Running Combat Encounters

“I rolled a sixteen to hit, an eight for damage, and I don’t want to hear how he died.” “Is it my turn yet? I want to shoot another one in the eye!”

The most important thing to know about running combat encounters is that you shouldn’t do it as much as most GMs do. It’s understandable why GMs tend to guide encounters towards violence. The combat rules in almost any major RPG (certainly in every edition of D&D) are more fleshed out than any one other field of activity, and that makes it easy to run. It’s important not to try and nudge players towards specific approaches towards encounters, though. For the same reason GMs tend to fall back on them, players will fall back on combat regularly as well. They’ll also want not to do that sometimes, however, and you should let them explore alternative options to resolving an encounter besides stabbing it to death. Don’t think of any encounter as a “combat encounter.” Players get to decide what kind of encounter it is. Don’t fall into the trap of feeling like you’re doing something wrong if your players end up talking their way through most or all of the encounters.

Describing the Fracas

But okay, your players have decided they do want to stab this particular encounter to death. Running an effective combat encounter comes down to two things, description and tactics. The right level of description depends a lot on both the group and the context of the fight. As the DM of the Rings quote above illustrates, some players like to hear how awesome they were when killing an orc and others just want to get another mini off the battle map. Generally speaking, it’s good to describe the first few times a certain monster type attacks, dies, or uses a special ability. The first time the party fights a goblin shaman, describe the goblin shaman reaching into his components bag for a sliver of never-melting ice and alchemical salt, slapping them together between his hands, and then when he parts them again a cone of cold shoots out to cover the battlefield. If the goblin casts another cone of cold in the same fight, shorten it down to something more like “the goblin claps his materials into a cone of cold.” If goblin shamans are common or long-lived enough to get three or four castings of the spell out, it’s probably best to just say he’s casting cone of cold again. Likewise with plain old goblins making regular attacks or dying off.

It can’t be stressed enough how important it is to know your group, though. If your descriptions of monster attacks are feeding into player descriptions of their own attacks or dodges, it might help keep the energy high much longer. It can also drag out the fight and kill the pacing.

Tactics and TPKs

Tactically, it’s important to make full use of your monsters’ abilities to the extent of their temperament. While it is reasonable that a dim creature might do things like charge out of a chokepoint to fight the party out in the open even when outnumbered, or that a cautious creature might conserve its special abilities until their situation is desperate out of fear that there might be another battle after this one is over, in general monsters want to win, know how to use their own abilities, and it is out of character for them to do anything else but what is most likely to defeat the PCs (with the exception of suicide tactics). Making good use of abilities makes enemies seem more distinct from one another, it puts pressure on the party to come up with specific counter-strategies to different enemy abilities and rewards their thoughtfulness if they succeed, and it helps maintain verisimilitude. Both challenge and fantasy benefit from monsters fighting smart, at least so long as they’re in an environment they’re familiar with.

Of course, smart monsters are more likely to TPK the party, and this leads to a temptation amongst some GMs to have monsters act like idiots, quietly lose half their hit points, or decide to flee just as they’ve seized the upper hand. Completely cleaning out the cast of main characters to replace them with new ones is not generally good for a story (even Game of Thrones rarely kills more than one major character at a time, and is usually illustrating a point about irrelevance of morality to military or political success when it does so), so GMs are rightly hesitant to do it, but trying to prevent it last minute is a sloppy solution. We’ll talk more about how to defeat your party without killing them all in the Art of Adventures and the Art of Campaigns. For now, what’s important to keep in mind is that monsters must fight to the best of their ability to survive in order to maintain both challenge and immersion. Being defeated is disappointing when it happens, but it makes it obvious that the victories are genuine, not just given to them by fiat because the DM is their friend and refuses to ever let them lose.

On a similar note, it was mentioned earlier that PCs might try specific tactics to try and neutralize monster abilities. Some of these are within the bounds of the existing combat rules, like positioning themselves so that no more than two of them can be hit by an AoE attack. Others, like trying to fire ballistae with chains attached in order to hit a flying dragon and force it to the ground where the melee characters can hack at it, will have to be improvised. A similar temptation to letting the party avoid TPK by fiat is to allow any plan the party comes up with to succeed by fiat, or worse, evaluating whether you think it’s clever or foolish and then deciding its success or failure by fiat.

