Bad Balance: Races of Destiny
I’ve made fun of Dungeons & Dragons 3.5 a lot, and I do it because it’s easy, and fun, and it’s funny, and let’s do more of it.
There were two sets of the first wave of expansions for 3.5, the class style books, known as the Completes, and the race books, known as the Races of books. The Completes touched just a tiny bit on the early days of the class role system – with the first wave being Complete Warrior, Mage, Psionics, Divine and Scoundrel, but expanding into things like Adventurer and Champion. The Races Of books gave you a run down of three races at a time, linked by a common theme, and that was a theme that was routinely straining at the edges. Halflings got treated as a ‘race of the wild,’ and gnomes as a ‘race of stone,’ as opposed to their proper place they shared of ‘races of hobbits fans.’
No time for gnomes or halflings, me, I know.
Continue Reading →Bad Balance: Pissing Contest
Back in D&D 3.5 there was a really weird balance problem that existed within the confines of “two of the most powerful classes” that were nonetheless so different in power that one of them was definitely a weaker option.
Let’s talk about wizards and sorcerers.
Bad Balance: Incarnum And Cognitive Load
One of D&D 3.5’s biggest problems was its magical system, which was by default bonkers and broken. Despite that, though, there was an almost constant attempt to expand the magic system, to fulfill every specific small variant visions of magic. The Expanded Psionics Handbook, the Tome of Magic, the Artificer and the Archivist, wings of spells in The Book of Vile Darkness and The Book of Exalted Deeds – there was a near-constant effort to expand the magic systems to do different things in different ways.
You can approach design from either a strong mechanical position, or a strong theme and Incarnum was a power idea that came hard out of a mechanical interaction. Magic of Incarnum brought its own system, called the Incarnum system.
The Incarnum system, as simple as I can explain it, is that you could create virtual items in your item slots, and then invest a small pool of points into these items to make them better. This could give you special abilities that got better, from turn to turn, and you could rearrange all these points every turn. When you needed lots of defence, you could sink those points into defenses. When you needed to kill something fast, you could put those points into offense. This system was pretty interesting and cool! You could really customise a character in a lot of ways, and there was this balancing act of choosing where your points were by default.
Incarnum however, was a really bad system, not because it was bad, but because as a player, you had to spend the bulk of your time juggling a small list of points for a small advantage. None of the Incarnum values were particularly large, and the niche utility of some of the shifting was as much a matter of pooling skill bonuses into your armour at the right moment at the right time, rather than really changing what you did. The system was designed to be careful enough as to not get out of hand like the existing spell system (which was broken), but still be an alternative worth playing (which was pretty hard, when the spell system was broken).
The real thing though that kills Incarnum is cognitive load.
Cognitive load is the concept in psychology that describes the amount of active memory you have to track to keep a task executed. It’s how you concentrate on something, it’s the work required in your brain to manage the information presented to you organised. Incarnum was a system that started with cognitive load problems, and it got worse as you levelled up.
You might sometimes hear a player describe a game as ‘smooth’ or hear a designer say something is ‘frictionless.’ Mostly, that feeling is attained by making sure your design does little to demand cognitive load without a reason. It’s one thing to concentrate on a complicated turn or a crucial strategy, but you don’t want the everyday operations of play to require you to make a lot of complex planning and contend with juggling information.
There’s a reason designers preach the idea of simplicity. It’s not for its own sake, it’s because you want to make it as easy as possible for the players to make decisions about what they want to do in your game, rather than have to do math on working out how what they want to do can work.
Bad Balance: The Problem With The Tome Of Battle (Which Isn’t What You Think)
Right at the tail end of 3.5 D&D, there was a book released that ruled.
There’s a lot of critical talk about the Book of Nine Swords. There were comparisons to Anime, as if that was inherently a dismissal point, as if Anime wasn’t regularly cribbing from D&D in the first place.
The Tome of Battle presented a solution to the problem of melee combatants in 3.5. As you levelled up, melee combat just didn’t keep pace with the kind of things spells could do. Spellcasters even in the early game had an edge on the melee characters, and increasingly, the game became about countering spellcasters rather than countering melee characters. You can view late-game combat as about trying to shut down the Wizard long enough that the Paladin could get some licks in.
But in Tome of Battle, melee weapon-wielders and armour-wearers got to stab things in the face real good. They touched on the core idea of 4ed, which is time spent in a turn is actually more valuable than hypothetical infinite options. It was a great book, created great characters, had a wonderfully varied lore you could use a little or a lot from, and mostly didn’t have total turkey prestige classes (as most books did).
Yet at the same time I am comfortable and confident declaring that Tome Of Battle is, as it stands, a mistake. Not because of anything the game did intrinsically, but because the book was released into a world with poor Tordek here.
The Fighter in 3.0 D&D was a really rough sell. The fighter at level 1 was already comparable to another class’ class features, and their design scaled up very linearly. The best levels of Fighter were 1 and 2, because the class was frontloaded enough to let you rush up to some sort of mid-tier trick slightly early (like Whirlwind Attack, the game’s idea of an ‘End Game’ Fighter Feat). The next best one was level 4, because at that point you had unlocked access to everything the Fighter could get access to and you only had to take one level that Didn’t Really Do Anything to get there.
There were a lot of things wrong with how they updated it for 3.5, one of which is the removal of the ‘Fighter’ subtype from the way they presented feats in the Players’ Handbook. Back in 3.0, feats that the Fighter could take with their bonus feats had the subtype [Fighter]. In 3.5, this was removed in favour of the new line
Special: A fighter may select [this feat name] as one of his fighter bonus feats.
This was in my mind a blatant mistake. They could have made it so that Fighter feats had riders or bonuses if they could check the number of Fighter feats you had, but only if Fighter feats had a subtype.
Anyway, the thing is, the fighter was pretty weak and attempts to fix the fighter had almost all missed, usually because they approached the problem in the totally wrong way – Fighters got better on a linear, additive scale, while spellcasters and things balanced against spellcasters got better on a quadratic scale. A level 20 fighter could do a decent bit of damage to a dragon in one full round of combat, assuming he could close – but the wizard could disintegrate the dragon or take over its will entirely.
But if you wanted to play a character in armour, with a weapon, whose primary interaction with enemies was hitting them with the weapon in increasingly skillful ways, though, the obvious look for most players was that the Fighter was your jam. And the fighter, as a character class, was made totally unnecessary by the Tome of Battle’s inclusion. In some cases laughably so – the Warblade got some bonus feats, got Weapon Specialisation (formerly a Fighter-only benefit) and got all those maneuvers that let the Warblade hang at the level of the Wizard.
The Tome of Battle classes didn’t wholesale replace all the fighter-style classes. Paladins and Crusaders compared to one another. The Sword Sage could do a lot of interesting tricks, but the Monk could still do other things the Sword Sage couldn’t.
But the poor Fighter?
The closest thing the Fighter could do is spend their bonus feats to buy Maneuvers. Once you had Warblades there really was no reason to play Fighters beyond their simplicity. They didn’t quite measure up to threats the game thought of as reasonable, but you could cover that with your party. They didn’t give you the kind of options they thought they did, but you might never notice that either.
But when the Warblade sat down next to you, did most of what you did, and a host of exciting, additional extremely extra things as well, it was hard to not notice.
Bad Balance: Why Balance?
Dungeons and Dragons 3.5 was absolute nonsense balance-wise, but it was remarkable because it was imbalanced in a whole variety of different ways that are good object lessons for designers to take on board when making your own RPG content. So, rather than one huge master-post explaining it, here’s one example:
Why Balance?
Inevitably, in this kind of conversation, someone will sit back, stroke their chin and say Yes Well, But Why Do We Need Balance Anyway? and then deliver a smug, eye-raised look as if they’ve just next-levelled the entire conversation. And sure, they’re kind of right – you don’t need things to be fair. You don’t. Heck, you don’t even need a rules system, man.
