Which Multiclassing Power Paradigm is Best? ToB? 4E? Else?

[RadiantPhoenix]

Bullshit. I want as many good, functioning, supported character types as the game designer can write. If you write content and then take it out because you believe I cannot handle it, you are:

1) wrong
2) calling me a liar
3) an arrogant shitheel
4) unable to produce game design ideas that won't be poisoned by your egotism and low opinion of your audience.

Having attempted to design Type-1 MtG decks digitally, I call bullshit on your bullshit. Deciding which of fifty-or-so different cards to run, and how many of each, is hard! (The only certainty is that I'm going to be running at least one basic land of each of the deck's colors; wouldn't want to get Path-to-Exiled and not be able to find something useful...)

[Username17]

Note that Lago originally believed his game would have ten kits per class. When I pointed out that this wasn't feasible he cast around for some new justification for why there would be more kits than he believed a player could handle. He hasn't put any thought into how many options he could actually write or test; his goal is to take away options for its own sake. Which is really fucking sad, because if his goal was "design a good game" rather than "dumb a game down" he might have encountered actual reasons to restrict kits to primary class only. (For example, if everyone with a rogue kit has to have the primary rogue track that includes Sneak Attack, then he could write kits that interact with Sneak Attack.)

I'm not Lago and i think his triple classing thing is stupid, but we already covered this exact problem when talking about FFXI style subclassing. Let's put some numbers to it:

• Imagine that we had 11 classes that could be taken as a main class or as a subclass, and that everyone chooses one of each that have to be different. That's 110 class combinations to choose from.
  
  Now let's add kits. If we have just two kits to choose from for the main class, that's 220 class combinations. If we add two kits to choose from but instead you pick them for your subclass, that's still 220 class combinations. If we pick from two kits for the main class and two kits for the subclass, that's 440 combinations.
  
  If instead of picking from two kits for each, we just picked from four kits for the main class and none for the subclass, that would still be 440 combinations.

At any finite level of complexity, adding a pick selection step for the subclasses is a bad deal. You can always get more apparent variety by adding more kits to select from for the main class than you can by adding more kits to select from for your subclass. Well, not always, at the point that you already have more kits for each class than there are subclasses to pick from, there is eventually a break point. But that's an edge case that is so far away that I don't consider it meaningfully possible. Even 4e does not have have more kits for any class than there are are classes.

So when you say:

Or they could, you know, ask someone else at the table "what's a good kit for X?"

It's much, much easier for a player to pare options down than to add new ones. If you offer lots of options people who want fewer can just not look at some of them or ask someone else for help. If you offer few options, people who want more don't have anywhere to go.

You're wrong. Like, mathematically wrong. And it was already shown exactly why in this thread. And I just showed it to you again. This isn't arrogance, this is simple math: the feature you are asking for is provably more trouble than it is worth.

-Username17

[virgil]

Bah, I lost track in the threads, but is there a problem with having a list of main classes that you add almost like descriptors as you level? Example, start as a level 1 Wizard, then at level 3 you get Fighter. At that moment (level 3), you'd be mechanically the same as if you started as a Fighter at 1 and added Wizard at 3.

[ModelCitizen]

Or they could, you know, ask someone else at the table "what's a good kit for X?"

It's much, much easier for a player to pare options down than to add new ones. If you offer lots of options people who want fewer can just not look at some of them or ask someone else for help. If you offer few options, people who want more don't have anywhere to go.

You're wrong. Like, mathematically wrong. And it was already shown exactly why in this thread. And I just showed it to you again. This isn't arrogance, this is simple math: the feature you are asking for is provably more trouble than it is worth.

-Username17

See, I said earlier that this feature might be a bad idea for reasons not related to the audience being too stupid for it. Read Lago's post about the triclassing thing again. He literally wants to make this feature work, then take it out. It is phenomenally nonsensical.

I don't think tying kits to any of the three classes would work either. Forget balancing them, just writing 40-50 kits that each have to work on three distinct and non-intersecting ability sets sounds like a godawful clusterfuck. Lago could have figured this out on his own if he wasn't so busy sucking his own dick about how he knows what people want better than they do.

