D&D 5E Multiclass in 5E

[Greg K]

I take it from this that d20M (odern?) has a cleaner version of the original? How does it work that makes it cleaner?

Base Classes: Base classes are based around one of the 6 ability scores (Strong Hero (Str), Fast Hero (Dex), Tough Hero (Con), Smart Hero (Int), Dedicated Hero (Wis), Charismatic Hero (Cha)). Base Class Levels go to level 10 and have multiple talent trees and a list of associated bonus feats (all characters also gain feats at level 1,2,3 and every additional 3 levels based on character level. These feats are not limited to the occupation feats). At odd levels a class grants a talent from one of its trees. At even levels, it grants a feat from its list. Despite the association, it is not necessary to put your highest ability core into the class associated with it, but you, probably, want to put something with a bonus.

By limiting a base class to 10th level, it is assumed that you are going to multi-class into an advanced class (a specialized experienced adventuring class that has some prescribed level abilities) or between multiple base classes (and taking their talents and feats rather than set class features).

The trees are
Strong Hero:
Extreme Effort (feats of strength)
Ignore Hardness
Melee Smash (increases melee damage)

Fast Hero:
Defensive Talent Tree (Evasion, Uncanny Dodge (2 levels), Defensive Roll, Opportunist (spend an action point to make an AoA against someone that has just been struck for damage)
Increased Speed Tree (3 levels): Each level increases base speed by 5 feet.

Tough Hero:
Damage Reduction Tree three levels starting at DR 1. Each time you take this DR increases by 1
Energy Resistance Tree: Choose one of the energy types listed and you ignore your con modifier in damage from that type. Each time you take this, you can add a new energy type.
Unbreakable Tree (Remain Concious, Robust (extra hit points), Second Wind (spend action point to recover hit points = to Con Mod , Stamina (increased healing))

Smart Hero:
Research Tree (Savant (add smart hero level to one skill from a list), Linguist (chance to understand encountered languages))
Stragegy Tree ( Exploit Weakness (use Smart Bonus instead of Strength or Dex on a DC15 Int check), Plan (when has time to plan, grants a bonus to an attack roll or skill check), Trick (opponent makes a save or is dazed for one round)

Dedicated Hero:
Empathetic Tree (Empathy, Improved Aid Another, Intuition)
Healing Tree (3 levels)
Insightful Tree (Skill Emphasis, Aware (add base Will bonus to Listen or spot to avoid surprise), Faith (spend action point to improve the result of an attack roll, skill check, saving throw, or ability check), Cool under Pressure (choose 3+wis modifier number of skills. The character can take 10 with them even when distracted or under stress)


Occupations: There are occupations which are the pre-cursor to 4e backgrounds. At first level , every character receives an occupation which makes at least two skills to the list of class skills and some other benefits (wealth bonus, additional feat, and or one or more additional skills). Say you take the Criminal occupation, you could apply it to any of the base classes. A strong hero with a Criminal background might be an enforcer. A fast hero might with Criminal be a burglar or pick pocket. A tough hero with Criminal might a gang member. An intelligent hero with Criminal might be a mastermind. A charismatic hero with Criminal might be a grifter.

Advanced classes : These are ten level classes that as mentioned represent experienced adventuring specialists. They can be entered as early as third level. However, Stan! points out in the Modern Player's Companion that many careers don't need to be represented by advanced classes (see below):

Examples: Soldier (someone trained to be be equally good in both melee and ranged), Martial Artist ( a melee combat) specialist), Gunslinger (a ranged combat specialist), Infiltrator (a cat burglar or spy), Daredevil (stuntment, Xtreme sports athtletes). Body Guard, Field Scientist (Indiana Jones, Steve Irwin, a storm chaser, etc.), Techie, Field Medic, Investigator (investigative reporters and journalists, detectives), Personality (TV star, Movie stars, Rock Stars, Politicians), Negotiator

There are multiple ways to enter Advanced Classes, but the fastest path for each of those listed above are each tied to one of the six base classes (there are two for each base class)


Advanced Class vs. Class Combos: As I mentioned above, many careers don't need to be advanced classes and can be handled by class combos (multiclassing between base classes). According to Stan! The Advanced Classes for a specific interpretation of a person with a specialty. If the vision of the advanced class does not fit the character and his variation of a type of specialist or a specialty doesn't exist as an advanced class, It is better to multiclass with base classes using talent and skill choices to model the character. Here are the examples Stan! provides for the use of Class combos in the Modern Player's Companion without his sample progression