Setting aside the obvious Let’s Take It Too Far And Show The Point Is Ridiculous, the question of balancing 3.5 D&D seems to always bring up someone who, weirdly, echoes the arguments of 2ed D&D before them. It usually comes down to it’s a cooperative game and there aren’t any winners or losers and therefore, balance is a phantom that need not be pursued, as if somehow, the game’s function is inhibited by balance, and that balanced things impose themselves between fun and the players.
Problem is, someone has to run the game.
Balanced characters aren’t balanced against one another to meet some arbitary philosophical goal. They’re balanced against one another so the person running the game has a reasonable, handleable piece of information about things they can present to the party. Big deal, the response comes: the DM can just tailor-make the experience for the party, every time. All they need to do is know what those characters can do, how good they are at it, how resistant they are, any rare abilities they might not use often, and how well they hit things, what they attack.
Which is to say: You don’t need balance. You don’t! Knock yourself out. You do you.
But there’s a virtue to having characters with roughly comparable ranges of power and utility, because it means that the people running the game aren’t faced with an enormous different task of challenge construction. That means that when someone wants to run a D&D game, they’re not presented with a cliff face of learning. It means that there are going to be fewer situations where the players try a thing they’re meant to succeed at and fail because it so happens that the enemy’s abilities fall into a venn diagram of specialised, obtuse weaknesses.
If you’re playing D&D 3.5, a game primarily designed around tactical movement and combat, and you think it doesn’t matter that it has a decent tactical movement and combat system that is reliably testable, provable, and functional, then maybe you want to play a different game? There are a lot of great games out there!
That’s the next lesson: Balance can be valuable for games, even cooperative ones, because it allows the person or people making the challenge to construct meaningful challenges.
Bad Balance: Your Part In Failure
Dungeons and Dragons 3.5 was absolute nonsense balance-wise, but it was remarkable because it was imbalanced in a whole variety of different ways that are good object lessons for designers to take on board when making your own RPG content. So, rather than one huge master-post explaining it, here’s one example:
Your Failure
You’ll find if you listen to any given D&D 3.5 player, they’ll usually have some memories of the things I talk about being total bupkis. I know I played alongside a cleric who wasn’t overpowered, and we had one game where the runaway behemoth was a telepath. As your friendly neighborhood min-maxer I had the game squealing under the heel of a bard, once. More often than anything else we’d see on the newsgroups players wondering about how they could play clerics well, because they thought their only job was standing by and healing, leading to an unfulfilling game of whack-a-mole. What’s more there are a lot of games where the wizard player felt worthless and ran away from goblins a lot with a terrible armour class. Once I heard the artificer dismissed as trash because a player could simply not imagine how to make it work.
This is one of the many ways D&D3.5 was unbalanced: It was entirely possible to play overpowered characters badly. Most of the characters who were busted were busted because of spells or magic items and that stuff was overwhelmingly available…
If you took it.
You could absolutely play a weak wizard! You could pick up the twenty totally worthless spells at every level, you could sink into the swamp of crap. You could take a level of sorcerer and a level of wizard, and then maybe level them up side by side and maybe you’d balance your stats and oh good god noooo.
You could be handed a high-octane chainsaw laser hammer and it was entirely reasonable for a new player, a player who had no reason to expect they were being given something totally broken, to sit down and tap nails in with the wrong end.
Bad Balance: Free Power
Dungeons and Dragons 3.5 was absolute nonsense balance-wise, but it was remarkable because it was imbalanced in a whole variety of different ways that are good object lessons. So, rather than one huge master-post explaining it, here’s one example:
Free Power!
A thing you’ll find in most games is there’s an opportunity cost to adding things to your character. Magic items occupy this space in D&D where there are slots, clearly recognising that there’s a good reason to limit the number of belts you wear, especially when those belts do magical things. Thing is, the item system isn’t the only place that came up.