[Lago PARANOIA]

And there it is again. You think you know what I want better than I do. Until you get over that you will have nothing to offer to anyone.

Oy vey.

When any game designer is designing anything they have to choose what to include and what not to include. They make this decision, assuming that there's no conflict between profit and enjoyment, on the basis on what they think that people will enjoy. When I leave the Use Rope skill out of the game going from 3rd Edition to 4th Edition D&D that is me making a decision that people will enjoy the game more with it not in. If I put Use Rope into the game going from 2nd Edition to 3rd Edition D&D that is me making a decision based on what I think people will enjoy. If I leave Use Rope in the game going from 3rd Edition to 3.5E that's me making a decision based on what I think people will enjoy.

Accusing me of arrogance on this front is just baffling. You can accuse me of not including enough choices or that the offered subset of choices were poor. But telling me that I'm being arrogant because I'm thinking of an upper limit of choices to provide to the game? Please.

Note that Lago originally believed his game would have ten kits per class.

Here is what I said:

I was thinking about 4 kits per class (sort of like the expansion options that 4E D&D characters get) and 12 classes per game.

Were you confused by me mentioning that testing 30 combinations is more difficult than 10? That wasn't a suggestion, that's a hypothetical. If you're doing a triple-classing system and class combination has ten kits, if you restrict kit selection to the first class they'll have 10 kits to look through. If you restrict kit selection to any unique classes a player has they'll have to look through thirty.

Now I also said that some players will swap around what their first class is based on kit selection. So if someone doesn't mind whether they're a Fighter/Wizard or a Wizard/Fighter even though they only have a menu of (assuming we're still sticking with 10 kits) 10 kits they'll actually look through 20 of them. I think that is a problem because it obviates choice winnowing.

[*] This is why I suggested like 4 kits, but that may be a bit too many. Maybe more like 2.

[*] When I say kit I'm thinking more of a small power upgrade equivalent to a 4E Wizard deciding they want the Orb of Imposition or Staff of Defense class feature. It should be just a bit of lagniappe to get people jazzed and should rarely be a deciding feature.

[*] To assist with the above point, kits should not have stupid generalist bullshit titles like Great-Weapon Fighter or Staff Wizard or Graduate of the White Lotus school. They should have specific, roleplay evocative titles to them, that way newbies and basketweaves can quickly churn through choices. Pit Fighter is fine. Charmmaster is also fine. So is Thug. Or Poison Druid. Min-maxxers might look past the name to see what the bonuses actually do, but they're the kinds of people who love sorting and option paralysis and dumpster diving anyway.

I don't think tying kits to any of the three classes would work either. Forget balancing them, just writing 40-50 kits that each have to work on three distinct and non-intersecting ability sets sounds like a godawful clusterfuck.

Having kits be class-agnostic is a defensible design decision and I could see myself supporting that. You would have to move away from kits/titles like 'Poison Druid' and go more towards 'Gladiator' or 'Field Marshal'. You still can't have kits like 'Crowd Controller' or 'Magic Item Lord' or 'Student of the Zodiac'.

Regardless there is still some finite number of kits you can include in the basic game before people start derping out over too much choice. Even if this wasn't true, which it is, you still have a limited amount of space and playtesting complexity.

[Lago PARANOIA]

I'm not Lago and i think his triple classing thing is stupid,

Butt... butt why, Frank?

[Username17]

I'm not Lago and i think his triple classing thing is stupid,

Butt... butt why, Frank?

I can see mandatory double classing, like the previously mentioned FFXI system. Characters have two shticks each and so on. But if you get more than that, what the fuck are you?

If it's important to you that the character be a hammer swinging, lockpicking, caster, you should probably just go Paladin/Thief, rather than Fighter/Magic-User/Thief. Because the whole point of character classes (as opposed to just buying unique abilities with points) is that the player characters can be quickly summed up and explained to the MC and the other players. You tell them that you're a front liner who augments that position with disposable necromancy pets (Hero/Necromancer) or that you are a nuker who root-kites (Wizard/Druid) or whatever. Having three classes to report makes that sentence unmanageable. Figuring out how the pieces fit together into a tactical unit is not merely a proverbial three-body-problem, it's an actual three body problem. And that means that after telling them what your classes are, you're going to have to explain how it works. Which means that the classes are providing essentially no information to the other players at all.

