4 Ways D&D’s Creators Tried and Failed to Balance Classes

The classes in today’s Dungeons & Dragons game are balanced to make sure that when players leave a session, everyone feels like their character contributed to the party’s success. No player should ever see their character routinely upstaged and wonder, “Why am I even here?” In a list of goals for fifth edition, designer Mike Mearls wrote, “All of the classes should feel competent when compared to each other at all levels.”

The game’s designers didn’t always aim for this target, and when they did the methods often failed. What methods of class balance have the game’s designers abandoned?

1. Ineffective in one pillar, strong in another

The D&D game focuses on three pillars of play: exploration, roleplaying interaction, and combat.

In the early D&D game, players spent most of their game time immersed in exploration: mapping, searching, and evading hazards. Good play meant avoiding combat and saving spells. Expert play meant getting treasure without a fight. The original thieves lacked any combat assets—not even backstabbing—but during all the searching, scouting, and evading, only thieves brought any useful, reusable abilities. They shined in the exploration pillar, and floundered in combat.

In an interview for Drache issue 3, D&D co-creator Gary Gygax explained, “D&D’s team aspect is important. In a D&D game, each player has to use his strengths at the right place. Otherwise, the group won´t be able to survive.”

Some of that spirit remains, Mearls writes, “We’re OK with classes being better at specific things. Rogues are good at checks and handling traps. Fighters have the best AC and hit points. Clerics are the best healers and support casters. Wizards are the best at area attacks and control effects.”

But the game no longer allows classes that prove ineffective in a pillar. “If each class has wildly different combat abilities and the game doesn’t account for that, the system falls apart,” Mearls wrote. Over the years, the thief class added a backstab feature, which became sneak attack and a suite of combat abilities.

See The Thief’s Strange Trip from Non-Combatant to Battlefield Domination.

2. Weak at low levels, mighty at high levels

In D&D’s early days, Gygax saw characters who survived to high level as proof of a player’s skills. By this notion, players able to raise a weak character to the top deserved rewards. Tim Kask, the first editor of The Dragon magazine, echoed this perspective when he wrote, “Anyone that gets an Illusionist [to high level] deserves whatever they can achieve.”

No class showed this attitude more than the magic user. Originally, magic users started with the no armor, the lowest hit points, feeble attacks, and just one magic missile or sleep spell. But while a high-level fighter just added more hit points and a higher attack bonus, wizards gained power in 3 ways: They gained more spells per day, higher-level spells, and more damage with spells of a given level. Their power grew to overshadow the other classes.

“Earlier, D&D balanced wizards by making them weak at low level and powerful at high level,” wrote third-edition designer Jonathan Tweet. “But we tried to balance the classes at both low level and high level. (We failed. Spellcasters were still too good at high level.)”

The current edition starts to get the formula right. Mearls explained his goal for fifth edition. “Attaining balance is something that we must do to make D&D fit in with fantasy, myth, and legend. Even if a wizard unleashes every spell at his or her disposal at a fighter, the fighter absorbs the punishment, throws off the effects, and keeps on fighting.”

See How fifth edition keeps familiar spells and a Vancian feel without breaking D&D.

3. Higher-powered classes require more experience points

Before third edition, every D&D class had a different table of experience points required to level. As far as I know, Gygax never explained this quirk. No one asked because everyone just assumed the higher-powered classes demanded more experience points to level. The charts hint at some of this: The mighty paladin requires more experience than the weaker rogue. But for the original classes of fighter, cleric, and wizard the differences seem quirky rather than systematic. “The system sometimes gave clerics more hit points than fighters because a cleric would be higher level than a fighter with the same XP total.” Until double-digit levels, the XP requirements for a magic user never left the wizard more than a level or two behind the other classes.

4. Classes with level maximums

Originally, Gary Gygax gave little thought to high-level characters. Kask recalled, “We figured the odds of even getting to level 9 or 10 were so high that it wouldn’t pose a problem. This was before the gross inflation of XP’s and the corresponding levels. The highest-level player in Gary’s Greyhawk campaign was a 7 or possibly 8 at that time, and they had been playing more than any other group with the possible exception of Dave Arneson’s.”

After D&D’s release, TSR co-owner Brian Blume lobbied to include the monk class in the game’s upcoming Blackmoor supplement. Kask wrote, “Brian rationalized the nearly super abilities of the monk’s high levels with the argument that nobody, or damned few, would ever get that high. (This illustrates a certain naivete that all of us shared back then. We had no idea people would play almost daily and rack up XP’s at a truly unimagined rate.)”

