Game Balancing

Discussion in 'Game Mechanics Design' started by Michael Caiola, Sep 29, 2016.

  1. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    I'm hoping this is in the right forum. I'm designing a game base on the Zodiac and Greek and Roman mythology. There are twelve heroes, each corresponding to a zodiac sign. For gameplay purposes, they're split into groups of three based on their signs' elements. So, three fire heroes, three earth heroes, three air heroes, and three water heroes. 95% of battle will be 3 heroes vs various enemies. There will a few team ups to take on bosses, so occasionally the battling will be 6 heroes vs the enemies, and the final boss will have all 12 battling.


    I'm good with story telling and I'm with the technical aspect of designing the game. Where I lack intelligence is in game balancing. That is, what should my hero and enemy stats be? I have a basic idea of how each character compares to other. For example, my Aries hero is the strongest, my Taurus hero has the highest defense, my elemental fighters like Libra and Pisces have low physical stats, but high magical stats. I also know that I want the heroes to complement each other. Where my Ares hero has no magic abilities, my Leo hero has low physical stats and uses magical skills.


    I'm lost when it comes to specifics, though. What should base stats be? How big should buffs or debuffs be? How strong should armor be? How much should it change stats? How strong should enemies be?


    Are there any resources that can teach me or help me understand game balancing? I know it's not as simple as typing numbers into the stat settings, but I'm eager to learn and study. This game is a passion of mine and I'll put in whatever work I have to.


    If you're willing to help me personally instead of referring me to sources, I would graciously accept it. But I'm perfectly okay with going to sources and learning myself.


    Thanks.
     
    #1
  2. Andar

    Andar Veteran Veteran

    Messages:
    28,627
    Likes Received:
    6,573
    Location:
    Germany
    First Language:
    German
    Primarily Uses:
    RMMV
    Game balancing has two key components: Mathematics and Playtesting.


    This basically means that you should start with a mathematical sequence where you guess that it should work correctly - that is based on the skill damage formulae and your decision on how much damage they do compared to the numbers in HP.


    Personally, I suggest using Excel or a similiar program to make the basic calculations.


    After that, you'll need a lot of playtests to check if the numbers really work the way you planned them - there are too many details (like how fast the player progresses on the map) to completely calculate the entire balance.
     
    #2
  3. Waterguy

    Waterguy Veteran Veteran

    Messages:
    1,188
    Likes Received:
    256
    Location:
    Brazil
    First Language:
    Portuguese
    Primarily Uses:
    RMMV
    I usually play the game as I develop, to know at what level a non-grinding but also non-escaping player would be. If the enemy is too hard for me at that level, and I know the way I was expecting to defeat them, then they are definitelly too hard for the player.


    That's why my game got two initial slimes. The gist was too strong for a party of two level 1, but not that much harder, so I made it a rare, harder version only found if you went "off-road". Two to three levels later it would be just as easy, but for a level1 it is not so it is a good challenge for not-so-rare-but-still-an-upgrade extras.
     
    #3
    Raol and Michael Caiola like this.
  4. Kes

    Kes Global Moderators Global Mod

    Messages:
    20,946
    Likes Received:
    10,652
    First Language:
    English
    Primarily Uses:
    RMVXA
    A couple of points to bear in mind when first starting:

    • How much variance do you want to give to your skills?  The default is 20% which I personally think is too high for your ordinary skills.  That is because it is 20% up or down, meaning that you have a total swing of 40%.  Take the example that @Waterguy gave.  If you get several hits coming in at the lower end of the possible range (i.e. variance of 20% less than the basic) then your players could get wiped out because they are unable to defeat the level 1 slime before they are knocked out.  On the other hand, 20% above would make it too easy and you might be training your player to spam attack all the time.  I tend to go for 8 - 10% except for things like high risk:high reward skills.
    • Healing items also need balancing against your damage levels.  This needn't be done straight away, you need to get a feel for how the damage is coming out before tackling this, but it is something to keep on your "to do" list.

    As has already been said, nothing replaces multiple play testing of your skills.  No matter how much thought you put into your skills, they are going to be affected by what equipment the player will likely have at that point in the game, the number of enemies in the troop and what skills you have given them, etc.  Only play testing will reveal if you have got it balanced or not.
     