I understand it when people want tier based class replacement, but I think that works better for a system where your "class" is a template you can trade out at other times as well.

The 3e class system was attractively packaged, but was basically a bad idea. By the time people figured out how to make the characters they actually wanted, they had lost all the advantages of a class based system and only had a fraction of the advantages of a point based system. Saddling people with more than two classes at a time is just asking for confusion.

-Username17

[Lago PARANOIA]

I'm aware of that effect, which is why I had the ratios be 50/35/15 instead of 33/33/33. I wanted it to be intentionally lopsided to reflect concepts that only have a tiny bit of another class's ability and also to provide an avenue for someone to pick up a schtick mid-adventure without fully immersing themselves in it the way dual-classing does--even though it's not a thing I totally agree with. The third class is effectively lagniappe instead of a schtick that's supposed to get screentime. IOW A Fighter / Wizard / Cleric has more difference from a Fighter / Cleric / Wizard than a Fighter / Wizard / Psion.

Now the reason why I still want that rather than just having the ability for people to put some points into some 'free' slot (for example having 85% of your character come from the class system, the rest from some ability pool system) is because I want the class system to be completely closed and inviolate. It's just been my experience from watching 3E and 4E D&D that allowing interactions with the class and level system that are not completely internal to the class and level system never lead anywhere good. Having that backdoor into the game invites abuse.

[ModelCitizen]

And there it is again. You think you know what I want better than I do. Until you get over that you will have nothing to offer to anyone.

Oy vey.

When any game designer is designing anything they have to choose what to include and what not to include. They make this decision, assuming that there's no conflict between profit and enjoyment, on the basis on what they think that people will enjoy. When I leave the Use Rope skill out of the game going from 3rd Edition to 4th Edition D&D that is me making a decision that people will enjoy the game more with it not in. If I put Use Rope into the game going from 2nd Edition to 3rd Edition D&D that is me making a decision based on what I think people will enjoy. If I leave Use Rope in the game going from 3rd Edition to 3.5E that's me making a decision based on what I think people will enjoy.

Accusing me of arrogance on this front is just baffling. You can accuse me of not including enough choices or that the offered subset of choices were poor. But telling me that I'm being arrogant because I'm thinking of an upper limit of choices to provide to the game? Please.

Your Use Rope thing is a false analogy. Removing Use Rope is a popular choice but the reasoning has nothing to do with OGOD TOO MANY OPTIONS HALP. But anyway, here's your original position on the triclassing thing:

[*] At first level, upon entering the second third of the game, and upon entering the last third of the game you get to select from a Kit, a Prestige Class, and an Epic Destiny that conforms to your Main Class. There's no reason why it can't belong to any of your classes (and in fact is a recommended house rule that will be in effect for more sperg-tastic games like the RPGA), it's just a way to cut down on option paralysis.

There are two problems with this:

1) You believe there is no reason the kits can't work with all three classes. I think you'd run into a lot of problems once you started writing it, but whatever. No one expects an off-the-cuff post to be a flawless design document.

2) In the hypothetical fantasy world where the kits work fine with all three classes, you then want to tell people arbitrarily that they're not allowed to do it. Your stated reason is, paraphrased, that they're not smart enough to do it. That's fucking insulting.


Furthermore it makes me wonder why anyone should read anything you write about game design. Reviewing from the "what people want" thread, you believe that you know what will make people happy better than they do. No one can argue against that on an intellectual level because there's no reason to believe you won't assume they're just too stupid to know what they want. And there's no reason for anyone to consider any of your ideas either. Once you believe that you know what your audience needs better than they do, the "correct" action is to trick them into accepting design they don't want. Anything you write must be assumed to be written in bad faith.

Carried to its logical conclusion your position is so arrogant and condescending that it removes you from the conversation. It makes you unable to talk with your "audience" in good faith and therefore unable to produce good game design ideas. So again, you need to drop this shit.

