Thursday, June 4, 2009

Balanced Level Progression

A very impressive feature of fourth edition Dungeons & Dragons is that the level progression has been balanced so that, on average, the various capabilities of players and monsters go up at the same rate as levels are gained.

In previous editions of Dungeons & Dragons, the distinctions between the various characters and classes started out at a fairly modest level, but grew greater and greater at higher and higher levels. For instance, at low levels a fighter had only a small “to hit” advantage over a cleric with the same strength. But as the fighter gained levels, the fighters hit advantage would become greater and greater compared to the cleric. And previous to third edition, the fighter would gain multiple attacks at a certain point, and the cleric never would. Of course, this was balanced out by the fact that at low levels the cleric spells were not all that amazing compared to the natural fighting ability of the classes, whereas at high levels they grew disproportionately and became massively powerful. However, the point is that at low levels the cleric could be a "secondary fighter" who didn't fight as well as the fighter, but fought decently. On the other hand, the cleric could have some pretty useful spells, but couldn't really concentrate on being a major spellcaster, he had to do a lot of conventional fighting. But at higher levels, the cleric’s fighting ability would become less and less important, and the spell ability more and more important. In other words, the essential purpose and style of different classes and characters would vary continuously as levels were gained. In theory, there is nothing intrinsically wrong with this. In practice, however, it made the game very difficult to design and balance correctly. Typically there would be a “sweet spot” level where the relative capabilities of the 2 classes felt about right, and the farther you were above or below this level, the less well the game seemed to work.

This issue of gains in levels changing the essential balance of the game occurred in many other places as well. For instance, in third edition you earned skill points every level, and it was possible to put the maximum number of skill points into one skill, or to divide up your skill points more evenly. Dividing them up worked interestingly at low level, but at high level it would mean that you were so far behind someone who had maxed out the skill that the GM practically couldn't design an adventure which would challenge one skill level without being either impossible or trivial for the other skill level. And it meant that at low level, an ordinary character might notice a stealthy opponent, but at high level, a character class which didn’t specialize in perception was practically blind against a stealthy opponent.

Designing the hit bonuses and defenses of monsters was also tricky, since with a d20 system it is important for the values to fall within a narrow range in order to prevent the rolls from becoming trivial or impossible, but when the characters diverge at higher levels is soon becomes impossible for maintain this balance. You don’t want characters to hit on 2’s or only hit on 20’s, but when hit bonuses diverge tremendously you can’t cover everyone.

Note that the fact that fighters got more hit points per level than clerics was not an example of this sort of divergence. If a fighter starts with, say, 50% more hit points than the cleric, and gains 50% more hit points than the cleric does every level, the fighter will always have 50% more hit points than the cleric. So for game design purposes, the relative abilities of the fighter and the cleric in this regard stay constant, even though both are becoming more and more powerful. But if the fighter hits in melee combat 20% more often the cleric at low level, but at high level he hits 80% more often and gets twice as many attacks, then his power relative to the cleric has tripled at high levels, drastically changing the nature of the relationship between fighters and clerics.

Fourth edition Dungeons & Dragons tries to solve this problem by having the capabilities of the characters improve at a uniform rate such that the characters remain more or less in balance compared to each other. I think this works really really well and is a great idea. Classes of all kinds get new powers, feats, magic item bonuses, skill bonuses, and attack and defense bonuses at the same rate. And this rate is carefully calculated to be equal to the rate at which the monsters improve. Some of my previous articles have noted some small discrepancies in his calculations, but in effect these mistakes are the "exception that proves the rule". That is, the fact that small discrepancies are actually noticeable and worthy of correction shows the extent to which the game has been balanced; such small differences would be lost among the vast disparities in powers and abilities between the classes in previous editions.

Now, the relative stability in capabilities between the classes as levels progress is not going to be 100% perfect. Since players have flexibility in what feats, paragon paths, and epic destinies they take, and in what magic items they discover, the various characters are likely to become more different from each other at higher levels just because they have more choice available. But this is no big deal, it isn't a system of inevitably increasing disparity with level. Whenever I see an ability that changes disproportionately with level (that is, becomes less effective or more effective than other comparable abilities as the character level changes), I call that out as a deviation from the normal level progression balance. In most cases, such deviations detract from the game, as the ability will be "just right" at some levels, and too weak or too strong at others.

A good practical example of this is the way that attack and damage bonuses work. A bonus of +1 to hit is equally valuable at all levels, because the increase in monster defenses means that your base chance to hit monsters of your level should stay the same as you gain levels. But a bonus of +1 to damage becomes less useful at higher levels, because the total damage you are doing is higher and thus the additional one point of damage is a smaller percentage increase in your damage output. Therefore, in order for a damage bonus to not become ineffective as you gain levels, the damage bonus needs to increase with every level you gain. This is also true with hit point bonuses. But with attack and defense bonuses, they need to stay the same, otherwise they become overly effective as you gain levels.

This is why the bonus from weapon focus increases with every tier; it has to increase in order to stay equally effective relative to the character. The fact that the damage bonus from Dwarf Weapon Training is a fixed +2 bonus means that it does not retain its usefulness at high levels.. It means that dwarves are incredibly potent with axes and hammers at low level, but lose a lot of this advantage at high levels.

On the other hand, the fact that the new Weapon Expertise feats give you increasing hit bonuses at higher levels means they increase in power at a disproportionate rate. At low level they are very useful, at high levels they are absolutely incredible and indispensable. This would be broken if it wasn't clear that these feats were meant as mathematical fixes rather than legitimate optional feats.

The idea that attack bonuses should not increase with level and damage bonuses should is a corollary of the underlying game system and seems to have been fairly well followed by the designers of features and powers, although there are a few mistakes. For instance, the fact that the wizard’s orb feature gives an ever increasing penalty to saving throws as the wizards level increases, despite the fact a penalty to saving throws, like just about anything else rolled on the d20, does not need to and should not increase as you gain levels.

I'm not going into examples of this right now. But there is one major area in which the game's wonderfully balanced level progression is disrupted at a more fundamental level; I shall describe this next week.

1 comment: