Monday, April 19, 2010

Hybrid Classes

The idea I thought was really cool in the Player's Handbook 3 was the idea of hybrid classes. Basically, this allows 4th edition characters to imitate 1st edition dual-classed characters, where you choose two classes as part of character creation and are equally good at both classes. In 4th edition, each of the two classes is called a "hybrid" class. You get all of the powers are both hybrid classes, but each hybrid class has only a portion of the powers of a full class. Numeric values like hit points and defense bonuses are, in general, half as much as a full class, so the final character is more or less the average of the two classes. Hybrid characters have the same number of total powers, but must split the powers within each category between the two classes.

An additional interesting concept is that for strikers, and some defenders, they came up with the idea of tying certain class features to class powers, so that hybrid strikers can only use their primary "striker ability" to pump up the powers derived from that striker class.

Here are a list of things I thought were pretty cool about the hybrid classes:

1. The idea of constructing specific half classes for every class seems really clever for game balance purposes. Instead of having to make some sort of generic rules and hope that they aren't totally unbalanced when applied to specific classes, you can craft each hybrid class to be half as powerful as a full class, and design the powers in that hybrid class to work well when combined with another class, while still having the freedom to design the powers in the main class without worrying so much about whether they will be balanced in a multi-classed character.

2. It is clever that the construction of 4th edition D&D, based on a common level progression concept for every class and bonus powers based on class, really works well with the hybrid concept. Each class can just have half the bonuses of the full class, and the level progression bonuses are either owned by the character as a whole, or divvied up between the two classes in a natural way.

3. Tying class features to powers makes the design of hybrid classes much simpler. Instead of having to figure out how to cut a feature in half (and many features don't really have a natural way to be cut in half), you just give the full feature and restrict it to combining with half of the hybrid character's powers.

4. Tying class features to powers seems very entertaining if both hybrid classes have tied powers. In particular, I like the idea of a fighter/rogue who can choose each round whether to "lock down" the enemy with his fighter attack, or backstab him with his rogue attack. It adds an interesting tactical dimension to the simpler alternative of a character who tries to both backstab and lock down the enemy, both with half effect.

Things that are not so great:

1. Tying class features to powers is cool when both classes have a tied feature, but is awkward when one does and the other doesn't. If you have, say, a striker/leader, you have a striker ability that can only be used with your striker powers, and a once per encounter healing power that is not tied to your powers in any way. This is OK with your encounter and daily powers; you will get to use the striker ability on about half of these (more like 2/3 in practice if you put your "odd" picks in the striker class). But with the at-will powers, you have two at-wills you can pick from, but one has a big bonus the other one doesn't. So it is tempting to use that at-will power an awful lot and skip the other one, which makes the character less interesting rather than more interesting.

2. Not only is having only one class with a tied feature less interesting, it is unfortunately also more efficient. A striker/striker can use only one striker power at a time, but a striker/leader can use a full striker power every round (abeit with somewhat less flexibility), and a half leader power on top of that.

3. What to do with armor is, and always has been, a tricky question. I really don't see a simple, elegant way to do this in 4th edition rules. I guess my best idea is to average the number of armor feats each class gets, then let the character buy extra feats without the statistic prerequisites until he reaches the armor type of the better class. Anyway, what D&D4 chooses to do is to use the weaker armor type of the two classes, then allow the character to spend his one and only hybrid talent to get the better armor type instead. This actually seems like a pretty decent way to solve the problem given the fact that averaging armor types just doesn't fit cleanly into the system.

4. The way that classes are tied very tightly to certain statistics in D&D4 is not very friendly to hybrid classes. With single classes this is relatively harmless, since you just pick your class first, then take whatever statistics it requires. But an awful lot of hybrid class combinations that might sound cool require incompatible statistics and just aren't practical (especially at high level, when you are going to fall seriously behind if you try to advance more than 2 statistics).

Each hybrid class has a "half-powered" role feature. The role feature is what gives strikers extra damage, defenders the ability to mark and tie down opponents, and leaders healing. Note that, in many cases, half of a role feature is more than 50% as good.

To see this clearly, consider whether a hybrid character with two classes of the same role is better or worse than a single class character.

A hybrid leader/leader has a slightly better healing ability than a regular leader. This is because, although he gets the same 2 healings per encounter, he can choose to use both on the same round. However, at 16th level he becomes worse, as he does not gain a third healing.