[Username17]

[*] At first level, upon entering the second third of the game, and upon entering the last third of the game you get to select from a Kit, a Prestige Class, and an Epic Destiny that conforms to your Main Class. There's no reason why it can't belong to any of your classes (and in fact is a recommended house rule that will be in effect for more sperg-tastic games like the RPGA), it's just a way to cut down on option paralysis.

2) In the hypothetical fantasy world where the kits work fine with all three classes, you then want to tell people arbitrarily that they're not allowed to do it. Your stated reason is, paraphrased, that they're not smart enough to do it. That's fucking insulting.

No, it is not insulting. You're being retarded. It has already been shown why you are wrong. Let's go through this again: for the same increase in complexity of letting a player select a kit for their primary, secondary, and tertiary class, you could cube the number of kits for each class. Not double or triple or anything shitty like that, but literally raise to the power of three. As in, instead of opening the floodgates for a selection from one of four kits for each of three classes, you could let each player choose just one kit for their primary class and write sixty four fucking kits for each class! That would be the same complexity level, and it's obviously too much. I don't want to work through the implications of 64 Paladin kits before completing my character, and neither do you.

Letting people pick kits from three different classes at once is obviously a shitty use of complexity limits, because it has explosively large numbers of things to test and it's still boring because there really aren't that many entries to read, just a lot of possibilities to procedurally generate and test.

You really need to get over being offended at the "condescension" of people wanting to protect you from literally hundreds or thousands of uninterestingly but game mechanically importantly different options to sort through before your character is finished. Because that shit is a really stupid fucking waste of chargen and playtest time.

Picking one kit for your main class and having 6 or 8 or even 12 kits to choose from would be way better than picking from 4 kits three times and would be several times less complexity to worry about. Iterative selection from short lists is lazy design. It requires a very small number of entries (short writing time) and takes fucking forever to playtest (so obviously you just don't playtest it). You are seriously sitting there being offended that someone wants to write more and make a product that people can playtest and use. That is insane.

-Username17

[Username17]

I'm aware of that effect, which is why I had the ratios be 50/35/15 instead of 33/33/33. I wanted it to be intentionally lopsided to reflect concepts that only have a tiny bit of another class's ability and also to provide an avenue for someone to pick up a schtick mid-adventure without fully immersing themselves in it the way dual-classing does--even though it's not a thing I totally agree with. The third class is effectively lagniappe instead of a schtick that's supposed to get screentime. IOW A Fighter / Wizard / Cleric has more difference from a Fighter / Cleric / Wizard than a Fighter / Wizard / Psion.

Now the reason why I still want that rather than just having the ability for people to put some points into some 'free' slot (for example having 85% of your character come from the class system, the rest from some ability pool system) is because I want the class system to be completely closed and inviolate. It's just been my experience from watching 3E and 4E D&D that allowing interactions with the class and level system that are not completely internal to the class and level system never lead anywhere good. Having that backdoor into the game invites abuse.

Can't be done. If your character's girlfriend is a saucy scullery maid that is basically a disadvantage because your character doesn't "get" anything for that, but the fact that you have a girlfriend who can be kidnapped and you aren't romantically available to the Elf Queen might come back and bite you. On the flip side, if your girlfriend actually is the Elf Queen, that presents a number of obvious advantages. And you can't readily tie either into the class and level system without making an Elothar Warrior of Bladereach, and doing that destroys the campaign's ability to generate new stories.

I understand the impetus to want to make every single thing your character can do be something that is derivable from their class and level, but you can't succeed. And history has shown that marching down that road leads to people trying and failing to make social advancement points and shit.

-Username17

[ModelCitizen]

Picking one kit for your main class and having 6 or 8 or even 12 kits to choose from would be way better than picking from 4 kits three times and would be several times less complexity to worry about. Iterative selection from short lists is lazy design. It requires a very small number of entries (short writing time) and takes fucking forever to playtest (so obviously you just don't playtest it). You are seriously sitting there being offended that someone wants to write more and make a product that people can playtest and use. That is insane.