Archaeologist: Smart/Dedicated/Charismatic
Bouncer: Tough/Charismatic/Strong
Counselor: Dedicated/Smart/Charismatic
Entrepreneur: Charismatic/Dedicated/Smart
Lay Clergy: Smart/Charismatic/Dedicated
Pro Athlete: Strong/Fast/Tough
Reporter: Dedicated/Charismatic
Rock Star: Charismatic/Dedicated/Tough

Fx users (Mages, Acolytes and Psis)
FX users are setting specific advanced classes. Prior to third level, a character with psionic power can take Wild Talent Feat which grants a 0-level psionic power useable 3/day. Surprisingly, there is no feat for casting 0-level arcane or divine spells in the core book although it would be easy to adapt the Wild Talent Feat. The supplement Urban Arcana, however, introduced the following for first level character
Divine Heritage: Prerequisite Wis 10. Choose 3 0-level Divine Spells. you can cast each 1/day. Must be taken at first level
Magical Heritage: Prerequisite Int 10. Choose 3 0-level Arcane spells. You can cast each 1/day. Must be taken at first level

Therefore, until 3rd level, FX users are going to only have access to 0-level spells.


For myself, I like the use of Talent Trees and Occupations rather than a lot of prescribed abilities. Advanced Classes are nice, but not truly necessary. I also like that spell casting is even weaker than D&D and characters have to start with their 0-level spells and have to wait to acquire their first level spells (then again, my house ruled 3e has spellcasters using the bards progression.

 

[Crazy Jerome]

I'm not pretending to have yet absorbed all of that sufficiently to comment on it intelligently, but I gather a D&D (partial) analog would be something like having only 10 levels for Fighter, Thief, Wizard, Cleric perhaps, then with advanced classes as Paladin, Ranger, etc? Or does the structure pretty much require you to have the more generic base classes, in order to work?

If the latter, I'm seeing that as basically going down one of two roads:

• Turning the classes into a thin, generic shell to continue to have "classes" in what has become a point-buy system in all but name.
• Effectively collapsing the ability score dimension into the class dimension of design, with effects I'm not sure what.

That is, I see how what you've written answers my question about a clean way to multiclass, but I don't see how it translates to D&D assumptions, without a bunch of (potentially contentious) side effects.

 

[Crazy Jerome]

Brainstorming and running off on tangents from too many recent posts to quote, trying to reverse-engineer from desirable outcomes:

1. Take Stalker's requirement as a starting place - no XP requirement to handle leveling. This assumes something at least passingly 3E-like in progression, with that kind of stacking.

2. Grant that not all class levels are equally valuable to a given character past first. (That is, multiclassing into 1 level of wizard doesn't buy a 10th level cleric very much compared to the 11th level of cleric.)

3. Assume for the sake of argument, however, that given levels of single-class characters are close to equal. A 10th level wizard taking another level of wizard is staying mostly in sync with a 10th level cleric taking another level of cleric.

4. Assume that there is some expected ratio for class effectiveness over levels, such that a character taking N levels is X better than before the levels were taken. (For ease of consideration and historical reasons, think roughly twice as powerful every two levels increase, though the ratio could be any number of things. This is not unlike 3E CR.)

5. Thus if a 10th level cleric gains two new levels, he is supposed to be twice as powerful now. Taking two levels of cleric will do that. However, based on #3 , taking 10 levels of any other class also should work. This means that taking a single level on this standard character chart (#1) outside your highest class, should provide 5 levels in the lower class, up until the point you equal your highest class.

6. Hmm.

Using other assumptions on the ratio and scaling, can an alternative structure be made to work?

For example, if each single class level is only x1.2 increase in power, instead of x2 every two levels, that drops the scaling considerably. The flat worth for x2 is 1,1.4,2,2.8,4 (roughly). The flat worth for x1.2/level is 1,1.2,1.4,1.7,2, etc. Or x2 power every 5 class levels. This means that a 10th level cleric branching into wizard could take, over his next 5 levels on the main chart, either 5 levels of cleric or 10 levels of wizard, to stay even. Or, as long as his cleric level remains higher, he can take 2 wizard levels for each "character" level gained.


It's too late for me to even try thinking about what this would mean for more complicated multiclassing. I do see a couple of gaping problems, though:

• If power doubles every five levels, that means a new spell level in a full caster gets picked up every five levels (instead of every two levels). A wizard would need to get to 11th level to get fireball. Aesthetically, I don't see that flying, even though it's just a kind of scaling (that could be adjusted in that main "character" chart). Though that certainly doubles down on Greg K's house rule to slow spell advancement.
• The classes are necessarily going to have a lot of "dead" levels at that rate.

