Multi-class hit points - once and for all...

[northrundicandus]

Scotty's view happens to match my view, which hasn't changed in many years - with the exception of the line:

d4+2 when the magic-user class XP requirements are met.

I think he meant "d4+2/3", which the mathematics nut in me translates as (d4+2)/3, so there's no confusion over the order of operations.

[rogatny]

And afterwards:
d8+2/3 when the cleric class XP requirements are met.
d10+4/3 when the fighter class XP requirements are met.
d4+2 when the magic-user class XP requirements are met.
Until the cleric ability is maxed out. Then
d10+4/2 when the fighter class advances
d4+2/2 when the magic-user class advanced.

My question is where is the support for this BtB?

Hit Dice Type shows the type of die to be rolled by a character of the appropriate class at each level of experience (q.v.) he or she has gained so as to determine how many hit points (q.v.) the character has. Multiclassed characters determine their hit points as follows:

1. Roll the hit die (or dice) appropriate to each class the character is professing.

2. Total the sum of all dice so rolled, and adjust for constitution (q.v.). Divide the total by the character's classes (two or three), dropping fractions under %, rounding fractions of % or greater upwards to the next whole number.

3. The number derived (quotient) is the number of hit points the multi-classed character gains with the rise in that experience level.

Note that when multi-classed characters are no longer able to progress in any given class, they no longer gain the hit dice for that class. (See CHARACTER HIT POINTS).

Emphasis mine.

It basically says... Single classed characters roll their hit points by the hit dice listed. Multi-class characters roll their hit dice as follows... roll the dice of each class and average for each rise in experience level. It then says, when a character can no longer rise in levels, you no longer average in that hit dice.

I see nothing that indicates you do something different with a rise in level than you do at character creation.

I'm sympathetic to the arguments against it, I just don't see the textual support for a contrary interpretation. I'd, frankly, love to see it.

So yes, this means a f/m/t with 100,000 xp will be leveled 5/5/6, and have basically rolled hit dice 14 times. While a single classed character with the same xp will have only rolled 7 or 8 times. If that IS the way that it was meant to be done, that would lend credence to those who say that the Con adjustment should just be added once to the die rolls before dividing.

BTW: If we were to assume that the C&C version of multi-classing presented in Yggsburgh represents Gary's current view, he simply rolls the HD of the class as it gains a level... Thus if a fighter-rogue gains a level in fighter, he gains d10 hp. A dubious argument to be sure... however, mathematically it works out about the same as the way I'm interpreting the rules.

But that brings us to the best argument against the way I'm reading the rules... There's simply no reason for it. You end up with about the same number of hp if you simply roll the HD of the class that gains a level without going into the silly averaging bit. Why on earth is all that averaging there? The text is pretty plain in that it's not just for the initial character creation, but applies to each rise in xp.

So that brings me back to that "professing" in point 1, and wondering whether that might have been a bad choice of words.

I checked all through the Sage Advice columns in my Dragon Archive last night and found nothing. (Not that Sage Advice is a particularly good resource for this kind of thing... there's so much wrong and contradictory stuff in there it's maddening. I also never realized that there was an about 3 year period from 1983 and 1986 where S.A. wasn't ever in Dragon.)

If anyone suspects that I'm enjoying this... I am. I once wrote a paper in law school about a Federal perjury statute in which courts have decided that an "and" actually means "or."

R.A.

[John Stark]

This to me is a nice bit of evidence that multiclass hit points are supposed to be averaged from each of the class hit dice every level gained. Notice how hit points are figured here per double class level for a F/MU.

That's just an average. In general, an F/MU won't progress in both of his classes simultaneously. I would agree with you if the progression were simultaneous, but it isn't.

This is a reductionist argument. Proclaiming that something is "just x" does not necessarily make it so.

Further, the whole point of that example, given the rules on pg. 19 for figuring multiclass hit points, is that an average between the character's class hit dice is the way you figure hit points per double class level.

When you say that "in general the F/MU won't progress simultaneaously in both of his classes," I would say that does nothing but bolster my point. Obviously, looking at the XP charts, no two classes are going to advance simultaneaously. If we wanted to give an accurate example of hit points for a multiclass character using a method that only takes into account each class's hit die when that class gains a level, we certainly wouldn't use the above example since it doesn't match that method at all.

