the Dungeon Master’s Guide has several other ideas for diverse encounter goals.
The trick to these kinds of encounters if to ensure there’s a clearly defined win condition for the players – but not necessarily one that you’re responsible to prescribe. Your players can deduce a lot about your encounters on the scant amount of information each dice roll will tell them. They can calculate where the enemy’s AC might lie, they’ll know when a foe is putting out more damage than they can sustain for multiple rounds, and they can react accordingly.
That said, one of the hardest things for newer players to learn is when they should run from an encounter. There are some situations that truly become untenable, and unlike a video game where everything is usually balanced around you being able to overcome it with the tools you have to hand, some battles in D&D might just not be feasible for you to win. I was in a game once where at level 4 we came across an ancient white dragon (CR 20). I had the most experience of all the players at the table, including the DM (who had rolled the encounter from his module’s table) and I immediately knew we needed to split. If it weren’t for a successful saving throw and a ring that conferred resistance to cold damage, one of our party members would’ve been killed outright by the breath weapon. (But we all made it out with the clever use of an illusion and Rope Trick.)
A Never-Ending Education
By the very structure of its rules, D&D is a more combat-oriented tabletop RPG than other contemporary systems. Encounters are something you as a DM will spend a lot of time cobbling together. All of your dungeons, your factions, and your wildernesses will be expected to have their own unique array of enemies to overcome. And designing these battles is a process that will never run out of things to teach you, and not just because every table is different.
At my table, it would be blasphemous were I to on-the-fly adjust an encounter I designed and make it easier. That doesn’t mean that design stops when initiative is rolled for every party. If you had enemies in reserve, but the players are getting thwomped just fine already, maybe they don’t need to show up. Maybe that high level spell slot lingering up their foe’s sleeve got used on something before this battle. My advice in these instances is to keep any of these adjustments under wraps, and whatever you do, don’t begin making tactical blunders that the enemies don’t have a justification to make. Nothing has killed a mood at my tables quicker than them being able to tell when a battle got easier.
Don’t be afraid to try new things with your encounters. If it doesn’t work, that’s okay. There will always be another roll.
As always, thanks for reading. Good luck out there, heroes.