However, since the only point of such a character chart is to equalize single-class and multiclass progressions, I suppose it could be pulled out of the direct 1:1 ratio with the actual class levels, and expanded. That is, pick up five "character" levels, and you can take a level in your highest class. Pick up a lesser amount, you can take a level in a multiclass. It's one of those things that might be cleaner in refined application than in theory, but right now it makes my head hurt!

 

[The Shadow]

Crazy Jerome, great stuff! I definitely think you're headed in the right direction with this analysis.

5. Thus if a 10th level cleric gains two new levels, he is supposed to be twice as powerful now. Taking two levels of cleric will do that. However, based on #3 , taking 10 levels of any other class also should work. This means that taking a single level on this standard character chart (#1) outside your highest class, should provide 5 levels in the lower class, up until the point you equal your highest class.

Not quite. Actually, the first "level up" put into wizard should get Wiz 7, and the second one takes him to Wiz 10.

Now, how to handle this in a way that won't make people whip out their calculators... Maybe a table is the way to go. How about this:

[sblock=Table]

Level     Power     Difference
1         1         -
2         1         0
3         2         1
4         3         1
5         4         1
6         6         2
7         8         2
8         11        3
9         16        5
10        23        7
11        32        9
12        45        13
13        64        19
14        91        27
15        128       37
16        181       53
17        256       75
18        362       106
19        512       150
20        724       212

[/sblock]

So the 10th level cleric has a Power of 23. Going up to 11th level would give him a Power of 32, a difference of 9. With Power 9, he can "purchase" 7 levels of wizard. Then, gaining another "level" (we need more terms!) he'd gain 13 more Power, for a total of 23 in cleric and 22 in wizard - or Wiz 9. (Due to round-off error, he doesn't quite make it up to 10th level.)

If he then wanted to branch out into rogue, he'd get 19 Power, for Rogue 9. So now he's a Cleric 10/Wizard 9/Rogue 9. Hrm, that seems rather extreme, doesn't it? If he'd remained a cleric, he'd be Cleric 13. Putting another "level" into Rogue would net him Rogue 12.

By the same table, a Cleric 19 would be equivalent to a Fighter 15/Rogue 15/Cleric 15/Wizard 15. That makes perfect "sense", given the assumption that 2 levels equals a doubling of power, but it definitely does not sound right.

Clearly, there needs to be some sort of penalty to take account of the fact of extra versatility from more classes. Or else we need to abandon the assumption that two levels equals a doubling. I think this method is on the right track, but it definitely needs work.

EDIT: In essence, by assuming a doubling of power every X levels, we're recovering the exponential XP charts of 1e. So it's not surprising that multiclass characters would lag only a few levels behind single-classed characters.

 

[The Shadow]

Okay! Take 2.

Since 3e-linear is too slow, and 1e-exponential is too fast, let's try quadratic. New table - with a handy pattern to it.

[sblock=Table]

Level     Power     Difference
1         1         -
2         2         1
3         4         2
4         6         2
5         9         3
6         12        3
7         16        4
8         20        4
9         25        5
10        30        5
11        36        6
12        42        6
13        49        7
14        56        7
15        64        8
16        72        8
17        81        9
18        90        9
19        100       10
20        110       10

[/sblock]

Now when our 10th level cleric devotes a "level up" to wizard, he gets Wiz 4. If he devotes a second one, he gets Wiz 6.

If he then picks up a level of rogue, he gets Rog 4.

Meanwhile, if we take a 19th level character as before and split up the power equally between four classes, we get 9/9/9/9. Which my gut says is closer to the mark than either 3e 5/5/5/4 or the previous table of 15/15/15/15.

What say you guys?

 

[Yora]

Not to rain your parade, and it's probably quite an interesting hypothetical concept.

But I quickly browse over it and my first thought is "way too complicated to be practical for the game".

 

[The Shadow]

I'd argue it's less complicated than 1e was. Here, you only have to look up one number in a table. There you had to divide your XP by 2 or 3, and then look up those 2 or 3 numbers in different tables.