    #4
    Raol, Dr. Delibird and Michael Caiola like this.
  5. StrawberrySmiles

    StrawberrySmiles The Talking Fruit Veteran

    Messages:
    1,722
    Likes Received:
    543
    Location:
    South Carolina
    First Language:
    English
    Primarily Uses:
    RMMV
    I'm a lot like the OP, actually. Math is my weakness, for one. The whole balancing aspect seems quite overwhelming, doesn't it?


    I even tried to find ways to just not have combat at all, though that didn't go far.


    Using Excel seems like a good idea, though. Might have to try that.
     
    #5
    Michael Caiola and Rayhaku808 like this.
  6. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    Math is actually one of my strong points. I understand the formulas and everything.


    I guess what I need help with is something like, should armor increase defense by 10%, 20%, 50% or something else? What should the proportion of base stats to armor and weapons?


    How strong should enemies be by comparison? For a simple grunt, should it take more than a minute or so to defeat it?


    The only one I really have an idea about is the final boss. Since he's facing 12 heroes, he needs to be at least 12 times as strong as any one of them. And probably stronger because of the heroes potential healing abilities and health items and such.
     
    #6
  7. Andar

    Andar Veteran Veteran

    Messages:
    28,627
    Likes Received:
    6,573
    Location:
    Germany
    First Language:
    German
    Primarily Uses:
    RMMV
    you're starting with the wrong questions.


    your first question has to be the way you want to handle combat from a time-based point. Do you want many short fights or few long ones or few short ones? How much time (in percent) should the player spend fighting?


    That will determine how long the battles should be in turns, and the enemy HP divided by the intended number of battleturns gives you the average damage to do per turn.


    Only after that you can think of the differences between attack and defense to create that average.


    Then one additional step is to decide on the pacing - how many levels should the actors have and how long should the game be will determine how long the player has to play for each level, and that in turn will give you a number of EXP to distribute per hour - on the number of battles per hour, which will determine the EXP per enemy.


    And so on...
     
    #7
  8. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    @Waterguy, that's the way I had planned on doing it.


    @ksjp17, I hadn't thought of the variance issue. That's a good point. I had played with the idea of no variance at all. Every attack or skill does a specific damage based on all the affecting parameters. Is no variance a good idea?


    @Andar, that was really helpful. That's kind of like reverse engineering it. I had put some thought into that. I want random grunt battles and planned boss, sub boss, and event fights. The randoms should be quick and the planned should be longer.


    The main focus of my game is the story and it's told mainly through exploration and boss fights. The grunts serve as minor obstacles and fuel for leveling up. That seems like a pretty standard model and I just want to make sure I get it right so people can play and not get bored.
     
    #8
  9. Kes

    Kes Global Moderators Global Mod

    Messages:
    20,946
    Likes Received:
    10,652
    First Language:
    English
    Primarily Uses:
    RMVXA
    Variance, the amount or even existence of it, does depend on what style of game and battles you want.  No variance gives a high level of predictability, but the risk is that it reduces the tension of a battle a bit.
     
    #9
    Michael Caiola likes this.
  10. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    I was going to make up for that with other things like enemy traits or surprise mid battle enemy entrants. I want the player to be able to plot things out, especially since my heroes are going to have ways to work together.


    For example, my Ares hero does get a fire sword at one point despite having zero magic attack. But my Leo hero will have a skill or trait that boosts the power of fire skills and attacks. So as long as she's alive, the fire sword can do extra elemental damage. If she gets knocked out, the fire buff goes away and the sword does hardly any elemental damage.


    So with things like that, I think variance would have a detrimental effect on strategy. Would it?
     
    #10
  11. Kes

    Kes Global Moderators Global Mod

    Messages:
    20,946
    Likes Received:
    10,652
    First Language:
    English
    Primarily Uses:
    RMVXA
    I'd be reluctant to commit myself to a definite opinion without play testing, as that can throw up all sorts of things.  There are many good ideas around, but it always comes down to the implementation of them in the end.
     
    #11
    Michael Caiola likes this.
  12. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    Duh, every skill has its own variance... So I can do it on a case by case basis.
     
    #12
  13. Waterguy

    Waterguy Veteran Veteran

    Messages:
    1,188
    Likes Received:
    256
    Location:
    Brazil
    First Language:
    Portuguese
    Primarily Uses:
    RMMV
    Variance is the random factor in the damage. With it at 0, you can calculate it through a formula. Variance makes it be just a bit different every time. It adds a naturality to it.
     