In most games there’s an opportunity cost. Every choice you make is an option. In 3.5 D&D there were a surprising number of times when there were no such choices. If you were aiming for a prestige class at level 6 onwards, your first five levels could sometimes look like utter nonsense – fighter 2, barbarian 2, ranger 1, for example, would give you a grotesque fort save, a handful of benefits and lose you a single point of reflex and will, which was just not a reasonable trade. If you were building a wizard, prestige classes themselves could look ridiculous, as you cherry-picked the opening benefits of four or five of them because none of them had a meaningful late game reward.
When you give players an option for something, you need to make it so that they’re giving something up if they take it. Not that every option is punishing – that’s its own bad idea – but that every option is a choice, and choices should be meaningful.
Bad Balance: Paralysing Potential
Dungeons and Dragons 3.5 was absolute nonsense balance-wise, but it was remarkable because it was imbalanced in a whole variety of different ways that are good object lessons for designers to take on board when making your own RPG content. So, rather than one huge master-post explaining it, here’s one example:
Paralysing Potential
Did you play a spellcaster back in 3.5? Did you play a Cleric? A Druid? A Wizard? God help you, did you play an Archivist?
The 3.5 spell list is an absolute swamp of bullshit, a completely festering mire of options that include procedurally generated X damage over Y area with Z range math-up messes that really form the basis for what you can probably handle, balance wise, to spells which are unimpressive with one basic form, and totally busted if you think about them innovatively, spells designed to be worse versions of the former, multiple spells designed to fill the same space by designers working on different books, grandfathered together, spells designed to duplicate other spells just with a different flavour to try and keep the spell schools reasonably balanced, then some complete out-of-context nonsense that didn’t have any combat-or-existing mechanical application but suddenly changed the context of how combat even happened. Feel tired at the end of that sentence? Good, because it’s worse than that.
Spellcasters 3.5 were broken and it was easy to get a modest amount of broken just by paying attention to a few exploitable spells, but if you wanted to go deep, if you were the kind of player who was willing to marinate deep in the dank shit of supplemental sourcebooks or even just read through the ramifications of everything in the player’s handbook, if you were the person who bothered to use Scribe Scroll and stockpile every level 1 spell you didn’t wind up using in any given day until you had literally a library of the dang things, then you knew how broad, how busted, and how blinding your potential was.
There’s no surprise that players – despite the weakness – really appreciated Sorcerers. All sorcerers needed to know was a small handful of useful spells, rather than try to learn all of the spells present in the entire danging game.
The way I coped with it, myself, was to opt away from the full-bore spellcasters. My few times playing Cleric or Druid were times that DMs quickly started nerfing things on the fly. If you instead limited yourself to doing a smaller handful of things really well, you might be less powerful, but you’ll at least be able to make a choice on your turn without ever being stymied by the thousands of things you could do. Levelling up and building your character was still a long, agonising process, but at least that concentrated the wait.
These older editions would sometimes present you with complicated puzzles in the form of a combat arena and expect you to answer that question with a spell template; a line, a point, a single target, a circular template or maybe a chain. But then, the Wizard didn’t have to do that, they could answer it by teleporting away, by charming something, by becoming invisible, by summoning thirty tons of stone directly above the enemy, by becoming something else, by making someone else into something else, and all of these options were presented to you.
Now it’s your turn.
Pick something.
Go go go.
Bad Balance: Accidentally Overpowered
Dungeons and Dragons 3.5 was absolute nonsense balance-wise, but it was remarkable because it was imbalanced in a whole variety of different ways that are good object lessons. So, rather than one huge master-post explaining it, here’s one example:
The Accidentally Overpowered
If you played D&D3.5 and you were into it, you knew you could spend hours working on a character, combing through dozens of races and classes and prestige classes, and even if you weren’t going to scrape through everything to milk every single possible advantage, it was still not uncommon to see builds like Wizard 5/Really Strong Wizard Prestige Class 10/Archmage 5.
That, however, is something you work for, something that comes at the result of a lot of system mastery, and, speaking as someone who had a lot of that mastery and really loved it, it was nice. It was good. I liked that. I don’t even feel that that was, itself a problem (though we’ll get to that).