 

[MoonSong]

But XP manipulation is needlessly complicated. Stacking of levels may lead to weaker characters, but it is very simple to grasp, I know there are concerns that "not every level is worth the same" (and I'm aware of them, like the infamous Fighter 3 vs Wizard 3 in third edition), but go too far with that assumption and suddenly multiclass becomes broken (like the multiclass system of certain "so-called-rpg" in which at certain point it takes way more xp to raise a high level class further than to rise a brand new class from nothing into an equal or almost equal level).
I really like the stack level approach from third edition, and it would be a shame to lose it on balance grounds, specially since the designers have told they were working on a less front-loaded approach, it is impossible to quantiffy the power level of a certain level, because many factors have to be taken into account, and they are different from table to table, including playstyle, pace and allowed materials (for example the healer class is way better at provide healing than clerics, unless you allow complete divine or spell compendium, at which point it becomes awfully bad at it), it cannot be measured in absolutes, any attempt to implement a "balanced" solution based on those grounds becomes arbitrary, and risks being very lennient with certain classes and very draconian with others, notice that the simple stack by level already suffers from those two points, but at least is simple enough, any further mechanical thinkering like the proppossed above will add complexity for no visible benefit other than increasing the power of certain combos, but will truly solve nothing .

 

[Crazy Jerome]

The thoughts behind it are complicated. When you consider multiclassing and all of its issues, the thoughts behind it are always complicated--unless you don't care that it breaks in places.

The real question is can the complicated parts be put into a system that handles the breakage while making it easy for the players to use? That's what we are playing with.

 

[Crazy Jerome]

Now when our 10th level cleric devotes a "level up" to wizard, he gets Wiz 4. If he devotes a second one, he gets Wiz 6.

If he then picks up a level of rogue, he gets Rog 4.

Meanwhile, if we take a 19th level character as before and split up the power equally between four classes, we get 9/9/9/9. Which my gut says is closer to the mark than either 3e 5/5/5/4 or the previous table of 15/15/15/15.

What say you guys?

I think the only thing missing is parsing out the levels of the multiclasses more smoothly. I was spitballing it last night as a common factor that could thus be extrapolated into a single chart. But if you try to merge your "level" and "difference" columns, we can see that won't work consistently.

So the next thing is a clean way to parse out the "difference" columns into parts so that our 10th level cleric taking wizard levels gets his 4 wizard levels over a reasonable progression, one at a time.



Alternately, if the system wants to have a multiclass penalty of some sort, you can use other things. For example, while going from 10th to 11th character level, the cleric gets his 4 levels of wizard. However, he can't pick up more than one wizard level per adventure, to account for training times. So that means:

• He is wasting XP (if used) and/or "bumps" if the DM doesn't use XP --that's the penalty.
• Or the group is going on easier adventures to spread that XP (or alternate measurement) out more efficiently.

In effect, you get those first 4, 6, etc. levels dirt cheap in relative XP costs, but time in the game is stretched out. Meanwhile, a character that starts multiclassing from near the start (e.g. roughly 50/50 fighter/wizard) is gaining each level at a time more appropriate to the adventures for his character level. It's a penalty for late dipping, and gets worse the higher level the character is.

Edit: To really hit it, make the character level be gained first, before the new class starts. Now our 10th level cleric must reach 11th level, where upon he picks up 1 level of Wizard. However, since he is "owed" 3 more wizard levels on the exchange, he can gain these 3 levels at whatever pace the group says is appropriate. He's paid up front, but doesn't get them until the story/group/house rules/DM says he can. Furthermore, he can't effectively buy more wizard levels until these are worked out. (He can buy them with character level, but will still only get them as they play out.) Eventually, it all works out balanced, but the character is radically slowed in the meantime.

 

[The Shadow]

So the next thing is a clean way to parse out the "difference" columns into parts so that our 10th level cleric taking wizard levels gets his 4 wizard levels over a reasonable progression, one at a time.

There is a really easy way to do this, but it does back away from the assumption that you can drop XP-tracking.

Instead of gaining XP, that you then use to look up your character level; instead the whole XP system is rejiggered so that XP translates directly into Power. Maybe every 1000 XP = 1 Power or something. So if you gained 5 points of Power last adventure, you can apply that toward your next level of cleric, or invest it into levels of something else.

(No, this still isn't a point-buy system, because you're still getting levels of classes in lump-sums of power. But it's true you don't have a well-defined "character level" any more. Just a well-defined "character power".)

Add a rule that you can only gain one level per session or something, and you're done. (Me, I'd also only give out XP after each adventure, rather than each session.)

Alternately, if the system wants to have a multiclass penalty of some sort, you can use other things.

It would be dead easy to adjust the table to include a penalty for branching out, if this is desired.