If a F/MU gains a level, and is supposed to get his normal HD divided by the number of classes that he has, then a proper example would show (5.5/2) for fighter hit points gained per whenever a fighter level is gained, and (2.5/2) whenever a MU level is gained. But what does the example say? It says that (5.5 + 2.5 = 8/2 = 4 hit points per double-classed level) are gained. A formula is being applied per double class level. What is a double class level? Its simply a level gained by a multiclass in one of his classes. Double classed level certainly does not mean "from both classes," as the writer of the text must have been aware that NO multiclassed characters will ever gain levels in two classes simultaneaously. Thus, I submit here that double classed level, in the context of pg. 19 (wherein a formula is given for determining multiclass hit points, and NOT a straight die roll as per single class characters) and in the broader context of the XP requirements of each class (which clearly show that no two classes will ever advance at the same time), means each level gained by the multiclass character. If double classed level is defined as such, then the formula from pg. 19 is applied, and the example on page 32 is an example of that formula.

What happens when the character in question gains a level in the Fighter class, but not in the Magic-user class? The answer to that question, I submit, is the key to interpreting how hit points are generated at first level.

A reread of page 19 is certainly in order I guess:

Hit Dice Type shows the type of die to be rolled by a character of the appropriate class at each level of experience (q.v.) he or she has gained so as to determine how many hit points (q.v.) the character has. Multiclassed characters determine their hit points as follows:

1. Roll the hit die (or dice) appropriate to each class the character is professing.

2. Total the sum of all dice so rolled, and adjust for constitution (q.v.).

3. Divide the total by the character's classes (two or three), dropping fractions under 1/2, rounding fractions of 1/2 or greater upwards to the next whole number.

4. The number derived (quotient) is the number of hit points the multi-classed character gains with the rise in that experience level.

Note that when multi-classed characters are no longer able to progress in any given class, they no longer gain the hit dice for that class. (See CHARACTER HIT POINTS).

I think that once this passage is considered in context, and the wording of the passage is carefully considered, that its meaning becomes eminently clear.

Point #1: What is the general thrust of the passage? To determine the hit die types for characters.

"Hit Dice Type shows the type of die to be rolled by a character of the appropriate class at each level of experience (q.v.) he or she has gained so as to determine how many hit points (q.v.) the character has."

Clearly the point of this note is to clarify what is meant by "Hit Die Type" in the table above this note on pg. 19. And what does it say? The "Hit Die Type" column indicates the type of die to be rolled "at each level of experience... gained."

Point #2: What comes next?

"Multiclassed characters determine their hit points as follows:"

This clearly follows after the former point, and needs to be read in that context, wherein the context is explaining what "hit die type" means as a character advances.

Point #3: A method is then given, with several steps listed on how to determine hit points for a multiclass character, and this method MUST be considered in the above context; to whit, the context wherein "hit die type" is explained and how that pertains to character advancement "at each level of experience... gained."

Point #4: No where in this context does pg. 19 say that the method given changes from 1st level for a multiclass to when other levels are gained.

"Multiclassed characters determine their hit points as follows:"

This passage does not give different methods for determining multiclass hit points between 1st level and later levels. Certainly the same method for determining hit points for a single class character is ALWAYS the same, whether at 1st level or later levels. This would also argue strongly that the given method for determining multiclass HP is uniform as well.

Point #5: The first step clearly states that dice are rolled "appropriate to each class the character is professing."

"1. Roll the hit die (or dice) appropriate to each class the character is
professing."

The above does not say, "Roll only the die for the class that has gained a level." This step, taken in the context of determining "hit die type" "at each level of experience... gained," rightly means to "roll the hit die or dice for each class the character is professing." This method is not varied from first level to beyond anywhere in this passage.

Point #5: The total of all dice rolled, considered in the proper context of "hit die type" "at each level of experience... gained" shows that there is a totalling of dice each level.

"2. Total the sum of all dice so rolled, and adjust for constitution (q.v.)."

There can only be a totalling "of all dice rolled" when there are more than one die rolled. This ONLY makes sense in two instances; at first level, when there will certainly be more than one die type to consider given that we are talking about a multiclass character, AND at later levels when an average between different class hit die types are being used to determine hit points when one of the classes has gained a level. If only ONE die is rolled when a multiclass character gains a level, then there would be no need to TOTAL up the dice.

And again, the proper context here is "hit die type" "at each level of experience... gained" and the unvarying method (i.e., the formula) given for determining multiclass hit points.

