Please see this IMPORTANT MESSAGE from The Angry DM!

The D&D Boss Fight (Part 2)

April 29, 2010

Three years ago, my party slaughtered its first dragon. Thus began my long love-hate relationship with solo monsters in 4th Edition. Recently, however, the folks at Radio Free Hommlet took a long, hard look at solo monsters and helped identify some of the real problems in solo monster design. I took a look at some of these flaws in the first article in this series. Now, I want to try and solve them with the help of a militant atheist Spartan with anger issues.

God of War III is the first action game I’ve played through since the glory days of the 16-bit processors. It was also the first game I played in the God of War series. Thus, the game was a very new experience for me, having mainly focused on RPGs and MMORPGs over the last few years.

God of War III begins with Kratos, a very angry gentleman with some sort of skin condition, leading an assault on Mount Olympus. In that respect, it doesn’t mess around or show any restraint. “Here is Mount Olympus,” it says. “You are very angry at it. Now, go kill it.” Literally, five minutes later, Kratos finds himself fighting a giant hippocampus, a sort of aquatic crab-horse-snake thing. Of course, this comes from the same mythology as the chimera, so I shouldn’t be surprised.

The boss fights have always been the centerpiece of the God of War series from what I can tell. They are huge and over the top, yes, but there is a great deal of strategy involved. The fights require you to play defensively and understand enemy attack patterns. Your most powerful attacks are rarely useful because they are easily interrupted or countered, so you have to content yourself with launching short volleys of quicker, weaker attacks while countering enemy moves.

All of this makes for a long fight. You need to take time at the beginning to test the enemy and see its patterns, you use weaker attacks, and you spend half the time concentrating on your own defense. But I didn’t realize how long these battles were until I looked at the clock. The fights never felt overly long or boring, and it never felt like I was grinding away uselessly.

The reason for this is that boss battles in God of War are staged. After you beat up on a creature for a little while, something happens to change the battle. For example, I was fighting the massive hippocampus while riding on the arm of the giant titan Gaia. After I had made some progress, the hippocampus dug its claws into Gaia’s wrist and she recoiled in pain. Now, the hippocampus and I were hanging upside down from her arm. My options for attacks changed a little, and so did the hippocampus’. Shortly thereafter, Gaia righted herself and we were back on firm ground for a final battle. The hippocampus was furious and thrashing and its attack pattern was faster and more chaotic with some slightly different attacks thrown into the mix. The whole thing seemed carefully scripted to give me just enough time to settle into a pattern and pat myself on the back for figuring out the fight before it changed. And each change was just enough of a change to make me need a new pattern without making me feel like I had started a new battle.

What is the point of this overly long introduction and discussion of a video game? Well, the more attentive of you have probably noticed that God of War III had inadvertently solved three of the problems in solo boss fights in 4th Edition: static fights, the lumping of all the coolness at the beginning of the fight, and the lack of sense of progress. And that is my starting point.

The Boss Fight

Before I go too far, I want to say that I don’t see rewriting every solo in the game. Sometimes, a solo monster can be a fun fight by itself and some solo monsters are really well put together. In particular, solo monsters of slightly low levels work really well when paired with one or two buddies. Instead, I am putting forth a new monster designation: the boss monster. They aren’t any more powerful than solos, really, but they are meant to be the culmination or climax of something big. A lich advisor to a king is an elite or solo monster. But the lich king of the Necropolis of Azard Kree who commands the undead army, he is a boss monster.

Design Goals

Before I go too far, I also want to explain exactly what I am trying to accomplish here. I am not just trying to fix solo monsters. Of course, that is one of my goals. But I am also trying to create a good narrative framework that gives structure to battles with creatures that are supposed to be epic. I am not as concerned with increasing the threat to the party so long as the fight feels fun and exciting. And I am not as concerned with shortening solo fights because, again, if it never starts to feel boring, it’s okay if it takes an hour to get through the encounter.

My most important goal is to accomplish all of this without uprooting too much of the established rules. By my nature, I avoid complexity and adding any new subsystem to the game increases complexity. In the end, I want the stat block for a boss monster to fit the existing framework of the game. That means accomplishing most of my goals with the addition of new powers to the stat block.

