5E Random Class and Race Tables

Hriston

Explorer
Hi,

I just wanted to share some tables I’ve started using to make characters. I converted these two tables from pages 175 and 176 of the AD&D 1st Edition DMG to use with 5E. I’ve tried to preserve the ratios from the original tables as much as possible. I roll up a race and a class and voila, instant character. Enjoy!

edit: I've added a new version of the class table and some subtables in a post below, which replaces the one here.

d100
Race
01​
Dragonborn​
02-06​
Dwarf​
07-11​
Elf​
12-13​
Gnome​
14-18​
Half-Elf​
19-20​
Halfling​
21​
Half-Orc​
22-99​
Human​
00​
Tiefling​

d100
Class
01-17​
Cleric​
18-20​
Druid​
21-40​
Barbarian​
41-60​
Fighter​
61-62​
Paladin​
63-65​
Ranger​
66-72​
Sorcerer​
73-79​
Warlock​
80-86​
Wizard (Other)​
87-88​
Wizard (Illusionist)​
89-98​
Rogue (Other)​
99​
Rogue (Assassin)​
00​
Monk OR Bard​
 
Last edited:

Connorsrpg

Adventurer
These percentages would obviously change depending upon the setting and how common each race is. I outline which races are common and uncommon in a region of a setting, then other races are possibilities to be from surrounding regions.

These tables clearly skew the results in favour of some classes and races, which is very cool to get a feeling of what is more common in an area. But again, not sure if any use to PCs who are often exceptions anyway.

For me, there are 12 classes. When rolling my character that is simple; I use a d12 (though in actuality, as I use non-WotC classes too it is more like a d20 :p)

Anyway, thanks for sharing. I like seeing tables like this and love that others do them too. (I have massive race tables for a GM creating a realm and populating them too). :)
 

Hriston

Explorer
These percentages would obviously change depending upon the setting and how common each race is. I outline which races are common and uncommon in a region of a setting, then other races are possibilities to be from surrounding regions.

These tables clearly skew the results in favour of some classes and races, which is very cool to get a feeling of what is more common in an area. But again, not sure if any use to PCs who are often exceptions anyway.

For me, there are 12 classes. When rolling my character that is simple; I use a d12 (though in actuality, as I use non-WotC classes too it is more like a d20 :p)

Anyway, thanks for sharing. I like seeing tables like this and love that others do them too. (I have massive race tables for a GM creating a realm and populating them too). :)
I think the tables on which these tables are based are an example of how 1E has an implied setting, but I wouldn’t take them to strongly imply any demographic information beyond how the typical adventuring party is composed. For anyone not too familiar with the 1E DMG, these tables are pulled from the Character Subtable, which is for generating randomly encountered NPC adventuring parties that are met with in dungeons, as psionic encounters, as henchmen in castles with “character-type” inhabitants, in uninhabited/wilderness areas with temperate and subtropical conditions, and (in slightly modified form) on the astral and ethereal planes. So what these percentages tell us about the implied world is what races and classes one would typically find in a party of adventurers, but not what races or classes one would typically encounter under other conditions or in the world at large. I think the prevalence of half-elves in adventuring parties is a good example of this. I don’t think it implies that half-elves are just as prevalent in the world as elves or dwarves, but rather that a typical party is as just as likely to include a half-elf as it is one of those other two races.
 

Connorsrpg

Adventurer
All true and very useful for a 'standard' game, whatever that is. People may find these useful. Just saying I already have ways of determining these things in my settings. And if playing a PC, I just do randomly and would never use a chart where there is less than 1% of playing a bard. As you said, adventuring parties are different and I doubt less than 1% have bards. Kinda the same point you made with half-elves.

No matter though. I hope people find these useful. :)
 

Lanefan

Victoria Rules
What's missing from the class table is any chance of the character (if higher than 1st level) being multiclass. I'd give it a 10% chance and if it hits, roll again on the class table re-rolling if you get the same class again or if you get something mutually incompatible e.g. an Assassin-Paladin.

To make it reflect 5e a bit more I'd knock 10% off the Barbarian chance and split it between Monk and Bard, as those two classes are a lot more "mainstream" in 5e than they were in 1e.
 
These tables clearly skew the results in favour of some classes and races, which is very cool to get a feeling of what is more common in an area. But again, not sure if any use to PCs who are often exceptions anyway.

For me, there are 12 classes. When rolling my character that is simple; I use a d12 (though in actuality, as I use non-WotC classes too it is more like a d20 :p)
I have sometimes played a randomly generated PC and used d12 to roll for class, but I must say that the uneven probability distribution is exactly what makes a random table interesting.