A hybrid striker/striker would have a very useful flexibility advantage over a regular striker, in that he can choose the better striker ability for each situation. As many of the striker abilities are situational and don't always work, this is a significant benefit. A rogue/ranger, for instance, could use hunter's quarry/twin strike whenever he was unable to arrange a sneak attack. The hybrid character does lose the ability to gain striker damage on basic attacks, but that doesn't seem a big enough penalty to offset the flexibility advantage.

Controllers don't have "role features" in 4th edition D&D. Presumably, in order to be game balanced, they have better powers. So, in effect, a controller's role feature is always "baked" into his powers, just like the role feature of a hybrid striker. However, a controller/controller doesn't seem too exciting; after all, all controllers have two at-will powers from a substantial list, the only difference with a controller/controller is that the two powers come from different lists. This just isn't as exciting as having two large striker powers with very different activation criteria. This is a good time, though, to consider the case of a half-controller, half-something else. In this case the one controller power starts to seem pretty good. A controller/striker, for instance, really would have two very different at-will powers, each with a very different situation they are good in, and would thus seem better than either a controller or a striker (in terms of the role feature).

The fighter hybrid has a role feature tied to class powers, and thus works like the striker. The other defenders work in a variety of different ways that are hard to describe in a generic way; each power has to be considered separately whether it is half as good as the full power, or more. My impression is that most are more than half as good as the full power.

The upshot of all of this is that a hybrid striker class feature seems rather better than half a striker class feature, and the same is generally true with controller and defender, but not with a hybrid leader class feature, so hybrid leaders seem to need more other stuff to really be equal. I guess the one saving grace of a hybrid leader is the ability to use your one healing per encounter to revive the real leader; but it seems like there are a lot of other, easier ways to get this ability.

Looking at the game balance as a whole, the advantage of a hybrid character is having better role features that a standard character, and more flexibility in selecting powers. The numeric values will be essentially comparable, except for minor things like having to round down, and the fact that the hybrid characters have one less skill. The hybrid character may, or may not, have inferior armor, and may, or may not, have more difficulty choosing optimum statistics. The real balancing factor is that a regular character will generally have two or more strong class features which are better than most feats; the hybrid character generally has few or none of these class features, but can get one (and only one) by spending a feat. This seems to me like a good overall balance, steering in the conservative direction of making the hybrid characters a bit weak, to make sure they don't overshadow the traditional classes.

Then there is the question of the balance of specific hybrid classes, whether they are strong or weak, looked at purely by themselves. That sort of thing is always an interesting exercise for me. There are two ways to do this – analyze each hybrid class from the ground up, or compare it to half of the real class. The latter is far easier, so I will do that.

My impression is that the typical hybrid class keeps a half class feature which is better than half as good as a full class feature, and loses two strong class features and half of the remaining features (if any).

Avenger: Typical half striker feature. Loses two strong features, Avenger's Censure and Channel Divinity. Loses Armor of Faith as well, which leaves the Avenger with pretty poor armor no matter what he combos with, instead of his usual fairly strong armor. Seems harsh.

Barbarian: Typical half striker feature (baked into the at-will powers). Loses Feral Might, which is really two features in one – seems tempting to get it back with the Hybrid Talent. Loses Barbarian Armored Agility, but this may not be too bad if the character can end up qualifying to buy a feat to wear heavy armor. Keeps the Rampage feature. Seems favorable.

Bard: Half healing power. Loses strong features Bardic Training, Bardic Virtue. Also loses minor features Multiclass Versatility, Song of Rest, Words of Friendship. Does get one extra skill, just like a full bard, and keeps the little Skill Versatility feature. Seems unfavorable.

Cleric: Typical half leader feature. Keeps the strong Healer's Lore feature, but effectively only half of it, since it only applies to cleric powers in the first place. Loses the strong Channel Divinity feature and the ritual caster feat. Can only get back half of the Channel Divinity feature. Seems OK.

Druid: Loses 3 decent powers - a third at-will power, +1 speed, and ritual casting. Seems OK. The loss of the Primal Aspect may make Con-based hybrid Druids less practical.

Fighter: Has a striker-like half defender feature. Loses the strong Combat Superiority and Fighter Weapon Talent features. Doesn't lose a skill, like a normal fighter. Has very nice armor proficiencies which will probably be lost in most hybrid combinations. Seems OK.

Invoker: Loses the strong Channel Divinity power, plus Ritual Casting. Keeps the pretty good Covenant Manifestation power. Seems favorable.

