Limiting focus fire in TotM play

[slobo777]

I'm putting together some software to "playtest" building encounters and adventuring days. I've done similar before for 3E and 4E.

So far, combat is one large spinning melee, a bit like a Taz whirlwind, that spits out survivors at the end. Inside it, each combatant attacks a random enemy on their action.

I want to add choices, of actions and who to attack. In practice that means each character will assess its options and choose something that reduces the balance of threat to their team (with "threat" backed by some maths on likely result and "balance of threat" accounting for risks).

I've done this before, and I know that if given free choice of who to attack, the computer will have each team focus fire on every turn. This is a common and natural consequence of ablating hit points (or any other resoucre-based defences, including e.g. actions used to defend).

In real games this happens a lot less. I'd like my program to represent real games in the long term, so it's time to think about this.

What I'd like to build is a simple, robust Theatre of the Mind environment that represents some typical (i.e. non-Taz) combat scenario, and that also restricts choice of who to attack.

Back in the day, my group achieved this in two main ways:

1) Lots of use of choke points - doorways, narrow corridors etc.

2) DM plays "nice" and generally doesn't focus-fire, e.g. rolling randomly or picking a focus for each monster.

these were often used in combination.

The computer can be made to understand (1), but not really (2) because it plays out so subjectively.

I'd really like to give the computer a "killer instinct" and show the game rules playing out with reasonable encounter balance. First because I like to play that way, but second because a big motivation here for me isn't D&D at all, but a chance to play with Genetic Algorithms

So, anyway, the point of my thread: I'd like to canvas for TotM setup, examples of play or adjudication that I can use for inspiration on how to implement this.

 

[I'm A Banana]

Ah. What you need is AI. And possibly aggro.

In actual D&D combat, on the monster side, there's a lot of flexibility in who you target. Perhaps some monsters are rutheless and target those with the lowest defenses (until prevented). Perhaps some are predators and pick on the smallest, weakest-looking member (and retreat when it becomes difficult to take them out). Perhaps some are cowards, only attacking one target en masse, and then fleeing when they're forced to go it alone (or even with equal numbers!). Perhaps some are simplistic, and attack "whoever's in front." Perhaps some are vengeful, and attack whoever deals the most damage. Perhaps some are persistent, and won't relent on a given target until that target drops.

On the player side, in D&D, the most typical limiting factor aside from terrain (which is still usable in ToTM, just modified) is who you're "in melee" with and who you're not "in melee" with. It might also vary with who dealt the most damage -- the biggest threat. If you focus fire on one guy with high defenses while the fragile speedster is decimating your HP's, you're gonna lose!

I've personally found that ToTM benefits greatly from more "defensive" abilities, as well. Things like auras that deal damage, counterattacks, insubstantial, resistances, etc. help change the calculus of who can do damage to what.

This isn't necessarily easy to program (Lawd knows videogame devs have tried to get believable AI for years, and still get this thing painfully wrong), but it's easy to DM.

 

[the Jester]

Are you going to build in morale rules for your monsters? Post-Basic, I've always more or less winged it, but I think it's important to factor in the "Oh crap, we're getting slaughtered!!" factor.

Of course, different monsters should respond differently to the same sort of stimulus (maybe kobolds route and flee, hobgoblins retreat to regroup, zombies don't respond at all and sahuagin fall on their own dead to feast).