Failures of D&D 3.x

I think I’ll start my ‘Thing a Day’ series with a description of why and how I think D&D fails. This is likely to be a work in progress, the points below are more or less off the top of my head.

I’ve been playing RPGs for almost thirty years, a wide range of them, and D&D has always been my go-to game. Sometimes this was because it was the easiest to find players for, but for the last ten years or so it’s been because D&D 3e (and D&D 3.5 even more, after that) did the best job of modeling the kind of world I wanted to explore.

So, D&D has always been my preferred game to play, and I think D&D 3.5 is the best version of it I’ve played. However, it has some flaws and limitations in how it works and what it models.

Quadratic Growth vs. Linear Growth

Let’s start with perhaps the biggest offender, since it is related to (and in fact direct cause of) several of the following issues.

Mundane characters (that is, those without the ability to cast spells or use other supernatural abilities) grow in a more or less linear fashion. Ignoring the maximum hit points gained at first level, their hit points go up in a linear manner, their attack bonuses go up in a linear manner, and many of their special abilities, such as they are, have linear gains at linear cost (class levels or feat slots). When improving or gaining new abilities, they tend to either pay to keep an existing ability (nominally) at a level-appropriate grade, or pay full price to gain a new ability at low level.

Fantastic characters (those with the ability to cast spells or use other supernatural abilities) grow in a quadratic manner, if not exponential. At every other level, a primary caster gets new abilities (spells) that are at a level-appropriate grade. A ninth-level wizard doesn’t need to spend four feats to get access to teleport, nor does a ninth-level cleric need to spend feats to get access to raise dead (though it does have a cost to cast). The characters can simply choose to learn or prepare these spells as needed.

For this privilege their linear growth (Base Attack Bonus, hit points, etc.) may be limited compared to the mundane characters. Over the course of ten levels, the wizard’s Base Attack Bonus falls behind the fighter’s Base Attack Bonus by five whole points. For reference, this is almost the same as the difference between rolling Strength 10 and rolling Strength 18. The wizard can work around this a little bit, though. For instance, a wizard could cast spells that use touch attacks (which are roughly the same in effect as the brilliant energy weapon quality – a +4 weapon quality – in that they ignore armor) or saving throws (that can be almost impossible to avoid because of how the saving throw DC calculations work).

Doesn’t seem quite fair, does it?

Multiclassing Doesn’t Work

On the face of it, 3e and 3.5 had the best multiclassing of all editions of D&D.

The earliest editions didn’t allow multiclassing at all. AD&D had two different types of multiclassing (‘actual’ multiclassing as done by demihumans, who were stuck with whatever selection of classes they made for their multiclassing, and ‘dual classing’ for humans where you could switch between classes but not back, and only as long as you followed certain rules about class abilities and had good enough ability scores). In short, as long as you could live without heavy armor and could find a DM willing to overlook the demihuman level caps (or played only at low level where they didn’t come up anyway) you wanted to have multiclass elves.

Starting with D&D 3e, you could freely and easily multiclass. Your fighter starts dabbling with magic? Give him a level of wizard and off he goes. If he later gets religion he might take a level of cleric, or if he’s larcenous, rogue (or maybe still cleric; we hear stories about some priests…). The linear bits still fit together more or less well – the oddity you get with ‘partial scores’ is easily corrected by tracking and applying the fractional values rather than the integers shown in the class progression tables.

However, this gets absolutely hosed by the linear vs. quadratic problem. Yes, a 12th-level fighter could decide to take his next two character levels as a wizard. He trades +1 Base Attack Bonus, a combat feat, +1 Fortitude (which is having trouble keeping up to save DCs already) and six hit points for the ability to cast magic missile for 1d4+1 points of damage a few times a day and +3 to his Will save.

Yay? He trades something that might be worth having (if it weren’t rendered irrelevant by his lack of spellcasting ability) for something that frankly isn’t worth having.