    #13
    Michael Caiola and kaukusaki like this.
  14. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    I already knew that. Thank you, though.
     
    #14
  15. Waterguy

    Waterguy Veteran Veteran

    Messages:
    1,188
    Likes Received:
    256
    Location:
    Brazil
    First Language:
    Portuguese
    Primarily Uses:
    RMMV
    ...So I don't get your logic that Variance is any worse in your example double tech than in any other...
     
    #15
    Michael Caiola likes this.
  16. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    If it's a simple attack with no other factors then variance could be useful. But in situtations where I'd rather have the player use strategy, I'd prefer it not to be random.


    That said, I realize my example was a poor one.
     
    #16
  17. bgillisp

    bgillisp Global Moderators Global Mod

    Messages:
    12,294
    Likes Received:
    12,510
    Location:
    USA
    First Language:
    English
    Primarily Uses:
    RMVXA
    For me, I started out balancing by throwing numbers at the wall and hoping they stuck. Seriously. My focus was on the plot first, and the balance was secondary. I actually for a while had all enemies with 1 HP until I found a script to let me just instant KO all enemies (for ACE) so then I guessed, using the party stats as a baseline.


    Now that I have my plot done in the game, I'm going back and balancing. I put in excel stats like the average player's HP, ATK, and DEF at certain levels, and my damage formula (which has low variance now), and used that to get an idea of how much HP to give my monsters, as well as how much ATK they should have, based on how much damage I wanted them doing.


    For EXP, I first figured out how many of each monster of equal level to the player do I want them to kill for a level up (some developers use 24 I've heard. Exact numbers will depend on how slow you want leveling to be)? Then I took the EXP required to level up, divided by that number, and then assigned it to the monster depending on what 'level' I assigned it.


    I'm still figuring out G balancing, so check back with me on that one in a few months.
     
    #17
    Miss Nile and Michael Caiola like this.
  18. DarthVollis

    DarthVollis Moogle Master Veteran

    Messages:
    410
    Likes Received:
    85
    Location:
    Kupoburg
    First Language:
    English
    Game Balancing can be a pain. Lay out your game first and get all the puzzles, dungeons, and placements correct before balancing. In the case of my game I playtest it all the time. So much in fact that I am sick of certain parts, but we have all been there. For enemies do not make them so strong that players will give up instead of trying to defeat them. That should be reserved for bosses but not all of them. Make some bosses weaker just to make the player feel good so that you can throw a hard boss at that them. That feeling will cause the player to try and be surprised at the boss. For armor and weapons NO GOD ITEMS! That can ruin a game. For example I am playing Dark Souls 2 right now and am enjoying it but of course I have died a few times (mainly from jumping wrong and walking off duh). God Items would ruin it for me and I would not play it any more.


    For the most part playtest, playtest, playtest.


    Of course let me know and I would be happy to playtest and give notes.


    Best of Luck


    DarthVollis
     
    #18
    Michael Caiola likes this.
  19. Michael Caiola

    Michael Caiola The Stone Bull Veteran

    Messages:
    392
    Likes Received:
    74
    Location:
    Albany/Saratoga, NY
    First Language:
    English
    @bgillisp, that's really helpful. Thank you. When you say "G balancing" are you referring to gold? I don't plan to have a buy/sell system in my game. The story does really allow for it. The heroes are navigating a hostile Earth to rescue people and fight the baddies.


    @DarthVollis, I've realized endless playtesting will be necessary. I'm fine with that. I like the boss tips. It's funny you should say no God items, because the heroes are based on Greek and Roman gods, so... they have god items and abilities. Lol. In the context of gaming, what does "god item" mean so I know if I'm making a bad decision?
     
    #19
  20. DarthVollis

    DarthVollis Moogle Master Veteran

    Messages:
    410
    Likes Received:
    85
    Location:
    Kupoburg
    First Language:
    English
    God Items means weapons and armor that make the player invincible or almost invincible. A weapon that makes one-hit kills should not happen or be so hard to get that it makes it worth it. That is what I consider God Items.


    I did think it was funny to mention that because of your project being about Mythology (one of my favorite subjects by the way).


    In my opinion never make it so that there is not a challenge at all because that ruins games.
     
    #20
    Michael Caiola likes this.

Share This Page