PDA

View Full Version : Player Help Playtesting Tabletop Characters



Tyeal
2014-06-01, 06:36 PM
My question is about testing a character, more in a combat situation than anything else. Or just tweaking stats with various items, mixing and matching classes and gear, optimizing or just trying to hopelessly break a character in terms of combat potential. Is there an easy way to do this other than waiting to get into a game? Like is there a standard or default combat scenario one can follow to get a feel for a character that you can slowly ratchet up in difficulty?

Something akin to the training dummy in Darksiders 2 that you can just wail on endlessly to learn/optimize combos, or the obstacle course from MW2. Keeping it to just Pathfinder for now, if you're curious about what system.

nedz
2014-06-01, 06:58 PM
Select a monster of the desired type and level and then roll some dice.

It doesn't really cover tactics though — just raw numbers, which you should be able to eyeball anyway.

Rhynn
2014-06-01, 07:58 PM
Just rolling dice is sort of worthless, you'd have to repeat the scenario hundreds of times to get real results.

Take average damages (damage multiplied by chance to hit), compare to hit points. Not just for attacks, but for special abilities, etc. Also, compare special abilities (e.g. "can fly, can cast spells, breath weapon/ranged attack" always beats "can't fly, can't cast spells, no ranged attack" and mostly beats "can't fly, can't cast spells, ranged attack").

It's rough, but it's much better than running individual simulations. It's also very valuable in D&D 3E, PF, and D&D 4E, which are numbers-over-time games (as opposed to games where a single lucky blow can kill anyone anytime).

Slipperychicken
2014-06-01, 11:08 PM
You could build a computer simulation of D&D combat, or at least the bits where things roll dice at each other and modify values. Although that could become really demanding on your computational power, and takes lots of programming know-how and a very deep understanding of the game mechanics (in which case you probably don't need a simulation anyway).


Typically, you can use resources like Anydice (http://anydice.com/) to help determine the chance of a given attack hitting a given AC, to confirm a critical, or to determine what probability there is to deal so much damage in a single hit. Also takes a bit of programming skill.


If the information were available, it would be possible to use machine learning to predict optimal builds by the relationship between character-building decisions and the character's success in-game (character success is obviously multifaceted, including factors like wealth acquired, goals achieved, fun experienced by player, number of times died, fondness with which the character is remembered later, and so on). Which would take some stats knowledge, and a lot of data which hasn't been recorded.