Edit: To really hit it, make the character level be gained first, before the new class starts. Now our 10th level cleric must reach 11th level, where upon he picks up 1 level of Wizard. However, since he is "owed" 3 more wizard levels on the exchange, he can gain these 3 levels at whatever pace the group says is appropriate. He's paid up front, but doesn't get them until the story/group/house rules/DM says he can. Furthermore, he can't effectively buy more wizard levels until these are worked out. (He can buy them with character level, but will still only get them as they play out.) Eventually, it all works out balanced, but the character is radically slowed in the meantime.

This works too.

 

[Crazy Jerome]

Another interesting thing about that "power" column is that it maps reasonably close to full casters getting a new level of spells every level, up to 7th level caster getting 4th level spells. But 5th shouldn't come until 10th level caster, and 6th really should be put off until 14th. I think the gap will keep expanding from there.

That means that either spells above 4th level would need to be toned down (at least the good ones), pushing the really powerful effects out past the 9th level spell bracket (or into some alternate "epic" system), or we would need about 36 character levels to accommodate the pattern, or we can compensate non-casters over 7th level with matching "over-powered" ability, comfortable that a multiclasser can only edge into these upper levels anyway.

 

[The Shadow]

Another interesting thing about that "power" column is that it maps reasonably close to full casters getting a new level of spells every level, up to 7th level caster getting 4th level spells. But 5th shouldn't come until 10th level caster, and 6th really should be put off until 14th. I think the gap will keep expanding from there.

I'm not quite clear on how you're seeing this, could you explain? For starters, which chart are you looking at, the first or the second?

EDIT: Wait, are you referring to approximate power doublings in the second table? The whole point of the second table is that power *doesn't* double at a constant rate.

 

[KidSnide]

4. Assume that there is some expected ratio for class effectiveness over levels, such that a character taking N levels is X better than before the levels were taken. (For ease of consideration and historical reasons, think roughly twice as powerful every two levels increase, though the ratio could be any number of things. This is not unlike 3E CR.)

5. Thus if a 10th level cleric gains two new levels, he is supposed to be twice as powerful now.

I don't want to rain on your analysis, but with bounded accuracy, the 3E CR rule of "two levels up, twice as powerful" probably isn't an accurate reflection of D&DN. (Putting aside for the moment the question of whether it was ever true in 3.x...)

Personally, I wouldn't be surprised to see each class include a special "multi-class" rule (maybe with its own level advancement table or maybe not) that explains how the class changes when it's taken by a multi-class character. I think the lesson from 4e hybrid characters is that you can give out standard character abilities if you keep a sharp eye out for unusually advantageous "splash" classes or abilities that synergize too well (i.e. are significantly more powerful in combination with abilities from other classes than they are in the context of the base class).

-KS

 

[sec_tcpaipm]

Instead of the table

How about - at every character level, for each class you have, you can increase that class by a level or add a new class at the old class level -2. So a 10TH level fighter can go to either F11 or F10C8 at character level 11. At 12th character level she(F10C8) can go to F11C9, F10C9W8 , F11C8W6 or F10C8W8T6. This preserves the CL + CL-2 = CL+1, Like CRs in 3.X. Other power doubling speeds can be handled by changing the level subtraction from a new class. Limits like: cannot add more than one class per character level, cannot raise a class level more than 2 above your first class, etc., might need to be included. We'd also need to be able to forego increasing our highest-levelled class to increase a lowered-level class by more than one level. Just as long as you can add up class levels using the 2CL = CL+2 and CL + CL-2 = CL+1 to get the character level. "Retraining" to decrease a class's level by 1 to free up room for other class's advance could also be incorporated.

 

[sec_tcpaipm]

I don't want to rain on your analysis, but with bounded accuracy, the 3E CR rule of "two levels up, twice as powerful" probably isn't an accurate reflection of D&DN. (Putting aside for the moment the question of whether it was ever true in 3.x...)

Personally, I wouldn't be surprised to see each class include a special "multi-class" rule (maybe with its own level advancement table or maybe not) that explains how the class changes when it's taken by a multi-class character. I think the lesson from 4e hybrid characters is that you can give out standard character abilities if you keep a sharp eye out for unusually advantageous "splash" classes or abilities that synergize too well (i.e. are significantly more powerful in combination with abilities from other classes than they are in the context of the base class).

-KS

Or the classes could "gestalt" instead of "stack", keeping the accuracy bounded to that of the best class(es), which would be necessarily less than or equal to a single-class character of that character level. Instead of a higher number, they'd get more options.

