Wednesday, November 11, 2009

Running a Raid, Part II

Ok, so now that we have the baseline, let's see about jazzing things up a little. There are three big areas we want to address: we want to make sure there is some reason to use one attack instead of another, we want to capture some way to simulate aggro management, and we want to generally be able to jazz up the fight with special events, terrain and so on.

The first is the easiest, and we solve a bit of the third at the same time. The trick is that you can offload most of the fiddly bits onto the bad guy. The simple gimmick for this is to assign vulnerabilities to the bad guys that reflect tactical situations, and those can be paired with resistances. To give a concrete example, suppose the Boss can turn into a swarm of bugs - for the duration of this it might not take damage from any attacks except those that can hit more than one square. Or perhaps he can still be hit, but he'll take an extra point of damage from those attacks.

Provided that these vulnerabilities and resistance switch over the course of a fight, you successfully capture the idea that if you know what to do during each phase, you can fight more effectively. Now, admittedly the fact that in a WoW raid this knowledge is gained by dying repeatedly until you get it right, but that knowledge can probably be gotten some other way depending on the venue - for a convention game, for example, it might be part of the prelude.

There's a temptation to address the second point (aggro) with complex systems, causing certain actions to increase aggro, marking to move it around and so on, but that is unnecessary complication. If you step back, aggro is really just a mechanized way for the boss to "fight smart", seeking to go after the most cost effective targets - the squishy ones in particular. As a GM, you can already make that decision, targeting healers and more lightly armored PCs, so the only thing we really need is a way to allow the PCs to inhibit this.

So we'll compromise on a simple mechanic: any time an attack would mark the boss, the attacker gets a taunt token. Taunt token's can be spent to cancel a menace point which it used to fuel an attack that does not include the PC as a target. This does not cancel the attack, just the menace point - the boss can spend another or, if he cannot, the attack changes targets to the taunting PC. However, we want to model the idea of a "Main Tank" so let's add one more twist: at the end of a turn (after the last PC, before the Boss), all the taunt tokens need to be given to one character. This also saves the trouble of havign to track everyone's threat tokens.

For the third and final bit, , we need to add in a little bit of extra variety - enemies that support the boss, terrain features and so on. We don't want these to be too fiddly, so we're not too worried about details - what these should take the form of are choices - opportunity cost decisions that give a reason to choose one action or another (rather than provide only one obvious choice).

There will generally take two forms: The first will be special actions that can only be taken at specific times and places. If the fight takes place in a fallen cathedral, perhaps there are candles to be lit (which takes an action from someone in the middle range), and lighting a candle reduces the number of threat tokens the boss has, or grants attackers +1 damage or otherwise provides some benefit. These are easy to design, and simply depend upon the description of the environment. The only limiter is that these are usually only useful in a single phase of the fight.

Representing secondary creatures is a little more difficult. Mechanically, they should be represented as one more power of the boss - he spends menace to make them appear, but since it's not an attack, it's not influenced by aggro. Support enemies will appear in a given area and be represented by some number of threat tokens. They work like the boss, in that players can attack them instead, and if they clear a threshold of damage (usually the boss's level -2) then they remove one of the tokens. If there are any tokens left when the boss acts again, then they get used. What they're used for depends on the monster: sometimes they'll be turned in for straight damage to a target or targets in the area their in, but if the creature is more built for harassment, then the boss may get those threat tokens back in addition to any it generates this turn.

Now, with the reminder that this is still just an interesting stunt, tomorrow I'll try to take this theory and create a sample or two, and maybe discuss how to use some of these ideas in a less over-the-top fashion.


  1. Now, admittedly the fact that in a WoW raid this knowledge is gained by dying repeatedly until you get it right...

    Or reading the wiki. :)

    Here's a thought: Run part of the battle as a cinematic flashback as per DMG2, showing the last raid that attempted this, and maybe a few escaped.

  2. I've been thinking, one area where the tabletop can shine over the CRPG style of raid is easily adding context and complications. I think I jabbered yesterday about SCs in the middle of the raid, but you can also give skills new light within a raid context. Monster Knowledge checks can give you invaluable information, as can perception or insight checks.

    You can weave that all into the gameplay experience and generally allow for greater player creativity than an MMO would typically allow.

    I'm interested to see some examples tomorrow.


Note: Only a member of this blog may post a comment.