What is Class Balance

I apologize because this will make no sense-- well, maybe it will make just a little sense. ^^ Skip to the summary and if it interests you, check it out I guess.

If there's one constant in WoW, it's that the majority of classes feel they are at the bottom of the food chain and demand rebalancing to be competitive be it in terms of PvE or PvP performance. As someone who used to PvE extensively, I definitely was as up in arms about these topics as anyone (First Year of WoW PvE endgame IMMUNE to our DPS tree at the time?), and have always been one of the players actively seeking out class changes be they for PvE or PvP; I constantly flooded the boards during TBC beta with commentary because one thing that is true about WoW is that the devs listen and respond to their players.

When I think about mages over the past 2+ years, there's been a ton of changes that directly came from the playerbase:

• New Rank of Water (4 Bottle at a time though)
• Mage Armor (To aid with downtime)
• +Dmg Coefficient Revision
• +Dmg Itemization Increase
• Addition of the stat, +Spell Hit
• Addition of the stat, +Spell Penetration
• Trainable Arcane Explosion, Trainable Evocate, Talent Changes to make "Elemental" Viable
• Water Elemental (Players called for it for months Pre BC) The pet was changed significantly based on player feedback during beta; reagent removed, cooldown decreased, damage increased, can cast while pet uses nova, etc.
• Spellsteal -- This was an ability in fake patch notes that blue said they thought was a cool idea.
  
• Slow - Again came from player suggestions
• Arcane Missiles Coefficient Revision - A series of lengthy threads lead to this change along with buffs to how +Dmg influences Area of Effect Spells
• Reversal of Counterspell Change
• Many more I skipped!

What's the point? The point is that players who can express coherently their opinions on gameplay can alter the game. This, of course, gives incentive to generate huge melodrama about "class balance" as every class does what they can to get buffs for his or her own class.

Now before I derail WAY too much, it's obvious that this cannot end; there's never going to be a day when everyone logs into the forums and posts "Yup, we're done, the game's balanced" as balance is highly subjective. Recently, there have been efforts to more objectively study class balance. Players analyze the class makeups of top arena teams and use these statistics to make arguments for why X class is over or underpowered. Are these arguments logical?

If I say:

Warlocks are clearly overpowered. Why is it that they are undisputed the strongest one on one class, they represent nearly 25% of the 2v2 players, 18% of the 3v3 players, and 7% of the 5v5 players? Shouldn't these ratios be closer to the general class population?

Are these types of statistics meaningful? There certainly are reasons why they aren't. Say warlocks are underrepresented in 5v5 teams, but overrepresented in 2v2s and 3v3s, are these players simply playing the smaller arena sizes because they couldn't land a spot on a 5v5 team?

Point being that people can come up with counterarguments for these statistics, be they ultimately logical or not.

But pose the question a different way:

If you were to bet what class is on the most 2v2 teams at the end of the season, what class would you bet on?

Take it a step further, what is the spread in percentage points between number of warriors on 5v5 teams and the number of druids? (So if warriors are 20% and druids are on 4%, the spread is 16) Would you sell me 10s? (Selling 10s means you think that the percent spread is less than 10 and for each point above or below 10 you pay out $ or collect it respectively)

When you think of some of this stuff from a trading/gambling perspective, it really turns the question into "what class would YOU bet on"; when you do this you can synthetically "price" classes with respect to one another for the various arena brackets and get a general feel with where the various class are with respect to one another.

Why think of the problem like this and not just use the "cold hard numbers" from end of season data?

Because the data will be attacked as invalid. The method for collecting the statistics will be attacked: How do you define "active" players? Rationalizing of the data will occur.

Using markets to naturally price the class relationships creates the beauty that if a person disagrees with the market's opinion: TRADE WITH IT. The problem is, well, of course, you can't. Forum polls and the like are useless as unless people have something tangible to win or lose from expressing opinions, they'll just theorycraft and argue all day long with no consequences; also what value is a market opinion where none of the participants have any "money" behind their opinions.

Breather. I AM talking about betting $$ on whether certain classes are stronger than other classes in a VIDEO GAME. Pretty silly...I think. But then again, does sports betting in a sense accomplish the same thing? That is, by establishing what the traded markets between teams is, one can estimate the future performance and relative strength of sports teams. And what about things like InTrade or TradeSports. These sites let uses bet on all sorts of things and they are extremely powerful for forecasting and analysis.

Breather #2: I've sidetracked to purely a PvP discussion, but realize that the ideas are similar in a PvE context. (think damage meters or something I dunno) What is class balance? It's a forward looking prediction that says how classes will perform relative to one another. If a certain class is forecasted to be dominating next season while another is predicted to suck, that is imbalance, and the predictions carry weight because they come from a large # of people with vested interests in their opinions.

Before I propose trading structured derivs on WoW class balance, does this make any fucking sense? No, not really. We gamers don't care that much; most of the actual players would not even be market participants. Scrap the whole real $ thing, can you do it inside WoW? A Casino! With sports boards like at Caesar's, trade class balance, which arena teams will perform well, etc. Bet WoW gold. Ignore the whole WoW Gold can equal real money side of things because that is a huge confusing mess. With this in place, would you have a real tool to understand how strong classes are with respect to one another? Yeah, probably.

But what about people with huge amounts of gold "bullying" the market?

Yeah, that could happen. But there are a lot of WoW players and you could limit the amount of gold a player can have net invested in the WoW Casino.

Weren't casino bots thrown out of WoW because the whole video game gambling thing is a messy subject?

Yes and who knows.

Summary:

Actually, what I wanted to get across is that class balance is more complicated than giving specific examples of one class ability relative to another class ability. Trying to piece by piece analyze class balance, which is how the WoW forums attack the topic, is logically flawed as what matters at the end of the day is how the class performs. You can measure performance by collecting statistics or by using markets to price predictive performance; I think markets are better because of difficulties in collecting meaningful and representative statistical data. Also, markets are forward looking starting at the present, while statistics just tell you about the previous season's class balance.

Creating tradable markets for things like this is difficult and probably only possible by trying to mirror sports gambling virtually inside the WoW world; this has other benefits because a Casino would probably be generally well received by the player base -- I picture...actually you REALLY don't want to know what I picture. ::shudder::

At the end of the day, all I know is that my class needs buffs and the rest of the classes need nerfs and hopefully I've convinced you all by now. ^^

I think for some the idea of using markets instead of statistics is highly counterintuitive. But, I have an example!

Suppose we play a game where I flip a coin, if its heads you get a dollar, if its tails, you get zero. How much would you pay to play this game?

Well the expected value of the game is 50 cents, so you could pay any amount up to that and it is profitable for you to play the coin toss.

You decide you want to make sure that the coin is fair, and watch me play the game with other players. After 24 flips, I've gotten 20 heads. Are you going to change your mind and pay 20/24 or ~ 85 cents to play the game? Hopefully not. ^^

But that's an extreme, what if after watching me flip the coin, I get 12 heads (what you'd "expect" lol). You notice the other players all paying me 55 cents to play the game. What suckers. Now a few people are paying 60 cents to play, soon a line of thousands of people are all standing in line to play, and are willing to pay 60 cents to play the game. Should you pay 55 cents to play?

What means more, your limited statistical data or the opinions of thousands of people?