Friday, February 22, 2013

Class Imbalance

Gamer A: Class balance is a tough problem, but designers must solve it. Without balance between classes, the game is broken.

Gamer B: Class balance is a phantom. It can't be achieved, it can only be imposed, and when you impose things on characters, the game is broken.

As we designed 2nd Edition AD&D, we didn’t fret much over class balance. That’s not to say we didn’t care or didn’t consider it important. The question of class balance was bandied around endlessly in the office, in letters to The Dragon, and at conventions.

We didn't fret over it for two reasons. First, no one could agree on what well-balanced classes would even look like. Second, the effectiveness of any character or group of characters is influenced by so many variables beyond the game designers' control that we concluded issues of balance are best left in the hands of individual DMs and dealt with at the campaign level.

In 4th Edition, D&D tried to take control of those variables. The trouble is, once someone “takes control” of a variable, it ceases to be a variable. It becomes a constant. One of the regular complaints against 4E is that all characters wind up being copies of each other, number-wise, no matter what the players do. Going up a level takes on aspects of “If It’s Tuesday, This Must Be Belgium.”

Keeping a tight rein on the numbers that way has a positive payoff—largely for the DM and the game's designers, whose jobs become simpler. It comes at a cost in verisimilitude and free choice, and the cost is borne largely by players. Some players are unwilling to pay it.

From a game designer's standpoint, creating a system that establishes balance between characters without interfering in the relationship between the character and the player is a very tricky and difficult problem.

Just because it’s difficult doesn’t mean game designers should throw up their hands and surrender. As my Nuclear Engineering 101 professor liked to say, "this is a technical problem, and it has a technical solution. We just haven't found it yet." Given enough time and money, engineers will find the answer. Seen in that light, the question is revealed to be, "how much time and money are you willing to invest in the search for balanced character classes?"

Before pressing on, I'd like to define two useful terms: "wide" balance and "narrow" balance. To define them, refer back to the previous post on Class Balance. In it, I outlined two simple roleplaying games. One assigned the numbers 5-3-1-0 to character abilities, and the other assigned the numbers 4-3-3-2. Narrow balance is typified by the 4-3-3-2 game. Wide balance is typified by the 5-3-1-0 game.

Narrow balance is relatively easy to achieve. The simplest approach is when the game tells you what your characters can do, and they can't operate outside those bounds. Chess is narrowly balanced. Characters in the D&D boardgames (Castle Ravenloft et. al.) are narrowly balanced.

Wide balance is difficult to achieve. It must leave characters great latitude to specialize in innumerable ways, yet somehow ensure that the character-building resources I invest in Diplomacy skill are as valuable as the resources you invest in combat. Balancing such disparate concepts is like trying to determine which is rounder, an idea or warmth.

That's enough definitions.

The point of the first Class Balance post was that some people like a narrowly balanced game, and some would rather play a widely balanced game. Neither is "true D&D," but one or the other is almost certainly the D&D you prefer.

Whether you want widely balanced, narrowly balanced, or completely unbalanced classes, I think most people can agree on this: If class balance could be achieved in a way that did not impinge between players and their characters, that would be a good thing. That type of balance would help players and DMs who want it and remain invisible to those who don’t want it or who don't care.

Narrow, 4E-style balance is not invisible. The limits are always present, and certain players will forever strain against them.

Wide balance, on the other hand, can be so invisible as to seem nonexistent. That's what led to so many arguments about balance in AD&D. Some players felt it had no balance whatsoever while others argued that it was excellently balanced, but the effect was so subtle it was easy to miss, like leanness in wine or the humor in Howie Mandel's standup routine.

Can Wide and Narrow Balance Coexist at the Same Table?

They can, except … those who reject the notion of wide balance prima facie will never accept that the situation actually is balanced.

A problem with wide balance is that it's so easily abused. It allows for the creation of “dump abilities.” A dump ability is just like a dump statan ability your character has the potential to be good at, but since you don’t intend ever to use it, you invest no resources into it. Who cares if your Charisma modifier is -4 if you arrange things so someone else makes all the Charisma rolls?

Let's look at it in the context of our wide game example, where a character's four abilities add up to 9 points. The standard character spread is 5-3-1-0. The gap between 5 and 0 is substantial, but the character excels at one thing, is average at a second, is familiar with a third, and is incompetent at the fourth. If players are allowed to distribute 9 points however they want, it will take about 6 seconds before an optimizer presents the DM with a 9-0-0-0 character. His combat ability is off the chart, but everything else about the character is a disaster. In fact, we’ll call him the “ability disaster.”

If the game allows this, a player can feel completely justified in creating such a character. Why not? Most players know that the DM won’t let a bad skill roll bring the campaign to a screeching halt, but they suspect or know that the DM will let characters die in combat. Under those conditions, putting everything into guaranteeing that you won’t die in battle and relying on the kindness of strangers for everything else is a logical, if selfish, strategy.