The problem that cropped up in 3.5, and you could see it happen in people’s stories of their games and problems adjudicating how characters work, in that some characters could, completely accidentally, be utterly honking busted, and those characters could be wildly out of whack with their friends and party members without ever trying to be. If your party featured a composition of a fairly boring Fighter, Cleric, Rogue and Sorcerer, one of those four characters is absolutely in a league of their own compared to the others, and as players play, it’s going to come out and get noticed.
Or what about an arrangement like Ranger, Druid, Barbarian and Wizard? The barbarian is better than Ranger, the wizard is better than both and the Druid has a single class feature that’s better than the Ranger.
The issue isn’t that a player could work to be broken, the issue is that players could accidentally be so good that other people would completely replace another character in the party, and if anyone noticed, it was just straight up feels-bad territory. It was really obvious – and it got worse as people levelled up and you had more room, more opportunities to make mistakes! What makes it even worse still is if players went and did stuff that was redundant! Fighter, Ranger and Barbarian were all the most easily grasped, handleable classes, and if a party has two of them, one of them was going to notice the gap!
Well, if they looked.
So you had to hope they didn’t look.
That’s not a good solution to the problem.
Bad Balance: The Impossible Spectre of Balance in 3.5 D&D
I recently went back to some old content I made for 3.5 D&D and found myself considering that the flavour, the tone, the purpose were all sound –
A quick aside.
When I say the flavour, I mean the way the game objects are designed to represent things in the universe; a ranged attack that deals a decent chunk of damage and requires an action to refresh could be easily flavoured as a gun;
when I say the tone, I mean the kind of other things in the universe that are necessary for the thing to exist; guns don’t work in a setting without advanced metallurgy, for example, but they also don’t work in a setting where you want fights to be back-and-forth exchanges of force;
When I say purpose, I mean what mechanical end I want this object to fulfill in the world; this gun may work as a way to give players with less physical stats a meaningful ranged attack and to show this region as being more focused on distributable technology than on magical advancement
– but that without a lot of refamiliarising myself with the rules I could not say for sure how balanced they were, or were not, in a D&D 3.5 or Pathfinder game. I went back to read the Tome of Battle and Tome of Magic, two books I love but which have
Let’s say problems.
3.5 D&D was a game with a fantasy of balance. It had a lot of systems for creating ballparks, and if you bothered to explore all those ballparks you could wind up finding one where all your players could play together. You had to avoid the situation where one player was playing a totally different sport in a different field, but it wasn’t like you were being fundamentally reasonably by limiting sources. The whole problem of the CoDZilla (“Cleric Or Druid”) of 3.5 was that in the core book alone they were still totally broken and other sources only made them moreso.
There were other systems totally weirded up; like the Sunder mechanic was either useless or amazing, and its side effect of destroying treasure was either dreadful or meaningless. The trip and grapple systems could be pushed to breaking, the summoning system had its narrow holes, and every single expansion or splatbook you can find only adds either new options that are too weak to make any difference, or totally new broken things.
This is the conundrum of 3.5: Nothing is balanced, but things have gravity. Things suck together, and you can find a balancing point acreted around one general family of busted stuff. This is something I really found comforting about it in hindsight, but is also a trap: If players were not in a position, skill-wise, to pull towards those same common spots, if they were drawn towards other thematic thing, that player was set up to have a miserable time.
So what’s the solution?
My gut is to make it so the broken options are easy to get. To allow for elegant, simple power. Make the four-prestige class stack-em-ups a bother to get. Make small rules tweaks that keep those kind of complicated builds being total upgrades, but don’t try and push players away from the powerful toys that are cool.
Towards the last of my 3.5 days, the builds I looked for to make were, as an ideal, as few classes and prestige classes as possible; as a designer, if someone brought me a character splashing a single level of four or five classes on the way to a prestige class, I was left considering that the jankiness was a problem. When your build was full of different stuff that you picked up because there was no investment to do so, it meant your play experience slowed down (hang on, do I have a thing to do with that?), and it also meant the costs for joining a prestige class or taking a level in another class were too low.
Overall this is heartening though: I don’t think I can make anything that’s too overpowered, especially ‘overpowered’ is a moving target.