Because, you know, everyone can make an equal distribution, and they are all the same, but an uneven distribution implies decisions and is therefore an expression of the author's ideas and feelings. :)
 

pemerton

Legend
I think the tables on which these tables are based are an example of how 1E has an implied setting, but I wouldn’t take them to strongly imply any demographic information beyond how the typical adventuring party is composed.

<snip>

what these percentages tell us about the implied world is what races and classes one would typically find in a party of adventurers, but not what races or classes one would typically encounter under other conditions or in the world at large.
Two thoughts:

(1) Have you ever tried to reconcile the table you used with some of the other tables found in the DMG? Obviously the ones for ranger and thief/assassin followers are very specific and hence have different spreads (although given that any race can be a thief one might think the table in that respect shouldn't depart too far from a more generic one; but it does, very significantly). And when it comes to class, there is the henchman one which is different from the encounter one, and then there is the different-again spread of adventuring classes on the city encounter table.

(2) What was your basis for interpolating likelihoods for the classes and races not found in AD&D?
 

Hriston

Explorer
All true and very useful for a 'standard' game, whatever that is. People may find these useful. Just saying I already have ways of determining these things in my settings. And if playing a PC, I just do randomly and would never use a chart where there is less than 1% of playing a bard. As you said, adventuring parties are different and I doubt less than 1% have bards. Kinda the same point you made with half-elves.

No matter though. I hope people find these useful. :)
I’m not sure what a “standard“ game is either. The type of game for which I think some tables like this could be useful is one that has a “Gygaxian” feel, which is my interest in converting them. I think this produces a markedly different result than, say, randomly generating class with a d12 and race with a d9, which I would characterize as more of a “kitchen sink” aesthetic, which I don’t mean to be disparaging in any way. There’s nothing wrong with that preference, and of course the default for both 1E and 5E is for the player to simply choose their desired race and class, but the premise of this thread is that it’s desirable to randomly generate those things, as one might for an NPC, and for my preference making all races and classes equally as likely doesn’t produce the desired result. I hope that clarifies my intent here.
 

Hriston

Explorer
What's missing from the class table is any chance of the character (if higher than 1st level) being multiclass. I'd give it a 10% chance and if it hits, roll again on the class table re-rolling if you get the same class again or if you get something mutually incompatible e.g. an Assassin-Paladin.

To make it reflect 5e a bit more I'd knock 10% off the Barbarian chance and split it between Monk and Bard, as those two classes are a lot more "mainstream" in 5e than they were in 1e.
I’ll have to look into how multiclassing is handled in the Character Subtable to tell you how I’d do it, but yes, I was mostly thinking about generating 1st-level characters in converting these tables, and I also sought to preserve the class-balance as given.
 

Hriston

Explorer
Two thoughts:

(1) Have you ever tried to reconcile the table you used with some of the other tables found in the DMG? Obviously the ones for ranger and thief/assassin followers are very specific and hence have different spreads (although given that any race can be a thief one might think the table in that respect shouldn't depart too far from a more generic one; but it does, very significantly). And when it comes to class, there is the henchman one which is different from the encounter one, and then there is the different-again spread of adventuring classes on the city encounter table.
No, I haven’t and don’t see any reason to. The followers tables, as you said, are specifically for generating followers for upper level characters. Perhaps I should state clearly that my intention with these tables is to provide a method for PC creation. PCs aren’t generally conceived of as followers of upper level characters, although I suppose they could be. And while it’s true that a member of any race could be a thief, the fact that for most demi-humans, Thief was one of a limited number of classes available, would lead to humans being underrepresented on a list of thieves when compared to the general population of adventurers, which is what we see on that table.

The rules for henchmen are interesting but depend on the DM‘s own breakdown of races that exist in a given area, and the class distribution specifically applies only to prospective henchmen, which is a subset of characters suitable for advancement, although it does seem pretty close to the distribution given in the Character Subtable.

I did take a look at the City/Town Encounters Matrix when preparing these tables, particularly the race breakdown, but again since my purpose is to generate a PC (or NPC) member of an adventuring party, I wasn’t particularly perturbed that demi-humans are found in higher proportions in urban environments than they are in adventuring parties.

The short answer is I’m not all that interested in the larger world-building questions that these sorts of comparisons bring up.

(2) What was your basis for interpolating likelihoods for the classes and races not found in AD&D?
I tried to limit arbitrary decision making to a minimum. I’m not sure if I succeeded.

The Class table departs from the original in two respects. First, it gives half the fighter’s space to the barbarian, the basis of which was the barbarian’s history as a subclass of fighter. I had at first thought to only give the barbarian a spread equal to the ranger, but it seemed too arbitrary, so I finally settled on dividing the fighter’s spread evenly.