This is the diametric opposite of the optimizers’ frequent complaint that characters who aren’t optimized for combat drag down the whole party, because someone else has to pick up their slack on the battlefield. In a balanced campaign, where characters face challenging episodes of combat, exploration, and social interaction, ability disasters like this character also drag down the party, because someone must pick up this character’s slack off the battlefield. If the other players don't do it (maybe they've all optimized for DPR, too, or they're sick of this player's single-minded self-centeredness), then it falls on the DM, who all too often winds up softening the game's noncombat biscuits so toothless characters can chew them.

What's the solution? Ultimately, I still think we had the right philosophy in 2nd Edition, even if our underlying math was off. Every character ability in the game, whether it relates to combat, exploration, or social interaction, is only as useful as the DM allows it to be. The most awesomely optimized slayer in the Ten Kingdoms is just another chump if the DM gives him nothing but kobolds to fight. All that excess power is wasted and he'll look like a fool every time he's challenged with witty repartee or a locked gate. The same thing happens to the cleric who poured all his points into battling the undead if the DM keeps the dead in their graves, and to the ranger who's all about fighting giants when the adventure trail leads to the land of pygmies. Even less fortunate is the rogue who's a master trap-spotter when the fighter finds a magic shield that detects traps automatically.

All of those conditions and a thousand more like them are beyond the game designers' power to control, no matter how symmetrically the numbers are polished in the rulebook. Again, that's not to say game designers should throw up their hands and declare surrender. Certain aspects of characters can and should be controlled. But the moment you place a limit on a number, you create a situation where the player who finds a way to exceed that limit gains an unfair advantage over everyone elsethe limit comes with built-in incentive to break it. Changing the limit only shifts the target without correcting the situation.

Either players accept absolute, narrow limits for the sake of well-tuned balance, or they accept that "balance" means characters are going to be approximately equivalent but never equal, like two shapes with roughly the same surface area but completely different angles and numbers of sides. And whichever way they choose, everyone needs to understand that the system's balance begins tilting out of whack the moment it collides with the "real world" of an ongoing campaign.

This square and triangle have equal areas. If they were characters, and their height, width,
and number of angles represented distinct abilities, would they be balanced?


  1. This is an excellent post and effectively conveys the message that to me 'balance' in the game is like the quest to find the holy grail, sure you can go on the quest, but you are very very very unlikely to find it.

    The game of D&D is not about balance and everyone being the same, it is about fun and having a good time with friends. I want my players to 'break the rules' and surprise me when I DM just as I want to break those same rules when I play, it is after all a game of imagination so lets imagine more and argue about rules a lot less.

    Great post! I am copying this to show others when the usual arguments pop up about balance (sigh)


  2. The best post I've ever read on the subject. Thanks Steve.

  3. Hear hear!

    This is why we still play 2e and why the 10th Age will forever be a 2e setting. Wide balance is what feels most like "home," probably because I grew up with 2e, but it also feels more realistic than the other narrowly-balanced games (which feel much more like games, as a consequence).

  4. I've always thought that class balance in D&D is a straw man, truth be told. To me, believing such balance is possible includes the incorrect assumption that one can predict all the possible outcomes/implications of class powers. The question of balance is a trick question if there ever was one!

  5. Excellent article. Do the rules allow for a creative player to turn the square on its tip?

    1. If they don't forbid it, you know someone's going to try.

  6. As a GM, my main goal in each session is to make sure every player has at least one opportunity to star, so that's how I build my encounters.

    1. As a DM, I once tried to plan encounters like that, but I realized that what I was a doing was, after a fashion, "stacking the deck" in favor of my players. To me, this is just as much a cardinal sin as stacking the deck against them.

  7. I don't like Wide game balance. It either makes players twiddle their thumbs when something is called for in which they have no ability, it depends on the DM giving every type of challenge equal spotlight and it leads to players optimizing the mechanics heavy parts of the game, usually combat, and then having softened non-combat biscuits where roleplaying replaces mechanics, often making the one character with the 5 in non-combat feel useless.

    I played a Thief for years in AD&D. I played a Bard in 3E. There is no point in having non-combat skills if the Fighter, Cleric and Wizard don't have them. Their players will not play second fiddle in non-combat encounters like the Thief has to do in combat.

  8. Why make a 9, 0, 0, 0? Isn't anything over 6 overkill according to the rules as written?

    1. Because some (not all) gamers love gaming the system. If a +6 to hit is standard, is +9 overkill? No, it means u r the roxxor, dood. Once that mindset shows up at the table, it tends to trigger a race for the biggest bonuses.