D&D 5E Should DnDBeyond include an Encounter Simulator?

Clint_L

Legend
Uh, yes, it would be "that hard".

You're talking about having an engine that will enact the mechanics of the entire ruleset, movement and all, and make combat decisions for each monster and character.

If you then added graphics, it'd basically be the engine for a full D&D computer game. And then what you are really testing is not the power of the characters vs monsters, but how smart the "AI" for combat choices is.
I don't think it would be perfect - you wouldn't have to simulate every possibility - but sure it could get more granular than what we have now. As mentioned above, haven't a number of folks already built encounter simulators?
 

log in or register to remove this ad


Umbran

Mod Squad
Staff member
Supporter
As mentioned above, haven't a number of folks already built encounter simulators?

It is one thing for a third party to build a thing for their own amusement. It is quite another for the publisher to build it and offer it up as an official tool. If a hobbyist does it, and it doesn't work great, we shrug and move on. If WotC built it, and it fell short, we'd castigate them for it.

Making your own toys and making official customer software tools are not the same thing.
 


Leatherhead

Possibly a Idiot.
I think people say it's too hard because they they get caught up in the trees while looking for a forest.

For starters, you know how you run monsters, you don't have to guess at what strategy you will be using for them. So that's half of the combat solved already!

Secondly, you don't have to account for every single player action, just the most likely ones. After playing for 10 levels, you should know that the Cleric is casting Bless, the Sorcerer tosses out Fireballs, the Barbarian runs up to chop the biggest threat with their axe, and the Wizard is going to be doing some kind of crowd control.

And the hardest part of that calculation, the crowd control, is way easier to account for than people think it is. All you have to do, is plug all the monsters into the action economy and subtract some actions. You just figure out how much hurt they do per action, and then guestimate how long they will be out of actions for.

Like if Web is the control spell of choice, you take 1 action away from everything likely to be caught in the Web. Because Web is designed to be broken or burned, and you can just save against it if you are lucky.

On the other hand, if it's something like Force Cage, well that target is just sitting out for the entire fight. But still easy to calculate because they don't start damage till everything else is gone.
 

Leatherhead

Possibly a Idiot.
But but more importantly, CR is internally calculated using a mathmatical formula that has never been published. It is going to be presumably published in the upcoming revision, alongside an overhaul of all the MM monsters. It primarily takes into account actions to kill a target, while providing lots of quirky ways for monsters to break the action economy by doing their non-damage stuff along side their normal attacks.

So basically, CR is being "fixed" right now, and it's all DPR based.
 


Stalker0

Legend
If combat was that predictable, playing the game wouldn't be fun.
There’s a difference between: a combat will always look like the sim and combat will on average look like the sim.

As already noted, versions of this have already been done, and done pretty well. Of course every party will be different but it gives you a baseline. If my simulated encounter says the party will win 99% of the time with “basic generic actions”…ok I know this encounter is likely too easy.

It’s not perfect, but it’s way better than just using a single static number
 

Clint_L

Legend
This is what I’m suggesting. Obviously you would never get (or want) a perfect simulation. But the encounter builder could easily work in a lot more of the data it has at hand: how many HP is the party at? What are the spell DCs, to hit and damage modifiers? What is the spell load out? How many spells and limited use abilities are available? How many ki points? And so on. All that data is sitting there.

I run simple combat simulations in my head all the time to help go to sleep (things like “how often will my level 10 monk beat a cloud giant 1v1 with the new rules vs. the old). DDB could do that level of calculation trivially, and it would be way better than what we have now.
 

Isn't part of the fun of the game discovering what happens based on the nigh infinite variables of PC/NPC/monster stats/abilities, player/DM decisions, environmental factors, and - perhaps most importantly - the fickle nature of the d20? Is it really that helpful to know, assuming said variables can truly be quantified accurately, that the likely combat outcome for the party will be 82.5% victory, 10% stalemate, 7% defeat, 0.5% TPK (or whatever)? Not to mention, it seems like an extra step of prep to worry about "likely outcomes" of combat when an encounter may not even be resolved by combat. Negotiation (see social interaction rules in DMG), Parleying (see Tasha's), Stealth, Evasion, etc. - the DM doesn't necessarily know what the party will choose to do in a given situation. Or will those need to be programmed into the simulation as well?

In practice, I tend to create encounters that are primarily appropriate for the adventure location - it is a secondary consideration that they are an appropriately "balanced" encounter for the party. In game play, when combat is the chosen encounter resolution, that means that sometimes the party will enjoy a cakewalk, occasionally the party will be in over their head, and sometimes the encounter is juuuuuust right (whatever that really means... I guess it feels challenging to some extent but doesn't end in a TPK/defeat). I used to employ Kobold Fight Club but haven't in some time - and it has made no difference in the enjoyment of sessions, IME. I also try my best to run combats according to the motivations of the baddies. Not every fight need be to the death. The Monsters Know What They're Doing is a great resource for this.
 

Remove ads

Top