Announcement

Collapse
No announcement yet.

Preliminary Monte Carlo Attach Mechanics Results

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Preliminary Monte Carlo Attach Mechanics Results

    Hi everyone, I have started a Monte Carlo approach to analysing attack mechanics. Thing I am looking at are how defense, armor and stunt points affect average damage output.
    I have built a model of exclusive use of Mighty/Lethal Blow and Lightning Attack. Model is on a zero armor, 10 defense paper and a character with zero strength bonus (min str is a flat bonus to all). I was specifically looking at what Lightning Attack (LA) should cost in SP so it's damage equals that of Mighty or Lethal Blow (MLB)
    • For a light weapon (1d6 dmg) MLB equals LA if LA is priced at 0 SP.
    • For a one handed weapon (2d6 dmg) MLB is closest to LA if LA is priced at 1 SP. If priced at 0 SP it is much higher than MLB (7.1 at 0 SP vs 5.5 at 1 SP where MLB is 5.7)
    • For a two handed weapon (3d6 dmg) MLB equals LA if LA is priced at 4 SP.

    The Monte Carlo analysis is done where the attack roll for a single trial is the same for L,1H,2H making the comparison stronger than if everything was independently random (think paired T test). 90 000 trials were calculated which is almost enough to capture the a 36 damage lightning attack with a 2 hander (less than 1 in 50000 chance). I have to press F9 a few times to get it to happen.

    Anyway, this is pretty good evidence about the effectiveness of daggers versus 2h. This improved pricing should be considered. While it may seem that making these two stunt strategies equal reduces a player's choice, it's still a gamble on higher moderate damage versus unlikely huge damage.

    Further work:
    • Creating a surface plot of damage (z) with ac and defense on the x,y axis
    • adding dual strike
    • adding pierce armor


    Any other ideas, thoughts on this analysis?
    [URL="http://herdingdice.tumblr.com/"]Herding Dice[/URL] - A tumblr where I put things about Adventure Game Engine design.

  • #2
    Re: Preliminary Monte Carlo Attach Mechanics Results

    Holy shit M! Monte Carlo?

    Talking about sledgehammers and nuts here, don't you think?

    I mean, go nuts as much as you like with the statistical analysis, but I for one think that's jumping the shark here.

    Please, don't take me wrong. I really appreciate every insight you delivered, but it's still only a game.

    That being said, have fun and happy statistics!
    Fantasy AGE Homebrew Wiki: Take me to the wiki!

    Free-Form Fantasy AGE System: Essentials - Version 2.5

    DracoDruid's Fantasy AGE: The Thread | Version 2.4

    Fantasy AGE - Revised Alchemy Rules: The Thread | Version 2.0

    Free-Form Dragon Age System: The Thread | Essentials

    Comment


    • #3
      Re: Preliminary Monte Carlo Attach Mechanics Results

      Originally posted by DracoDruid View Post
      Holy shit M! Monte Carlo?

      Talking about sledgehammers and nuts here, don't you think?

      I mean, go nuts as much as you like with the statistical analysis, but I for one think that's jumping the shark here.

      Please, don't take me wrong. I really appreciate every insight you delivered, but it's still only a game.

      That being said, have fun and happy statistics!
      If you want to figure out what actually happens in the game as a whole, its either that or analysis from exhaustion. Otherwise what you think is going on is colored too much by confirmation bias and other issues. The hobby is full of people who think mechanics are doing one things when the numbers don't bear them out, and while statistical analysis can miss issues if you don't apply it properly, in the end, its the only tool you've got for anything beyond subjective opinion.

      Comment


      • #4
        Re: Preliminary Monte Carlo Attach Mechanics Results

        Well, my first contact with Monte Carlo was structural lattice calculations for solid state physics on a computer cluster, soo...

        What I wad trying to say was:

        It's still just a game. Don't overanalyse everything.
        Fantasy AGE Homebrew Wiki: Take me to the wiki!

        Free-Form Fantasy AGE System: Essentials - Version 2.5

        DracoDruid's Fantasy AGE: The Thread | Version 2.4

        Fantasy AGE - Revised Alchemy Rules: The Thread | Version 2.0

        Free-Form Dragon Age System: The Thread | Essentials

        Comment


        • #5
          Re: Preliminary Monte Carlo Attach Mechanics Results

          So the first post is a bit mis-interpreted when I stated what stunt points could be repriced to. I was making comparisons in cost as if you needed to spend stunt points for a roll like you would spend mana. Really what I should (and have started doing) is running against a 0 defense paper so that all rolls "hit" and generate SP.

          However, i think the work still speaks strongly to the imbalance of daggers,1h,2h weapons. I put dual strike into it today and the charts looked cool. The 2X multipler effect that is dual strike makes the damage graph saw toothed instead of smooth like mighty/lethal blow and lightning attack.

          One idea I had was a new Lightning attack that says "You make a second attack against the same enemy or a different one in range for 1d6 plus your weapons ability bonus (str or perc)." This makes it a whole second attack for a light weapon but a smaller bonus but still useful for a 2 hander. You would not roll to hit for this lightning attack.
          [URL="http://herdingdice.tumblr.com/"]Herding Dice[/URL] - A tumblr where I put things about Adventure Game Engine design.

          Comment


          • #6
            Re: Preliminary Monte Carlo Attach Mechanics Results

            Well, if the rules show two-handed weapons as superior to daggers, then I would call this a reality win.
            The same goes for complains about daggers having almost no effect against heavy armor. Well, duh. That's what it's for. And that's why big ass weapons were made.

            As to MB/LB vs LA, maybe LA should cost a number of SP equal to the weapon's MinSTR.

            Side question: did you still price MB/LB with 2 / 5 SP in your calculations?
            Fantasy AGE Homebrew Wiki: Take me to the wiki!

            Free-Form Fantasy AGE System: Essentials - Version 2.5

            DracoDruid's Fantasy AGE: The Thread | Version 2.4

            Fantasy AGE - Revised Alchemy Rules: The Thread | Version 2.0

            Free-Form Dragon Age System: The Thread | Essentials

            Comment


            • #7
              Re: Preliminary Monte Carlo Attach Mechanics Results

              Originally posted by DracoDruid View Post
              Well, my first contact with Monte Carlo was structural lattice calculations for solid state physics on a computer cluster, soo...

              What I wad trying to say was:

              It's still just a game. Don't overanalyse everything.
              One person's overanalysis is another person's simple analysis. Bluntly, if you don't need someone else's analysis, ignore it; that doesn't mean someone else won't want it.

              Comment

              Working...
              X