Thursday, April 7, 2011

Balance: Third times the charm. Right?

I suspect many people who read my blog don't follow the comments when those comments go on at great length and over several days so I want to summarize a bit what came of the last balance post I made on tuesday.  lebkc came up with a definition for a balanced game that read like thus:

"equally successful for all players under perfect play"?

I think I would rewrite it slightly to be like thus:

"A balanced choice is one where any of the options will produce a similar average result in terms of performance when chosen by a player that otherwise plays perfectly."

Thing is, lebkc's argument doesn't work for game balance on a whole game level at all but I think it works great for a single choice.  I don't know which way he intended it.  It isn't doing anything particularly innovative there but it certainly sums up the idea that if a set of choices all have similar performance results when played perfectly they are all balanced against one another.  However, I was talking about defining what makes a game balanced or not rather than what makes a choice balanced or not.  I don't think I was particularly clear on that point which is probably why the discussion has been so vigorous and we have been so much talking past each other.

Any choice within a game can be put into the set of choices that is balanced (as per definition above) or unbalanced.  The trick to defining a balanced game in my mind is to figure out how we derive overall game balance from looking at the individual choices.  Clearly all games of any complexity have many choices that are widely unbalanced but this does not prevent us from thinking of the game as unbalanced at all.  In chess you can spend every move moving a knight in and out of his starting position - this is not balanced as a strategy against "play well" but it doesn't make chess unbalanced. In WOW you can refuse to put on gear and this is not balanced against putting on gear but yet that isn't going to change our minds about whether or not we call the entire game balanced.  It is clear that there are some specific choices that will trigger us to think that a game is imbalanced but that the size of balanced/unbalanced sets is not the trigger and I can't see any other clearly identifiable pattern.

As such I don't think we can derive a definition for a balanced game that can be arrived at without looking at the preferences and beliefs of the players.  Clearly humans have tendencies and we can make good educated guesses about what they will think and reasonably call things balanced or not based on those guesses and I don't think there is any other way to resolve whether a game is, as a whole, balanced or not.  Individual choices can be mathed and played out to arrive at a defensible, logical conclusion as to their balance or lack thereof.  I don't think a whole game can be approached in this fashion.


  1. I know I've been aggressively critical of your propositions for a definition of balance, but I think the discussion has been very illuminating. I think both of your proposed definitions highlighted challenges in getting at the idea of the balance of an entire game.

    If I had to try to explain game balance in a simple way (not offer a definition, but explain it to someone who is not used to hearing the term) I would say that when we talk about balance in games we are often talking about two different things being equivalent in some way. So for classes in an MMORPG it would be their ability to perform their function in groups. For factions in an RTS, it would be that no faction is clearly better than another or clearly dominates another in direct conflicts.

    Then we could go on to say that a game is balanced if all of the "important" parts of the game are balanced. If they asked how we knew what was important then I might fall back to something about broad consensus - and there are the preferences and beliefs your are talking about.

    One of the questions I always consider when I'm trying to get to the bottom of what I'm talking about is whether we are talking about something or not.

    When people talked about heat thousands of years ago they said a lot of wrong and crazy things. They were trying to talk about this sensation they had that seemed to line up with certain things that happened in the environment, some of which they could predict and some of which they couldn't. I'm sure you would have found people defending the idea that whether something was hot or not could not be objective measured because you had to take into account the different feelings of heat experienced by different people.

    Of courses this was nonsense. Whether or not something qualifies as "hot" in common parlance is certainly a matter of opinion but it turns out that all the time people were talking about heat they were actually talking about something. There is a thing, and it is heat, and they were trying and largely failing to get a grasp on it (at least I think this is the best way to interpret the progress of science).

    So the question is, when we talk about balance of a game, are we talking about something, or are we using it as a shorthand or metaphor. It could be that when we describe a game as balanced all we are doing is saying that a list of properties of that game are balanced (and that list is not strictly determined by the game itself), so it's a manner of speaking about the community's relationship with a game rather than a property of the game itself.

    But it could be that when we describe a game as balanced we are really talking about something. That doesn't necessarily mean we have to be able to figure out what that something is.

    Usually I err on the side of assuming we are talking about things, but this discussion has really made me wonder about that. I can clearly conceptualize what balance means when there is competition between players (even if that competition is not in a PvP game - competition for raid spots), but if there is no competition between players I can't really see how to apply the term.