Paladin: The Divine Challenge defender feature does about half damage or a little more, and is clearly rather better than half as good. While the damage is halved, it is every bit as difficult for the monster to avoid the damage. So either it inconveniences the monster just as much to have to attack the paladin, or the monster takes damage more often. The paladin loses two strong class features, Channel Divinity and Lay on Hands. These are pretty good, and the paladin is likely to lose his massive armor as well. On the plus side, the loss of Lay on Hands may allow the paladin to skip the Wisdom statistic entirely. Seems OK at best. On the other hand, the Divine Challenge is exceptionally easy to abuse with a ranged character, so who knows, maybe the hybrid Paladin is quite deadly.

Ranger: Standard half striker feature. The ranger has three mediocre class features – Prime Shot, a free feat, and a free skill. The hybrid keeps the free skill, and only loses two weak class features. The ranger seems like a pretty sweet hybrid.

Rogue: Standard striker half feature. Loses the strong Rogue Tactics and First Strike powers, plus the ability to fight well with daggers and shurikens. Seems OK.

Shaman: Standard leader half feature. Loses the strong Spirit Boon and opportunity attack powers. Keeps the Speak with Spirits power. Seems OK.

Sorceror: Standard striker half feature. Loses a couple nifty but not all that strong powers. Seems pretty favorable.

Warden: The half defender power affects one adjacent enemy instead of all of them – this certainly seems more than half as good. Loses the strong Font of Life power. Also loses Guardian Might, which is odd, as this is a small power combined with an AC feature needed to balance the class. So you have to find a way to get a good armor class, which may or may not be difficult, depending on what other class you choose. So it is hard to say how favorable the Warden is.

Warlock: Standard half striker feature. Loses the strong Pact Boon and Shadow Walk powers, and the Prime Shot power. Ouch. The poor warlock never seemed that great to begin with, and just gets to keep the weak striker power and loses all the cool warlock stuff. Seems weak.

Warlord: Standard half leader power. Loses the very strong command presence feature, but keeps the solid combat leader feature. Not sure if this is good or bad.

Wizard: Loses two lesser features – ritual casting and spellbook – and keeps Cantrips. Loses the one primary class feature, Arcane Implement Mastery, but can get it back. Seems pretty good.


  1. Hey man! I got into some financial difficulty and had to cancel The Internets, lend out the computer for awhile, but now I've got a shinier computer and internet. Whee!

    Good to see you're still chewing on stuff even without an audience. I'm really surprised your blog isn't more popular, as there are few that analyze from the angle you do; typically the analysis is from a player or GM perspective, moreso than the rules themselves. I guess most people prefer to theorycraft on a forum where they can try to outshout people they disagree with for several pages long.

    At any rate, I've got a huge backlog of reading it looks like, but real quick here's the rule hack I explain to players who are interested in hybrids.

    Like you said, there's definitely no elegant way to do it. I personally hate the idea of making a player spend their hybrid talent feat on something dull like armor proficiency, so I've tried to work up a suitable "averaging" of armor profs. I don't even mention it to players; if they ask about it, I tell them that they can take it, and then at paragon tier they can take a real hybrid talent feat, or vice versa.

    Averaging your armor:

    if both classes have shield prof, take the highest shield prof.
    if only one class has shield prof, you get prof with light shields.

    (If a class has chain prof or higher, I've waffled on whether or not to count hide as a step, since hide and chain seem to be treated as sidegrades which both follow from leather)
    the difference is only one step,
    use the lower armor prof.

    the difference is anything larger, then you are proficient with the lowest armor prof +1 step.
    You can take armor feats that would normally come free for one of your classes without meeting prereqs. Furthermore, one feat will buy you two armor profs.

  2. That is to say, I don't tell players that they can spend their hybrid talent feat on armor profs. I try to avoid that ever coming up.

  3. ...apparently I just needed to look back a little further to see the audience lurking in the shadows.

  4. Hey Gumby, good to see you back!

    Your idea for armor proficiency is pretty close to the idea I thought of. Mine was to count leather armor as 1 proficiency, hide or chain as 2, scale as 3, and plate as 4, then average the two classes and give them the armor proficiency corresponding to the averaged number, rounded down. You would only get hide proficiency if one of your two classes have hide. You can then buy more armor proficiencies without meeting prerequisites.

    This is pretty much the same as your system, except that in my system the average of cloth and plate is chain, not leather, and that if the two classes are exactly 3 steps apart in armor type (cloth/scale or leather/plate), it takes 2 feats to get the best armor type, rather than 1 in your system.