-Username17

You are still not understanding what Lago actually said. He stated that there is no reason to restrict kits to primary class other than "option paralysis." You're arguing a very different position than he is.

[Username17]

You are still not understanding what Lago actually said. He stated that there is no reason to restrict kits to primary class other than "option paralysis." You're arguing a very different position than he is.

No. We are not. Option Paralysis is the thing that happens when your complexity gets too high. You have a complexity budget before your game becomes unplayable as a game and people either shut down or start making choices at random.

Now we can argue about what that complexity threshold actually is. But it exists. It's a thing. And Lago was saying that letting people select kits for secondary and tertiary kits contributed too much towards complexity limits. After you got offended by this blatantly obvious observation, I pointed out that it increased complexity by an order of magnitude or more for a benefit that was actually very small.

So whatever you think the complexity limits of your target audience actually are, it is extremely obvious that kit selection for secondary and tertiary classes is well beyond them. However many kits you think players can handle, cubing the things to select from is probably unwise. There are much better things you can introduce into character generation that contribute to option paralysis less and contribute to the feeling of character concept breadth more.

Lago's statement here is basically trivially true. And you flipping out about it is a failure of analysis on your part.

-Username17

[ModelCitizen]

Jesus fucking christ.

Your position, as far as I can tell: The feature would produce too many options to write or test, and also too many options to manage in play.

Lago's position: Writing and testing the feature would not be difficult. However, players would not be able to play with it.


You see how these positions are fundamentally different?

[hogarth]

No. We are not. Option Paralysis is the thing that happens when your complexity gets too high. You have a complexity budget before your game becomes unplayable as a game and people either shut down or start making choices at random.

3.5 D&D is a playable game. So if you're saying that you should probably make your game less complicated than 3.5 D&D, that is not a difficult requirement to pass!

[Username17]

Jesus fucking christ.

Your position, as far as I can tell: The feature would produce too many options to write or test, and also too many options to manage in play.

Lago's position: Writing and testing the feature would not be difficult. However, players would not be able to play with it.


You see how these positions are fundamentally different?

No. He pointed out that writing such a feature would be easy. Indeed it would. You just write however many kits you were going to write and then let people select three of them instead of one. Actually playtesting it would be essentially impossible.


Making iterated choices that can be mix-n-matched makes writing a very large number of options go faster. You write X options for column A and Y options for column B and you have X*Y options. So you write X+Y things and get X*Y options. For any X and Y greater than 2, that's a huge increase. It makes playtesting hard. First because you have X*Y things to test, and secondly because the playtester has to procedurally generate every actual option before testing it - so if there are infinite loops (or things just as bad) hidden in there they will indeed remain hidden.

I think Lago's suggestion is bad. But he is objectively not pissing in your Cheerios the way you are accusing him of doing and I don't know what your fucking problem is.

-Username17

[ModelCitizen]

[*] At first level, upon entering the second third of the game, and upon entering the last third of the game you get to select from a Kit, a Prestige Class, and an Epic Destiny that conforms to your Main Class. There's no reason why it can't belong to any of your classes (and in fact is a recommended house rule that will be in effect for more sperg-tastic games like the RPGA), it's just a way to cut down on option paralysis.

"There's no reason why it can't belong to any of your classes." So when you say that there are reasons to restrict the kit to primary class (such as that the additional options would be impossible to playtest), that means your position is fundamentally incompatible with his.

As for why this is relevant: Lago's position rates his own ability to manage a large list of options much higher than that of his audience. That's part of what makes it insulting.

[Username17]

[*] At first level, upon entering the second third of the game, and upon entering the last third of the game you get to select from a Kit, a Prestige Class, and an Epic Destiny that conforms to your Main Class. There's no reason why it can't belong to any of your classes (and in fact is a recommended house rule that will be in effect for more sperg-tastic games like the RPGA), it's just a way to cut down on option paralysis.

"There's no reason why it can't belong to any of your classes." So when you say that there are reasons to restrict the kit to primary class (such as that the additional options would be impossible to playtest), that means your position is fundamentally incompatible with his.