The Boss Fight: Basic Framework

The basic framework for a boss fight is a fight in three acts. Each boss monster consists of three stat blocks. When the party has “killed” one stat block, something happens and the monster is replaced by the next stat block. This isn’t necessarily a transformation of one monster into another (but it can be), simply a change in stance, tactics, and powers. Of course, this does create other possibilities which I will discuss in brief below.

The entire encounter together is worth the same experience as a solo monster and the total hit points of all of the stat blocks is equal to the standard hit points for a solo monster. The individual components do not have a ‘bloodied’ state. Rather, the third state of the monster is always ‘bloodied.’ This allows PCs to use abilities that key off of being bloodied, though they do come up a little later in the fight.

The shift between one state and the next is the most important part. At the moment that the component stat block is ‘killed’, it will use an action of some kind to signal its change. Through this action, it may or may not damage the party or inflict a condition, but it will always disengage from the party. After that, it becomes the new monster and all conditions currently affecting it are removed as if a new fight started.

I’ve thought long and hard about this. The player’s advocate side of my brain balked at this idea because it weakens the effectiveness of ‘until end of encounter’ powers. But in the end, the fact is I have reduced them to one third as effective instead of one fifth as effective, which is what they are in a standard encounter with five monsters. The player’s advocate side of my brain also mentioned the idea of overdamage: what if the player uses a 3[W] daily and most of the damage is lost because the battle shifts to a new state. Again, the analytical side pointed out that, in a standard encounter, overdamage is always a possibility and is possible five times more often.

The real key though, is this: DMs need to identify a creature as a boss and let the party know what that means. If the party knows that every two or three rounds the fight is going to reset, they will manage their own resources more evenly. Just like they manage dailies and action points throughout the day, they will now manage them throughout the boss fight. Some players will hold back until later stages. This will help reduce the grinding feeling because the party is now spreading its resources and should have interesting things to do in the later stages of the battle.

Forcing a disengagement also works well to prevent the fight from becoming too static. It gives the solo a chance to pick more favorable terrain and get out from under the dog pile, however briefly. With the deck stacked against solos so much, they need the terrain and the breathing room.

At the end of the stage change, each party member is allowed to recharge one encounter power and spend one healing surge for free. The fight has taken a momentary breather. Again, this reduces grind and allows the party to absorb a little extra damage as they adjust to the monster’s new tactics and new attacks.

Finally, at the end of each stage change, the solo monster rolls initiative again and it has a new place in the initiative order. This mixes up the action just a little.

In play, each stage of a boss monster will have a power that looks something like this (taken from a red dragon boss):

Fury of the Inferno (free (special); when reduced to 0 hit points; encounter)
Close burst 2; +12 vs. Reflex; 1d12+4 fire damage and the target is pushed 2 squares; miss: the target is pushed 2 squares; effect: the dragon moves its speed and cannot end this movement adjacent to any enemy, each enemy in the battle regains the use of one encounter power of their choice and can spend a healing surge. After this action is resolved, replace the red dragon with the furious red dragon as if it were a new creature entering the fight. The furious red dragon rolls initiative normally. Special: the dragon can take this action even if a condition or effect would normally prevent it from doing so.

The Individual Stat Blocks

Each individual component stat block actually doesn’t have to look that much different from a standard solo monster. This is a design feature. The art of solo design still applies pretty much as written in the DMG and DMG2. Individual component stat blocks don’t need as many encounter powers because each one only lasts for one third of the encounter. And the powers don’t need to recharge because the creature will get a new encounter power at each new stage. This works well because it essentially matches a ‘5, 6’ recharge on average.

Action points are worthy of note here. Each individual component stat block has two action points and this goes a long way toward addressing the problem of having too few actions. Coupled with the standard array of solo powers (immediate actions, minor action attacks, and auras), four extra actions in the combat helps a great deal.

The Boss Fight: The Possibilities

Now, we have a solo monster that comes in three parts. Each part is essentially one third of a monster. At this point, a lot of new possibilities are opened up. For example, the middle stage of the boss fight could be replaced by a complexity 2 or 3 skill challenge to chase down the fleeing monster and corner it. Success would allow a more favorable position. Failure gives the creature time to set an ambush. Or, the fleeing monster could call upon underlings to hold off the party while it retreats to its private chambers. The XP of that encounter would be one third of the solo monsters original budget. The party would not earn experience points or reach milestones for these individual stages as they are all part of the same encounter. It should be clear that if the party tries to take a rest, the monster will reengage them.