Whether or not Constitution is applied to each hit die type in the formula is a question of dispute. I'm inclined to say that the whole point of "totalling" the hit dice points to the idea that you do this before you adjust for Constitution. Otherwise, this passage would say, "Adjust each hit die for Con, then total." Either way, it doesn't make alot of difference in terms of the method (i.e., formula) that is being given for determining hit points for a multiclass "at each level gained."

Point #6: Step 3 is self explanatory and clear. It would pertain regardless of our interpretation of this passage as a whole.

"3. Divide the total by the character's classes (two or three), dropping fractions under 1/2, rounding fractions of 1/2 or greater upwards to the next whole number."

Notice however that this states to "divide the total by the character's classes (two or three)." How do we divide "the total" if there was nothing to "total up?" We certainly don't "total up" a single hit die roll, as there is nothing to "total up;" we merely read the result of the die roll if it were the case that we only roll one die when a multiclass character gains a level in one of his classes.

Point #7: Step 4 clearly points us back to the original context of "hit die type" "at each level of experience... gained."

"4. The number derived (quotient) is the number of hit points the multi-classed character gains with the rise in that experience level."

In other words, the method given in 1-4 are the method used "with the rise in that experience level." No other method is given in this passage.

Point #8: The final sentence given after the method laid out in 1-4 points to the idea that each class level gained means we use the method given.

"Note that when multi-classed characters are no longer able to progress in any given class, they no longer gain the hit dice for that class. (See CHARACTER HIT POINTS)."

As I stated up thread, why would this need to be stated if multiclass characters gain only the hit die relevant to each class when that class gains a level? It should be obvious that a multiclassed character would no longer gain HD from a class that is maxed out if we only roll the hit die that the class would normally get when it gains a level.

However, if HD are determined for multiclass characters via a formula wherein hit dice for each class is rolled every time any one of the classes gains a level, this final note makes perfect sense. In other words, this statement at the end of the check list would not be needed if we weren't using the given formula of averaging class hit dice each time a class gains a level. Its only when that formula is being used throughout the multiclassed character's career that we would need this statement to clarify what we should do once a class has maxed out.

In light of this passage on pg. 19, and its context, I think the example given on page 32 for figuring hit points for a F/MU shows the proper method to be used; i.e., averaging the hit die types of each class every time a multiclass character gains a level in one of his classes.

The words of the passage have meaning, and make sense only in the given context. Once the context is properly understood, the rest of the passage falls into line pretty clearly. Its only when we ignore the context in question that confusion arises. The point here, once again, is to determine what the passage actually says so that we know the RAW. After that, if we think the method is too complicated, unwieldy, or unbalanced, its certainly within our purview as DMs and/or players to use a different method. But let's not declare a method "BtB" or the "RAW" that does not take into account the actual language and context of the passage in question.

[rogatny]

John,

Thanks, that's a great step-by-step of what I was trying to say. I agree with you that, that's what the books says. Whether it SHOULD be like that, or whether that's what the author intended might be another question, but that is the rule as written.

And like you suggested earlier... That pretty much puts to rest any question I've ever had in my mind about the "poor demi-human has to deal with level-limits" crap. Multi-class, get a ton of hit points, and be happy.

R.A.

[John Stark]

You end up with about the same number of hp if you simply roll the HD of the class that gains a level without going into the silly averaging bit.

I didn't think that this was correct, and indeed was the reason you were changing your view in terms of how to calculate multiclass hit points.

Here's an example of the differences between the two methods being described:

F/MU with 160,000 XP

Method 1: Average the hit dice of each class everytime one class gains a level.

F/MU 1/1: (d10 (5.5) + d4 (2.5))/2= 8/2 = 4
F/MU 2/1: (5.5 +2.5)/2 = 8/2 = 4
F/MU 2/2: (5.5 +2.5)/2 = 8/2 = 4
F/MU 3/2: (5.5 +2.5)/2 = 8/2 = 4
F/MU 3/3: (5.5 +2.5)/2 = 8/2 = 4
F/MU 4/3: (5.5 +2.5)/2 = 8/2 = 4
F/MU 4/4: (5.5 +2.5)/2 = 8/2 = 4
F/MU 5/4: (5.5 +2.5)/2 = 8/2 = 4
F/MU 5/5: (5.5 +2.5)/2 = 8/2 = 4
F/MU 6/5: (5.5 +2.5)/2 = 8/2 = 4
F/MU 6/6: (5.5 +2.5)/2 = 8/2 = 4
F/MU 7/6: (5.5 +2.5)/2 = 8/2 = 4
F/MU 7/7: (5.5 +2.5)/2 = 8/2 = 4

Total HP: 52

Method 2: Roll only the die type for each class when it gains a level.