Gygax published a class that imposed harsh limits to high-level monks. For monks “there is only one man at each level above 6th.” So to rise above 6th level, a monk character had to find the one monk of that level and win a fair fight. “There will always be a higher level to fight, even if there is no player character in the role.” The class topped out at 16th level.

A year after Blackmoor, gamers had completely disproved the theory that few characters would rise to high level. So Gygax returned to the monk class’s scheme for limiting the new Druid class in the Eldritch Wizardry supplement. Kask explained, “Every advance beyond level 11 meant fighting and defeating a fellow druid in either magical or physical combat—and the occasional 11th-level challenger of one’s own to deal with!”

In practice such limits only steered players away from choosing the classes they wanted to play, or blocked characters from advancing with their peers in a high-level party.

Next: Number 5.

13 thoughts on “4 Ways D&D’s Creators Tried and Failed to Balance Classes

  1. majorgs15

    There are a lot of “assumptions” made in this article (and in Mike Mearls’ opinions of what is/was needed in 5e – but hey, that’s his job”.

    I, for one, do NOT believe that all classes need to be balanced at all times. Not every character needs to contribute *THE SAME AMOUNT* to every session/adventure. If I build a table, I need a saw, hammer, screwdriver and tape measure. Each contributes something the others can’t (not just do BETTER than others), but if the “mission” off building the table is accomplished, all will have contributed to the “victory”.

    Furthermore, since the concept of roleplaying interaction is “class-agnostic”, even those classes not using their skills (at the moment), can contribute in interactions, and in exploration, even if to a LESSER (but not zero) extent in combat.

    And I don’t think that “in practice” the “king of the hill” concepts used for monks and druids dissuaded anyone from playing them – on the contrary, just like kids playing “king of the hill”, they held out the challenge of climbing to the top and staying there.

    And the fact that few characters might rise to high levels fits exactly with my readings of fantasy. Merlin was the only good wizard and Morgana the bad in the Arthurian stories, but there were a whole ROUND TABLE of fighters. Fafhrd and the Grey Mouser were but 2 thieves, but there were plenty of sell-swords, etc.

    I don’t know that having different XP thresholds for each class was important “back in the day”, but today, with the use of “milestone advancement”, it was a way to differentiate between how hard it was to rise in level depending on class and experience gained. Today, that doesn’t seem to matter – everyone advances together when the “railroad” module/campaign chapter says they need to in order to be able to weather then next set of encounters.

    Bottom line – I am not sure that the “old/original” way was as “bad” as the light in which folks like to cast it today. It was just different, and different for reasons that have as much validity as the current “every character can be a jack of all trades” class “balancing”. It would seem the need to “work together” to obtain the skills of the saw, hammer and tape measure, is not important any longer.

  2. ericscheid

    “No one asked because everyone just assumed the higher-powered classes demanded more experience points to level.”

    They did explain, in a manner, in the AD&D 2nd edition DMG, in the section Creating New Character Classes (Optional Rule):

    “To use this method, choose different abilities you want the class to have. You must include some aptitudes such as fighting. But other abilities, such as spellcasting, are optional. Each ability you choose has a multiple attached to it. As you select the abilities for your class and add the multiples together. After you have chosen all the abilities, multiply the base experience value (see Table 21 ) by this total. The result is the number of experience points your new class must earn to go up in levels.”

    So, yup, more or better capabilities translated into more XP required to level.

    1. Dan

      True, but they also admitted up front in that section that you wouldn’t be able to precisely reverse-engineer the original classes using those tables. Simple classes like fighters came close, but druids were way off, for example.

      1. ericscheid

        True, the optional rules as provided wouldn’t produce exactly the same XP tables.

        Also true though .. the provided rules demonstrated the general principle: more capability requires more XP.

  3. Dave Dalrymple

    Another point to consider is that AD&D classes weren’t necessarily intended to be equally attractive. In practice, most people I played AD&D with chose Fighters or Rangers. Wizards, thieves and clerics were much rarer. And I was the only weirdo I knew who played Bards in 2e. 🙂

    The result was a party makeup similar to the Heroes of the Lance: 60% Fighters.

  4. Full of Opinions about game design.

    Hilariously, I’ve found that 5e is no more meaningfully balanced than any of its predecessors in terms of actual play. Surely, it lacks the “everything is a save or die” mechanic of 1e, and the munchkin optimizer shenanigans of 3.5 (and to a lesser degree PF1e).

