I'm looking at my first foray into developing a game, and would love to know whether you guys have any thoughts on game balancing on limited multiplayer games.

The game I have in mind involves a neutral player that has to achieve a goal, with two supporting "deity" players who are one of 'good' and 'evil' - One of the deity players would try to help the player achieve their goal, while the other would try to thwart them.

Any thoughts or pointers on how I can ensure the deities are balanced?

If you want me to expand, I will, just didn't want to give away too much of the game play before I finish it.

  • \$\begingroup\$ Simon, I edited your question a bit to ask about strategies for balancing. I hope that's OK. I've seen other balancing questions that do well, but questions that ask for ideas or books to read sometimes get closed. BTW, sounds like a neat game. \$\endgroup\$ – MichaelHouse Oct 11 '12 at 16:54
  • \$\begingroup\$ Thanks, that's cool, just signed up here, so I'm not 100% on what the standards are yet. I'm pretty excited about building it! \$\endgroup\$ – NotSimon Oct 11 '12 at 17:00
  • 4
    \$\begingroup\$ Lots of play testing and lots of trial and error. Once you have a playable alpha/beta, reach out to other developers and designers to see if they are willing to help test and give feedback \$\endgroup\$ – Justin Skiles Oct 11 '12 at 17:13
  • \$\begingroup\$ @JustinSkiles You should make that an answer. It will probably be accepted. I'll certainly upvote it. \$\endgroup\$ – Engineer Oct 11 '12 at 17:34

The simplest way to ensure things are balanced is to make them the same. Give both sides the same powers, give them different names and change their visual effects.

I agree, that's pretty boring and likely won't give you very interesting game play. However, it's a good starting point.

Once you've got a good starting point, with good and evil having very similar "base" abilities, you expand. But you always expand both good and evil, in the same ways. Either give both sides similar abilities or give them equal but opposite abilities. Giving the evil deity a ability to create walls? Give the good deity an ability to remove walls. Giving the good deity an ability to heal over time? Give the evil deity a way to harm over time.

OK, things have gotten a bit more interesting. Deities don't have the exact same abilities anymore with just different visual effects. But, still kind of boring and not really there yet.

Next up is adding unique abilities for both sides. This is where you want to do some heavy play testing. When you're adding abilities that don't have a "counter ability" on the other side, it's important to play test. Obviously you'll want to play test with all the abilities, but these are the most likely to introduce unbalanced play.

  • \$\begingroup\$ I'm not so sure that you should create abilities and counters on a one-to-one basis. Combinations of abilities can create situations that are vague to define, but still need a counter. The counter will probably be equally vague. \$\endgroup\$ – AlbeyAmakiir Oct 11 '12 at 22:31
  • \$\begingroup\$ @AlbeyAmakiir I'm not sure I understand your comment. If a ability doesn't have a counter then it would go into the "unique abilities" category I talk about last. \$\endgroup\$ – MichaelHouse Oct 11 '12 at 22:35
  • \$\begingroup\$ Ah, I misread your last paragraph. Ignore me. >_< \$\endgroup\$ – AlbeyAmakiir Oct 11 '12 at 23:45
  • 1
    \$\begingroup\$ Healing and DoT are not opposite. You can use a DoT and kill a unit with it, but you can't resurrect with healing (or can you?) \$\endgroup\$ – Markus von Broady Oct 12 '12 at 20:52
  • 1
    \$\begingroup\$ @MarkusvonBroady It would depend on how you defined the DoT and Hot. I agree it's possible to err at this stage. Hopefully, the playtesting should reveal such inconsistencies. \$\endgroup\$ – MichaelHouse Oct 12 '12 at 21:43

Play test. Play test some more. And keep play testing.

One good metric during play testing for balance is win ratio. A truly "balanced" game would have each side winning 50% of the time. Keep this in mind, as it can help with the "boringness" of equal-but-opposite that Byte56 suggests to a degree. Here's how to expand on his excellent answer:

By making some skills not quite equal to their opposite, but gaining a bonus elsewhere, the game can feel more strategic since certain skills have an advantage or disadvantage. Some examples: evil can make 2 walls, good can only break 1. But good can give a minor speed buff for X seconds. Evil can stun or slow for X/2 seconds. Evil's damage hurts for just a little more than good can heal, but the heal has a slightly lower cooldown or cost.

  • 1
    \$\begingroup\$ Excellent points. One should definitely consider the entire skill set when balancing, not just isolated sets. \$\endgroup\$ – MichaelHouse Oct 11 '12 at 18:32
  • \$\begingroup\$ +1 just for the playtest part. This can not be overestimated. Get experts in the genre to playtest, if you can. They will push the system to it's limits, and try things you never expected. \$\endgroup\$ – AlbeyAmakiir Oct 11 '12 at 22:33
  • \$\begingroup\$ Just remember that if evil can create 5 wall segments, good will be fine with destroying just one segment to make a hole, making it unbalanced, because evil builds the wall longer (5 clicks) then good destroys it. \$\endgroup\$ – Markus von Broady Oct 12 '12 at 21:48
  • \$\begingroup\$ What if Evil doesn't have to put the walls next to each other? If the arena is small enough, or the walls are long enough, evil can make a switch-back path where destroying one wall only reduces the number of switch backs by one. \$\endgroup\$ – Orin MacGregor Oct 15 '12 at 11:33

Aside from the tips in the other answers, you can add self-balancing mechanics to the game. This is not a substitute for balance, but can make the job of balancing easier.


  • Introduce prediction and bluffing. Useful if the player won't have time to react to everything, and must prepare in advance. "The opponent has multiple options, and he could do any of them, but this option is safer, and he keeps doing it. On the other hand, he's probably realised I have noticed his pattern, and may choose to do the other option to catch me off guard." See: Every good fighting game, strategy game, gambling game, ever.
  • Minimise the chance of permanently locking down the opponent with an endlessly repeatable strategy, or repeatedly using what is found to be a "best strategy". I know you're not creating a fighting game, but as a great example, Guilty Gear XX and BlazBlue both have many of these mechanics, allowing their characters to be quite bizarre, but still somewhat balanced. The more hits you make in a single combo, the weaker they get. Gravity increases as you juggle the opponent. Block too much and you get weaker. Among others. All characters have access to all of these self-balancing mechanics.
  • Have multiple winning conditions. This relates somewhat to the first point, but also allows you to react to poor situations by changing goals. It creates more complex resource-management decisions; go all out or save some in case things go badly? Fighting games usually have both knockout and time-out goals. It doesn't need to be built in; in StarCraft, for example, while the goal is only to wipe out your opponent, you will see top players concede and end the game early for any number of reasons.

David Sirlin has a great set of articles on balancing multiplayer games here, and an in-depth one on the self-balancing systems of Guilty Gear XX here.


Play testing is great, and not just for balance issues. It does run the risk of tuning the balance to your testers's skill level.

I am a big believer in using maths for balancing. If you can find a way to define what you want to balance as a number (often damage output) and add in a few reasonable assumptions about context (often player skill and environmental factors) then you can pull it into mathematical equations that can help you make decisions.

For some fabulous examples, check into theory-crafting done for WoW and probably other MMORPG's. (I haven't been a part of that in ages, but I believe elitist jerks is still going strong)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.