 

[KidSnide]

Or the classes could "gestalt" instead of "stack", keeping the accuracy bounded to that of the best class(es), which would be necessarily less than or equal to a single-class character of that character level. Instead of a higher number, they'd get more options.

If you mean a 5th level fighter, 6th level cleric gets the class abilities of a 5th level fighter and a 6th level cleric (with the higher magic and weapon attack modifier), then yes, I think that's very likely.

I think the issues mostly come from splashing one level of a class. For example, if you splash cleric as presently written, you can take the war domain to get a load of proficiency bonuses. If you splash the current rogue, you get some fantastic skill abilities. These are the things that might prompt WotC to spread out these abilities in a special multi-class paragraph.

The other issue is how to make spell casters multi-class well. Bounded accuracy (for saving throws) and making spells no longer scale with caster level helps, but I don't have a grasp on whether it helps enough. I speculate that a 5th level wizard / 5th level sorcerer is still going to suck compared with almost any 10th level single-class character. That would be a shame (mixing casting traditions is an interesting choice), but I don't know if the designers will really fix that problem in D&DN.

-KS

 

[sec_tcpaipm]

If you mean a 5th level fighter, 6th level cleric gets the class abilities of a 5th level fighter and a 6th level cleric (with the higher magic and weapon attack modifier), then yes, I think that's very likely.

I think the issues mostly come from splashing one level of a class. For example, if you splash cleric as presently written, you can take the war domain to get a load of proficiency bonuses. If you splash the current rogue, you get some fantastic skill abilities. These are the things that might prompt WotC to spread out these abilities in a special multi-class paragraph.

The other issue is how to make spell casters multi-class well. Bounded accuracy (for saving throws) and making spells no longer scale with caster level helps, but I don't have a grasp on whether it helps enough. I speculate that a 5th level wizard / 5th level sorcerer is still going to suck compared with almost any 10th level single-class character. That would be a shame (mixing casting traditions is an interesting choice), but I don't know if the designers will really fix that problem in D&DN.

-KS

Would it help if a wizard/sorcerer 5/5 is the same character level as a wizard 7 or a sorcerer 7? if it is as difficult(or just slightly less) for a 5th level wizard to become 7th level as it is for a 5th level wizard to become a wizard/sorcerer 5/5?

or in order to get that first rogue level, you must already be 3rd level and take that rogue level instead of that next level of whatever, and then they increase together?

EDIT: OTOH there are some class features that a character should only get if the class is the highest-level class for that character (first class in case of a tie).

 

[Crazy Jerome]

I'm not quite clear on how you're seeing this, could you explain? For starters, which chart are you looking at, the first or the second?

EDIT: Wait, are you referring to approximate power doublings in the second table? The whole point of the second table is that power *doesn't* double at a constant rate.

I'm referring to the way that a wizard, cleric, or other full caster is generally considered to double in power each time he gains a new level of spells. It's fairly inexact, has exceptions (e.g. 7th caster level), and depends on both the higher level spells and the proliferation of lower-level slots. So it might not all apply to Next.

In your chart, the power increase is close enough to this ratio to keep casters getting a new level of spells every level up to 7th level. After that, though, the cumulative effects of how you toned down the power doubling rate begin to become too strong to ignore. So something has to be done to compensate, or casters cannot continue to get standard D&D spells every odd level past that point.

As for KidSnide's point, part of the problem with the non-casters versus full-casters in 3E is that the ratio doesn't hold equally. It does trail off a little for casters at the higher levels, but trails off more and sooner for the non-casters.

That said, rain on my analysis all you want. I'm only throwing it out there to see if it sticks, intrigued by the difficulties of addressing Stalker's requirement in a way that I would find tolerable.

 

[Crazy Jerome]

The other issue is how to make spell casters multi-class well. Bounded accuracy (for saving throws) and making spells no longer scale with caster level helps, but I don't have a grasp on whether it helps enough. I speculate that a 5th level wizard / 5th level sorcerer is still going to suck compared with almost any 10th level single-class character. That would be a shame (mixing casting traditions is an interesting choice), but I don't know if the designers will really fix that problem in D&DN.

I don't think 5/5 compared to 10 staight can be resolved with traditional D&D classes. (It might be with the d20M stuff Greg K was discussing. I'm unsure.) Equal to a 10th level wizard is going to be something between 5/5 wiz/sor and 9/9 wiz/sor. Exactly where depends on how you do the classes.

I do think it gets easier to manage if most basic abilities don't stack in such a system. Take your best hit point formula from a single class, whatever that is. Skills and other discrete options are where it gets tricky, as you indicated.