Let’s look at the other side of this. The fighter’s buddy, a 12th-level wizard, decides to get a little more butch and take a couple levels of fighter. He gains +1 Base Attack Bonus, +3 Fortitude, six hit points, and two combat feats. Mind you, it cost him only +1 Will base save bonus and… seventh-level spells?

Back that up, I think he’d rather be able to, say, travel to another plane (plane shift), safely teleport almost anywhere, or change the weather for miles around himself. He’d be trading something that certainly would be worth having for something he can easily do without.

Even when you don’t get gross disparities like this, many other classes multiclass poorly because the level-appropriate abilities they grant are based on class level. This means that in order to get what ‘good stuff’ the class offers you have to stay single class. The limitations put on the class abilities work okay when a character single-classes, but break when a character multiclasses.

Level Adjustments Don’t Work

If you want to play a cool race, something with some nifty racial abilities, you pay extra for the privilege. Specifically, you pick up a level adjustment that causes you to be treated as higher level than your Hit Dice indicate.

One reason this fails is that you often end up weaker than your ‘adjusted level’ indicates. You might have some shiny abilities in limited circumstances, but you lose some key durability and strength you should otherwise have for your level. A 2HD creature with a +4 level adjustment is not a sixth-level character. He’s simply too fragile to survive, and unless the special abilities gained directly apply to the situation may well be unable to act effectively at all.

This fails also because the ability you pay precious level adjustment for might be rendered meaningless as you gain levels. For instance, DR 20/magic sounds awfully impressive until you realize you probably can’t get it until a level such that most things you’re likely to fight have magic weapons or DR/magic themselves (either of which means your DR doesn’t help at all). Similarly, many racial abilities might be unusually effective in some limited circumstances when you first pay for them, but don’t scale well and become irrelevant.

Magic Trumps Mundane

This might be a personal peeve and it’s not actually a problem. It seems it’s very easy to find magic that makes actual skill use irrelevant. Whether it’s the knock spell, fly, invisibility, or fireball, it seems it is often possible for spellcasters to render other characters meaningless. The other characters can still be meatshields, of course, but this is not a very satisfying career.


  1. Joseph Charpak

    Nice post, pretty much sums up the problem. Wizards should have been harder to kill than they are in early levels, and should have been easier to kill than they are in later levels.

  2. SS

    Nice summary. The “Magic trumps mundane” was fun to read as it basically confirms my rant on about (high level) mages being the do-all-don’t-need-help WMD iWin-button class.

    Give me a ranger any day. Straight forward (kinda boring) but _honest_ class.

  3. hadsil

    A main fighter who multiclasses a couple of levels in wizard/sorcerer does not/should not take Magic Missile. Shield is good choice for when using a two-handed weapon. Grease is useful. Charm Person can help. Silent Image is only limited by your imagination. Expeditious Retreat is very useful when you fight in platemail. (Get a wand of it.) When you take into account those spells, then multiclassing wizard/sorcerer is more appealing.

    Granted, a wizard really shouldn’t just multiclass a level or two in fighter. If a player really wants to be a warrior-mage, otherwise known as a gish build, he does need more thought. Prestige Classes help, such as Spellsword and Abjurant Champion. Alternatively, the player could just play a Duskblade.

    • Yes, there are some utility things that are… useful. That’s what ‘utility’ means, after all. However, many of these things can be had for chump change by this point and are probably not worth investing a character level in.

      The relevant prestige classes are an attempt to patch this failure, and are somewhat successful, but I think that that you have to plan for this sort of thing or you fall into a trap is not a good selling point. Similarly, I’d rather see it work out of the box than depend on a class like Duskblade that is specific to the character concept. Most of the other core classes are relatively generic (and the ones that aren’t I’d be inclined to make prestige classes anyway), so having something so specific to get around the design problem doesn’t set well with me.

  4. hadsil

    Whether it’s worth multiclassing a level or two for the spells is up to the player to decide. The player may like the added ability of using wands and staves. The +2 boost in Will save is nice too for a one level dip, +3 if two levels. It costs a BAB and a few average hit points. What you get is a lot more than what a feat could provide, which is its own problem getting back to your linear/quadratic point.