As for why this is relevant: Lago's position rates his own ability to manage a large list of options much higher than that of his audience. That's part of what makes it insulting.

Are you stupid? He said that it was perfectly possible to select a kit from anything, but that it wasn't set up that way for reasons of option paralysis. That is in fact exactly what I said, too. It's a piece of shorthand about something that is demonstrably mathematically true. Reducing the number of kits you select from reduces complexity thresholds by huge amounts for what is essentially a fairly limited trade-off in apparent customization. Seriously: get off your high horse about this. Lago's statement is actually mathematically derivable from simple axioms, you're basically frothing about how P=/= NP is a condescending elitist imposition from mathematical know-it-alls who can't even talk to normal people because they are too full of themselves.

You. Are. Wrong.
Lago. Is. Right.

At least, about this minor facet that you have chosen to attack him on. His tri-class juggling act is still a bad idea, but not for any of the reasons you have ranted about.

-Username17

[Seerow]

Where was Lago's actual tri-classing system discussed? Because from the bits I've seen here, it sounds a lot like what Legend does (ie each class has 3 tracks, you can pick all 3 tracks and be single classed, or mix and match the tracks to be multiclassed).

[ModelCitizen]

Frank, go to my post six posts up from yours. Read it again. Then read back down to last post. If you still don't understand that you and Lago are arguing incompatible positions, go back and do it again. And again, if necessary. Keep doing that until you get it. I'm sure you'll figure it out eventually.

[Lago PARANOIA]

I understand the triple-classing complaints and I agree that there's really nothing you can do about it.

Having a main/sub class account for 85% of your character and choosing from a 'free' bank for the rest of it would do pretty much the same thing. I still hold a white-knuckled fear over that because I can easily see people doing things like making Power or Class Feature X mandatory for certain builds or combos. Making people go through the threshold of having to actually declare that package as a class curbs abuse because there's still a segment of the population who will balk at saying that their class is a Wizard but not if their Fighter got kicked out of the Magic Academy soon after learning Magic Missile. Also giving abilities out as a package instead of a discrete selection reduces the average number of internal combos someone can make, both because people will write new abilities faster than new classes and to increase the chance of giving someone a schtick that's balanced but not combo-able.

Why am I against dual-classing? I'm not against it, but I do think that it's inelegant when trying to model characters who are competent in a schtick but not defined by it. Zuko is a really good swordfighter but he loses more from not being able to firebend than not being able to swordfight. Quite a few characters are defined by such a relationship. The Grey Mouser might be able to do a trick or two with magic but isn't defined by it. So is Tidus being able to throw out some time-related magic in addition to being a Sports Ace. Or Dr. Doom being able to summon a demon now and then. Or Vincent Valentine being able to transform into a B-movie monster. So on and so forth.

Now the problem goes away the more classes you publish for the game. When your game gets to be 60 classes you could probably find some sort of dual-class combo that represents all of the above character facets. But I'd rather be able to have that option in the first book.

[Juton]

Guys, I'd actually like it if something useful came out of this thread. If Model Citizen/Lago/Frank feel it's absolutely necessary to continue ranting about whatever, can the do so in another thread pls?

[Lago PARANOIA]

Don't blame me, Juton; I've already made my peace with ModelCitizen. I'm ready to talk multiclassing.

[OgreBattle]

How do you handle the Paladin and Swordmage in this multiclassing system

Should the Paladin be its own class that multiclasses with others, or should Fighter/Priest be the way to construct a Paladin (and Priest/Fighter as cleric?)
What are the advantages and disadvantages of each?

If the goal is more simplicity it seems that having no Paladin, but Fighters and Priests, would be the better choice as the Paladin is a natural result of main/sub multiclassing.

[Juton]

Also, should classes be engineered to be equal at every level, or would it be acceptable to have some classes be more powerful at a certain level, weighed against them having a higher point cost? Having a working system where one level of Bard doesn't cost the same as one level of Rogue can allow for a lot of interesting possibilities. If you can make it work, which could be a big if.

If you where going to make a new edition of D&D would it be better to make every level of every class equal or would it be OK to make some classes mature in power at different rates and price accordingly?