F/MU 1/1: (d10 (5.5) + d4 (2.5))/2= 8/2 = 4
F/MU 2/1: (5.5)/2 = 2.75 = 3
F/MU 2/2: (2.5)/2 = 1.25 = 1
F/MU 3/2: (5.5)/2 = 2.75 = 3
F/MU 3/3: (2.5)/2 = 1.25 = 1
F/MU 4/3: (5.5)/2 = 2.75 = 3
F/MU 4/4: (2.5)/2 = 1.25 = 1
F/MU 5/4: (5.5)/2 = 2.75 = 3
F/MU 5/5: (2.5)/2 = 1.25 = 1
F/MU 6/5: (5.5)/2 = 2.75 = 3
F/MU 6/6: (2.5)/2 = 1.25 = 1
F/MU 7/6: (5.5)/2 = 2.75 = 3
F/MU 7/7: (2.5)/2 = 1.25 = 1

Total HP: 28

That's quite a disparity there. Which is why the method as laid out in the book seemed so incongruous or unbalanced to you, I thought.

If anyone suspects that I'm enjoying this... I am. I once wrote a paper in law school about a Federal perjury statute in which courts have decided that an "and" actually means "or.".

Heh. All of my professors in college told me I should have proceeded to law school after my undergrad work was done, precisely because I enjoy arguing minutiae like this. It drove them nuts.

[rogatny]

You end up with about the same number of hp if you simply roll the HD of the class that gains a level without going into the silly averaging bit.

I didn't think that this was correct, and indeed was the reason you were changing your view in terms of how to calculate multiclass hit points....

Nope. I was referring to the Yggsburgh version where you simply roll the HD of the class gaining a level and don't divide by anything. Thus, when a fighter thief gains a level as fighter, he gets d10 hp. When he gains a level in thief, he gets d6 hp. It averages out the same as d6+d10/2 for every level.

R.A.

[John Stark]

And like you suggested earlier... That pretty much puts to rest any question I've ever had in my mind about the "poor demi-human has to deal with level-limits" crap. Multi-class, get a ton of hit points, and be happy.

Oops, cross posted with you RA. I see we are in agreement.

[TRP]

And afterwards:
d8+2/3 when the cleric class XP requirements are met.
d10+4/3 when the fighter class XP requirements are met.
d4+2 when the magic-user class XP requirements are met.
Until the cleric ability is maxed out. Then
d10+4/2 when the fighter class advances
d4+2/2 when the magic-user class advanced.

My question is where is the support for this BtB?

I'd say the support is in the (q.v.).

Foster and Storm have it right on, and it looks like Scotty and north agree. It stands on 2 legs; first it can be supported by the (q.v.) and second, it makes sense intuitively.

This was my interpretation 26 years ago, and I haven't seen a compelling argument yet to change it.

[John Stark]

I'd say the support is in the (q.v.).

Which (q.v.) are you referring to, and which debate are you referring to?

There are essentially two debates running in this thread. The first is how you determine hp for multiclass characters, particularly when one of their classes gains a level. The second debate is merely a subset of the first, dealing with how any Con bonus is applied upon the gaining of any given level for a multiclass.

It'd help if you clarified which debate and which (q.v.) you are referring to.

And honestly, I think that your "side" would need to do a point by point analysis of the language and context of pg. 19, dealing with the many indicators that I laid out which strongly suggest that your method is incorrect, before declaring a "winner," heh.

[rogatny]

I'd say the support is in the (q.v.).

Which one? There are three in the relevant passage.

One refers to "level of experience," another refers to "hit points," and the final one refers to "constitution."

"q.v." simply means "which see," referring the reading to the section on the relevant term. I see nothing in the constitution (pg. 12), hit points (pg. 34), or experience (pg. 106) sections of the book that are instructive in this regard.

R.A.