Even the baseline boss monster in three stages has some interesting possibilities, each of which can be reflected in its statistics. Some defenses could go up very slightly while others go down. Its tactics could change slightly. A clever DM, writing for his own players, might use the different stages to give different characters a chance to shine. So, whereas defenders can easily lock down stage one, stage two becomes more mobile and the ranged strikers and controllers get to open up.

The thing to remember is that the stages should tell a story. The red dragon, for instance, begins as a fairly standard soldier monster. After all, it doesn’t consider the party a very serious threat. But when they push it into stage two (Furious Red Dragon), it unleashes its full elemental power, perhaps gaining an aura that makes it harder to engage (aura 2 that does slight damage and creates difficult terrain). In stage 3, the party has clearly beaten it down. It loses the elemental fury and begins fighting like a cornered, desperate monster.

When the party confronts the white dragon on the surface of an icy lake, things are a little different. In stage 2, the creature launches itself downward, through the ice, bursts up somewhere else and retreats deeper into its lair. The party must contend with a skill challenge to clamber across the shattered lake or get out of the water without taking too much cold damage. If they take too long, the dragon has set an ambush behind a collapsible ice wall. In this case, stage 2 is replaced entirely by the skill challenge. When the party catches up to the dragon, it is in its stage 3 (bloodied) state.

And the lich-king? He begins as an artillery monster with some control abilities. When the party begins to gain the upper hand, he launches a wave of necrotic energy and summons three skeletons to harass the party while he encases himself in a sarcophagus of necrotic energy to recover. He only breaks free when his skeletons are dead. Although bloodied, he is more powerful and dangerous with the infusion of necrotic energy pulled from the Shadowfell.

Why Three Acts?

I chose three stages rather than any other number for one reason: this is the standard structure for dramatic/narrative tension. In the first act, the party engages the foe and starts to gain the upper hand. A reversal occurs and the party must deal with a new and more dangerous situation in the second act. The party turns things around and they push forward to victory in the third act. It also works out that each stage will last about two rounds, possibly three. A fresh party has enough dailies and encounters to fill that easily. A tired party has enough encounter powers to fill most of that time with a daily or a couple of at-wills interspersed.

The Action Denial Problem

The staged fight and the additional action points have done a great deal to solve most of the weaknesses of solo monsters and to add some exciting new possibilities. The staged structure also gives the solo a chance to shake off all negative conditions twice during the fight. But the action denial problem still looms large over the boss fight. The +5 to saves does not mitigate the situation nearly enough and, as parties learn the structure of boss fights, they will learn to game the system.

There are a number of ways to solve this problem, some of which can be quite unfair to the players. One could simply make boss monsters immune to certain conditions, as many video games have done, but that, to me, is taking it too far. The party needs a chance, albeit a slim one, to actually have some of these conditions stick.

Thus, on each boss monster, I add the following two powers:

Boss Monster Stability
When an attack would knock the boss monster prone, the boss monster makes an immediate saving throw to avoid falling prone with its normal +5 bonus.

Boss Monster Resilience
At the start of each of the boss monsters turns, if the boss monster is dazed, immobilized, or stunned, it may choose one of these conditions and make a saving throw with its normal +5 bonus. If the saving throw is successful, the condition ends immediately and the boss monster can act normally on its turn, even if the condition does not normally end on a save.

These two powers leave a 25% chance that a given instance of prone and a given action denial condition will affect the boss monster on its turn. This works out well given that normally, such conditions can only be inflicted on 20% of a standard encounter group. More to the point, limiting the resilience to one condition allows the party to stack them up if they want them to stick, but this reduces their ability to inflict such conditions in future rounds by expending limited powers.

In Conclusion

The staged boss monster framework along with a few new powers  has given us a very nice framework with which to work. In the next part, we will look at some of the specifics of designing boss monster fights and I will provide two heroic-tier examples for you to try out in your home game. But I must caution everyone that this is still very much a work in progress. I am only in the early play testing stages, but I will post results as they arise. I also invite any feedback and encourage everyone to come up with their own spin on these ideas.

Tags: , , , ,