  2. I think you are hitting it dead on. Balance in terms of a well defined choice or in terms of players competing against one another is fairly easy to define and certainly is a real thing. Balance in terms of 'is this game balanced?' isn't the same sort of measureable, real property of a game in the same way. I think that when we talk about balance of a whole game we really can only look at it from the perspective of the experience of the players and that it is not a properly of a game at all until the players are brought into the equation. That isn't at all the sort of conclusion I expected when I started asking myself this question, nor is it the sort of conclusion I like, but I think it is the correct one.

  3. Hey, long time reader here, I don't remember ever posting though. On to my point...

    In a game of numbers like WOW and Civ5 its easy to measure because there are numbers; there unit costs and there is range and power, all things that can be balanced against each other. But sometimes there is a balance of skill-needed.

    I'm not sure if you ever played DotA but it had a hero with three passive abilities, a slow, a life leech and a movment/attack speed boost (not sure but I think Craix was his name) this hero is by no means unbalanced in a game of numbers, but a new player can easily farm out of sight forever and come back at the end game with the best items and right-click his way to victory. Right-click because he needen't use anything, hence the low skill barrier.

    They fixed him now of course, which goes to show that it was a recognized imbalance even though good players always knew how to kill'm.

    So this adds an even more complix layer to the discussion of balance, should we balance how easy it is to learn a class/faction/hero?

    An old gripe of mine is China Infantry in C&C:G; they get their transporters already full, which really takes off a lot of the micro managment needed. Even though I can manualy fill my GLA transporters and probably win, the fact that his mind and hands were free to do something else meanwhile is an inbalance.

  4. Naix is I believe the hero you're referring to. He definitely had scaling issues with gear and if you let him farm freely he'd dominate the late game. Gank him early and often and you'd win easily.

  5. It is certainly true that a choice that is balanced for experts can be unbalanced for new or weaker players. However, because games played by new or weaker players tend to be quite random and the amount of balance has a much smaller effect on the outcome I pay much less attention to it. A strategy that utterly dominates low level play is a balance concern but as long as weaker players have a reasonable shot at defeating it if they are aware of it I don't see much reason to worry about it.

    We balance for experts because expert play is the only place where you can really determine what *is* balanced with any accuracy and because experts are the ones who care. Ignoring low level play entirely is a poor idea but having lower standards is both common and entirely acceptable.

  6. One of the biggest problems with balance discussions is how hard it is to play a certain strategy. When you describe a situation where it's easy to beat a certain hero, but if people don't realize that they need to pay attention to him he dominates, it's pretty impossible to say whether that is balanced or not in a general sense.

    One of the reasons why WoW class balance discussions work is because everyone seems to have agreed to discuss the balance of the maximum possible output. This creates huge problems for people who aren't at the maximum. Nerf Survival Hunters or Cats because they do too much damage and you'll find that for everyone except the best players, they just plain suck. Nerf Mages because they do too much damage and the nerf applies pretty evenly from the best to the worst players.

    Complicated games, if they are balanced, are generally balanced for either the average player or the top end player. Whichever you balance for, there is a good chance the other will find the game somewhat degenerate.

  7. I don't think it only affects new players though; even experienced players can be disadvantaged if their counter stratgy is a lot easier then what they are doing.

    Say there is a class in wow, say deathknight, can reach the same maximum DPS as a mage; the mage reaches this max DPS by pressing 1 every two seconds while making sure his dot (number 2) and a reactive spell (number three) are always off CD.

    The deathknight though needs to do some serious resource managment or he could end up with free GCDs, which is a huge DPS loss considering.

    Both of them played by expert will reach the max on a dummy; in a fight where there is a very real need for situation awarness to the 0.5 second, the mage, regardless of "proness" will out do the deathknight if the deathknight's cycle is complex enough.

    Of course this isn't a real example, but you can see where I'm coming from, skill barriers should be put on all factions/classes.

    A situation where army A is known as the noob army is not good game design

  8. I agree. I don't think having one army/class/faction/etc. be super easy to play is a good design at all even if other choices are capable of keeping up at maximum skill level. It certainly leads to all kinds of imbalances at lower brackets and there isn't any need for it - having results be skill dependent is generally a good thing.