[rogatny]

Playing with numbers:

An elvish f/m-u maxes out in levels at 5/9... Using the interpretation John and I understand (average of both HD at every level obtained), he'd have an average of 44.5 hp. Under the other interpretation (half one HD with each level obtained), he'd have 31.25

Average hp of other character classes at 270,002 xp:

9th lvl C = 49.5
11th lvl Dr = 49.5
9th lvl F = 49.5
8th lvl P = 44
9th lvl R = 45
10th lvl M-U = 25
10th lvl Ill = 25
11th lvl Th = 37
9th lvl As = 31.5
8th lvl Mo = 22.5

[JCBoney]

The q.v. regarding Constitution can simply refer to the section on CON as in "this is the definition of Constitution." The book is rife with references like that. It does not necessarily mean "see the section on CON and note that it says you should add the CON bonus to every die in regards to multi-class PCs."

See, there's a problem that those just now getting into 1e have: 1e was originally tailored as a step up for those already playing OD&D. It was almost assumed you knew how to play...the PHB and the DMG simply codified the new rules that were mostly gathered from the Dragon over time. I never played OD&D but I had the benefit of starting out on B/X D&D, so I knew basically how to play when I moved on to advanced.

Now we're trying to retro-read the 1e rules and make sense of it, but without taking into consideration that the basic ideas of play were pretty much assumed to be known by the reader, you get nebulous interpretations like the ones we're wrestling with here.

[John Stark]

And like you suggested earlier... That pretty much puts to rest any question I've ever had in my mind about the "poor demi-human has to deal with level-limits" crap. Multi-class, get a ton of hit points, and be happy.

Probably should start a new thread to open this can of worms, heh, but I'll say this.

I think multiclass level limits, given the method we've winnowed out of the text, make perfect sense. The multiclass character gains a ton of advantages from his different classes (particularly if he triple classes as a thief) and alot of HP (if he survives to high levels, that is).

In the end though, I think single classed demihumans get left out in the cold. For NPCs this is fine, but for a player I think it actually detracts in a "gamist" sense. Why play a single classed demihuman at all? Your answer is "multiclass and get over it," heh. As a DM though, I'd think we want to encourage players to try all sorts of combinations of class/race, and single classed demihumans fall under that rubric. Heck, even as a player, I'd like to try most of the classes and races available. For high level play though, a level 7 elven fighter isn't going to cut it, nor a 6th level halfling fighter. Its at high levels of game play where the initial advantages of being a demihuman (race abilities) wear thin when the character is stuck at a medium level with his single class.

The two solutions, IMO, would be to simply eliminate single classed demihumans (which would take us back to more of an OD&D or Basic D&D "race as class" idea), or to allow for higher (or even no) level limits for single classed demihumans, with something else to balance out the initial racial advantages that pertain to such characters. For my own campaign, I use alignment restrictions as an "in game" means of balancing this. Just as many find playing a LG paladin restrictive, and thus opt for a fighter to escape that restriction, so will players choose a single classed human if they don't want to be roped into playing a "good" elf or a "lawful neutral" dwarf. Higher starting stats for single classed demihuman characters might be an option as well, which would mean that DMs would need to enfore the "you play what you roll up" at character creation notion (which, admittedly, doesn't get enforced much, but which I personally think is more of a pure "gamist" approach to starting play).

Hopefully this doesn't turn into a "demihuman level limits" threadjack. Your point of view simply spun me off into this line of thinking, and perhaps I should have started another thread. Of course, the subject has been so beaten to death, maybe I should just stop typing, heh.

[Stormcrow]

I have a singular question for those here who think each die rolled is individually adjusted for CON:

2. Total the sum of all dice so rolled and adjust for constitution.

If the individual die are adjusted for CON, as Lord Cias purports, then why does the above statement call for a totalling, and then adjusting for CON? How can you adjust for CON after the dice rolls are totalled?

Like this:

Fighter/Magic-user/Thief 5/4/6 with 18 Con

Total the sum of all dice so rolled…
5d10 + 4d4 + 6d6

…and adjust for constitution.
+ 20 + 8 + 12

David
Stardate 6668.7

[rogatny]

Like this:

Fighter/Magic-user/Thief 5/4/6 with 18 Con

Total the sum of all dice so rolled…
5d10 + 4d4 + 6d6

…and adjust for constitution.
+ 20 + 8 + 12

Awright...

I see where Stormcrow is coming from. The text on pg. 19 is in reference to Character Class Table I on the same page.

"Hit Dice Type shows the type of die to be rolled by a character of the
appropriate class at each level of experience... he or she has gained
so as to determine how many hit points... the character has."

To mean:

To determine how many hit points an 8th level paladin has, you roll a d10 eight times.

To determine how many hit points a 4th lvl fighter/5th lvl thief has, you do thus and so...

Let me think on this.