The key to success in D&D isn’t supposed to be guessing what the DM wants you to do, nor guessing what the DM thinks is cool. That’s metagaming, and you shouldn’t be punishing your players for avoiding metagaming. As much as possible, the key to success in D&D is supposed to be about interacting with the world as though it were an objective thing. That’s roleplaying, and in order to encourage it, the DM needs to be as objective as possible. That means not asking yourself “is this plan clever enough that it deserves to succeed” or “will the success of this plan be exciting?” What you should be asking is “what happens next?” The details for how to do that are in the Art of Rulings, but this is left here as a reminder that the short term excitement from a successful plan is not worth the long term ennui that comes as the players gradually realize that the game they’re playing isn’t about actually being clever, but rather about sounding clever enough to impress a capricious or condescending DM.

What about the opposite problem to TPKs? What happens when a boss, due to lucky rolls on the player side, ends up going down almost immediately? This can be anti-climactic, so there is often a temptation to give the boss some extra HP to drag the fight out. We strongly advise against this. It's the same problem as above, except making climactic encounters actually climactic, while a worthy goal, isn't even as important as keeping the protagonists of the story alive. The entire point of the dice is to avoid making combats deterministic, and giving the boss a pile of extra HP to cancel out high damage rolls or a string of crits nullifies the random element and makes the fights deterministic. Letting the players have their spectacular, lucky win is an option that should be considered for most fights, even if it's the boss of the adventure, but this can sometimes kill pacing. A villain who's received three months of build-up going down in one round can feel like a letdown even for the winning side.

For these climactic, arc-ending encounters, give the boss double HP and then have it lose 10% of its (newly increased) HP each round for the first five rounds, to a minimum of one. After five rounds, it's lost 50% of its HP total, which means you doubled its health, then cut it in half, so now it's back to its original starting health, minus whatever damage the PCs have done. A string of lucky crits that killed the boss in one round under the normal rules will instead shave off half its health, and it will still be losing 10% every round. The string of lucky crits did make a difference to the fight, and the boss monster is very likely to go down before round five, when the health depletion ends. In the more typical case when the party doesn't get a bunch of lucky crits, at round five (usually around the point when a boss encounter should be wrapping up), the boss will have lost all of the extra HP you gave it in addition to however much damage it took from the players so far.

This way, the boss is no more likely to ultimately kill the players (as it would be if you just doubled its HP without adding the degeneration effect), the players' still gain an advantage from their lucky crits or clever strategies reducing the bosses' HP faster than normal and will never loop around to doing so much damage that most/all of it gets ignored to let the boss stick around longer, but the odds of a one-round boss kill are still dramatically reduced.

Combat Encounter Variety

When designing opposition for combat, it’s important to avoid just opening up a portal to the Elemental Plane of Orcs. Enemy variety is one of the most critical things to keeping combat interesting, and that means not just describing enemies as being different from one another, but also giving them different combat abilities. Not just different ability scores, an enemy with 2 less HP but 1 more damage per attack is barely distinguishable, but combat abilities. A powerful attack on a recharge, an ability that does lots of damage but has a one round wind-up and can be interrupted if it’s damaged, an ability that inflicts a status effect, an ability that changes the positioning game (for example, flight, burrowing, or incorporeality), an ability that limits the players’ ability to deal damage like fire resistance or immunity forcing the Wizard to dig through the spellbook for something other than his standby fireball, and so on. Don’t just give these abilities to bosses, hand them out to mooks as well. Mooks who are fire resistant or who can fly and attack at range or who explode one round after they die make the entire adventure interesting and memorable, not just the final encounter.

An encounter almost always benefits from having at least two types of enemy who have some kind of synergy with one another. For example, a weak enemy with a powerful ranged attack on a recharge will cower behind bigger, tankier enemies while trying to recharge. A healer and buffer who keeps the front line healthy in the melee. Debuff casters who reduce the movement speed of party members, while ranged attackers kite them. Sometimes these encounters work best with one mini-boss supported by lots of weak enemies, sometimes as two different mooks mixed together, and sometimes just two or three powerful enemies, each with unique abilities. The “two or more types” isn’t a universal law, only a general rule, so don’t apply it to situations where it doesn’t make sense (many wilderness encounters) or where a lone monster really would be more appropriate (a powerful, legendary monster fought in its lair probably doesn’t need any help). Default to multiple enemy types in each encounter, but don’t feel straitjacketed by it.

Don’t tell the players directly what the enemies can do. Give them the same information that enemies will use to select targets. Some enemies have heavy armor and others don’t, and just like the monsters, players can probably guess that the lightly armored monsters are reasonably likely to be casters, especially if they don’t have any obvious weapons, and obviously once a monster has cast a fireball, the players can be pretty confident that the monster can do so again, and that other similar looking monsters likely have similar abilities. Clever monsters might intentionally mislead players by giving casters weapons they can’t use, just to make them look like archers, but generally speaking monsters are dressing to impress one another with the abilities they actually do have, not confuse enemies into thinking they can do things they can’t.