    But in terms of play at the table, the classes are simply not at all balanced against each other with minimal apparent designer intent. In glorious retribution for the dark ages of 3.5, the bard is somehow the best class in the game even before you get to the sillier colleges. I could go on about the failings of 5e, but they’re honestly rather irrelevant in most cases – casual players won’t notice or care, and more experienced players will either play a more developed system or hack it to the point that it’s barely recognizable for houserules, supplements, and homebrew.

    This comes from extensive play and GMing. Don’t get me wrong, 5e can be fun and is a great gateway into the tabletop community. I mostly don’t enjoy running it – the core mechanic is brilliant, but everything around it (classes, spells, monsters, etc.) feels underdeveloped and thrown together.

    I’d compare this to PF2e, which is a significantly less newcomer friendly system exceptional maturity of design – very taking the lessons of the past decade of gaming into account but still managing to create a very solid and exquisitely balanced system. That said, PF2e isn’t my favorite system to play, but it stands as a standard for what game balance without making everything equivalent can look like.

    TL;DR version: 5e is hardly balanced but still a good starting point for new players or resting point for casual players. Also you can hack 5e into a very serviceable game but that is honestly more work than just finding a game that works to begin with.

  5. Daniel Boggs

    >>>” Originally, magic users started with the no armor, the lowest hit points, feeble attacks, and just one magic missile or sleep spell.”

    Not originally. As published “originally” in 1974, the Magic-user and the Cleric both had the exact same 1d6 HP to start. The Fighter had a tiny advantage at 1d6+1. All three classes did exactly the same damage regardless of weapon – 1d6 HP.

    However Magic-users did not have Magic Missile spells because that spell was invented later. In fact Mu’s had no offensive spells at all until 5th level when they could learn Lightening and Fireball.

    But that’s a total red herring because Magic-users of all levels can use magic items and scrolls of any level spell. In OD&D Mu’s should be spending their money on spell scrolls to take with them on adventures. It was never the intent that they would rely entirely on memorized spells. Potions and scrolls are among the most common items in the treasure tables and a low level Mu heading into a dungeon and well equipped with such things at the start could reliably predict they would find more to replenish those burned up as they went along. The Mu’s real weakness was in lack of armor – the “price” for magic.

    Anyway, the point I think I’m trying to make here is that the original game was more equitable when it came to classes – much more equitable than is usually realized – and that’s one of the many reasons I prefer the original rules. That all changed when Gygax published Greyhawk in 1975, with new classes and new rules to “balance” them.

    1. sapphirecrook

      Scrolls are so underappreciated. Either to have a very circumstantial spell at the ready (resist energy, comprehend languages, etc.) or to always have an emergency casting of something good ready.

      I can’t believe how HARD it is to get spellcasters to even CONSIDER using scrolls or wands or anything. Yes they take extra actions to use; so make them non-combat spells. You’re not buying fancy weapons and armor, use that money.

      Recently convinced a cleric to carry scrolls for some of their less used spells instead of prepping it and never using it. Like I opened their eyes to a world never seen before. I guess people get too caught up in their class features and listings to even consider outside help…

  6. Corvus Corax

    I’m guessing Number 5 is Ability Score Prerequisites.


    Class Ability Minimums


    Class Str Dex Con Int Wis Cha

    Fighter 9 — — — — —

    Paladin* 12 — 9 — 13 17

    Ranger* 13 13 14 — 14 —

    Mage — — — 9 — —

    Specialist* Var Var Var Var Var Var

    Cleric — — — — 9 —

    Druid* — — — — 12 15

    Thief — 9 — — — —

    Bard* — 12 — 13 — 15

    1. sapphirecrook

      I highly doubt it. This doesn’t actually balance anything.

      Like, who wants to play a Wizard who can’t even cast spells? If it was intended to balance, instead of ‘hey you rolled X stats then Y classes are the best idea’, it does a weird job at it.

    2. Dave Dalrymple

      That would be number 6, I guess. But, yeah, the reason for the strict ability score minimums for the “advanced” classes was because the designers felt that these classes should be hard to qualify for. Nowadays, we see this as an arbitrary “luck gate”. But at the time, it seemed reasonably fair: a powerful character SHOULD be hard to get.

  7. Pingback: Paladins, Barbarians, and Other Classes Once Balanced by Rules of Behavior | DMDavid

Leave a Reply