    Prestige Classes as a patch to multiclassing is a point that multiclassing has problems, but that the patch exists is a good thing. Prestige Classes are best used as that patch or to specialize in one particular area for a single class character. Prestige Classes become a problem when the DM allows a player to multiclass Prestige Classes.

  5. The caster-multiclassing point is something which Trailblazer attempts to remedy, and does so fairly successfully. It introduces a “Base Magic Bonus” which acts as caster level, which even fighter-types get at a rate of 1/3. As a result, a Fighter 6 / Wizard 3 has a BMB of +5, which is sufficient for 3rd-level spells, which are still relevant at that level. We’ve seen a ton of rogue/wizards and rogue/sorcerers in TB, all of whom were quite effective, as well as some terrifying monk / buffer clerics (monks get 1/2 BMB, so a monk 2n / cleric n gets nth level spells). You still sacrifice some utility by multiclassing into and among caster classes, but the gains are much more significant. Trailblazer also uses this as a fix for Theurge-type classes; a cleric n / wizard n has access to both nth-level cleric and nth-level wizard spells. However, they come from one pool of slots and ‘ready’ spells per day, which makes this significantly less powerful than a Theurge; compared to a monoclass wizard, you sacrifice some power for versatility by doing this.

  6. David

    My fix was fairly straightforward. Best of all worlds, approximately. Classes beyond the first (favored for humans) cost 1000 more per level than if you had started with them, but otherwise advance as if from 1st level. This quickly explodes certain traits, but the fix for that is simple, too. Compare total XP spent to character level values to determine number of HD, Saves, BAB, skills, i.e. the generic qualities, always favoring the character. Prestige classes were always a full character level. Add remaining XP to next class type if some remains.

    Example: CL 8. Total XP: 30,000. Arrangement: 5th level Barbarian (10000 XP-1st class), 4rd level Druid (10000 XP), 4th level Rogue (10000 XP).
    So HD are:
    10000 XP –> 5d12 from Barbarian +
    10000 XP –> d8 from Druid (with 5000 XP remaining going to Rogue) +
    5000 +10000 XP –> 2d6 from Rogue + Druid excess (with 2000 XP floating)

    Total 5d12 + d8 + 2d6
    Similarly for BAB, Saves, Etc.

    Kept the generic stuff in tow, and allowed the multi-class folks to become useful pretty quick but never overshadow the specialists in their own field.

    It did make some of the prestige classes redundant, though.

  7. Someone Else

    I think the issue with linear vs quadratic growth is primarily a shift in gameplay styles rather than mechanics.

    In the past, if you went wizard, you were taking a massive risk in the hopes of it paying off later. Basically, you had to survive in order to become powerful, and that was very challenging. Or you could go martial, less powerful later but you were far more likely to actually reach “later.” So it was a risk/reward trade off.

    Gameplay style changed though, and now players basically expect to survive unless the GM is a dirtbag. And that takes away the risk, leaving just the reward.

    • Yep. There was always a disparity between martials and casters. In AD&D there were two disparities (fighters were a little more durable than wizards — more because of armor than hit point totals, nobody had lots and lots of hit points — at low levels, wizards were devastating at higher levels). In 3e they filed off the risks of low-level casters, as you say, but kept the good parts at higher levels (and made them much better, even — more spells, faster spells, save DCs that outpaced save bonuses by a significant margin).

      • Someone Else

        It wasn’t the 3e rules that reduced the risk. It was the shift in popular play style. How people run the game can have an even larger effect than the rules themselves and this is a fine example. You could easily run 3e just as dangerous and lethal as earlier editions, but no one did. Heck, a ton of 3e rules were ignored or overlooked by the community. I still haven’t found a single gm willing to use the witch example class change from the core dmg, and no one I know uses a wide variety of encounter difficulties as directed in the dmg.

        The community took the rules but literally played contrary to the system’s design, with results such as the 15-minute workday and other “design problems.”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to Top