Second, I divided the former non-illusionist magic-user’s space evenly between the sorcerer, warlock, and non-illusionist wizard, conceiving of “magic-user” as formerly encompassing the concepts of all three 5E classes, however ineffectively. I had originally thought of giving the wizard enough of a spread that each subclass could have equal representation, but I found the result too arbitrary, especially as it impacted the space remaining to the sorcerer and warlock.

In practice, when using this table, once a class has been generated, if determining a subclass is desired and one of the two that appears on the table didn’t result, then I would make a second roll, if necessary, giving equal weight to each subclass of the resulting class. Without taking stock of how many subclasses each class has (I generally use the PHB only), I think a second roll wouldn’t be necessary for Barbarian, Wizard (Other), or Rogue (Other), but rather the spread of each result could be divided equally among the (remaining) subclasses.

The Race table was more straightforward. I kept the proportions between the existing races the same and gave dragonborn and Tieflings a value equal to the least represented race on the original table, half-orcs (coincidentally, perhaps, also a “monster race”). My thinking was influenced by the inclusion of these additional races in the PHB as “rare” races. After rounding, the extra 2% that these races take up essentially came out of the human’s share.
 

Hriston

Explorer
@Lanefan, I had to look at the 1E PHB to remind myself how multiclassing (and human two-classing) works in 1E.

Yes, 10% looks right on the money if you want to extend multiclassing to humans, as 5E does. Since I’m going for the Gygaxian aesthetic, I’m tempted to just stick with the multiclass percentages given for each demihuman race, which leaves dragonborn, humans, and Tieflings out in the cold, but I guess they could still multiclass going forward (which begins to look a little like replicating the humans with two classes rule).

Either way, if starting at 3rd level or higher, I’d give any character that qualifies as multiclassed an additional 25% chance of being triple-classed.

All levels would be distributed evenly between classes, priority going In the order rolled, and the first class rolled would be treated as the character’s 1st level class.
 

Hriston

Explorer
(2) What was your basis for interpolating likelihoods for the classes and races not found in AD&D?
After further consideration of my answer to this question, I've found that I'm dissatisfied with how the class table came out and my decision making about that. My approach to the race table, on the other hand, was to treat the additional races (Dragonborn and Tiefling) as equivalent in frequency to the least represented race on the original table (Half-Orc). This approach allows for the in-fiction rationale for the absence of the two additional races on the original table to be that they were omitted or overlooked because of their rarity. I find this approach to be the more satisfying of the two.

So applying this approach to the class table, there are three classes and one subclass that I don't think are represented on the original table: Barbarian, Rogue (Arcane Trickster), Sorcerer, and Warlock. Giving each of them a value equivalent to the least represented classes on the original table (Monk and Bard, with 1/2 percent apiece), I got this result (and the following subtables):

Edit: These tables have been edited as per the discussion and suggestion from @pemerton down-thread

d100
Class
01-02​
Cleric (Death)
03-04​
Cleric (Knowledge)
05-06​
Cleric (Life)
07-08​
Cleric (Light)
09-10​
Cleric (Nature)
11-12​
Cleric (Tempest)
13-14​
Cleric (Trickery)
15-16​
Cleric (War)
17-19​
Druid, see subtable
20-38​
Fighter (Battle Master)
39-57​
Fighter (Champion)
58-60​
Fighter (Eldritch Knight)/Barbarian, see subtable
61-62​
Paladin, see subtable
63-65​
Ranger, see subtable
66-85​
Wizard (non-Illusion), see subtable
86-87​
Wizard (Illusion)
88​
Sorcerer/Warlock, see subtable
89-98​
Rogue (Thief/Arcane Trickster), see subtable
99​
Rogue (Assassin)
00​
Monk/Bard, see subtable

Druid subtable
d6
Circle
1-3​
Land, see subtable
4-6​
Moon

Circle of Land subtable
d8
Terrain
1​
Arctic
2​
Coast
3​
Desert
4​
Forest
5​
Grassland
6​
Mountain
7​
Swamp
8​
Underdark

Fighter (Eldritch Knight)/Barbarian subtable
d12
Class
1-10​
Fighter (Eldritch Knight)
11​
Barbarian (Berserker)
12​
Barbarian (Totem Warrior)

Paladin subtable
d8
Oath
1-2​
Ancients
3-4​
Devotion
5-6​
Oathbreaker
7-8​
Vengeance

Ranger subtable
d6
Archetype
1-3​
Beast Master
4-6​
Hunter

Wizard (non-Illusion) subtable
d8
School
1​
Abjuration
2​
Conjuration
3​
Divination
4​
Enchantment
5​
Evocation
6​
Necromancy
7​
Transmutation
8​
Roll again

Sorcerer/Warlock subtable
d12
Class
1-3​
Sorcerer (Draconic Bloodline)
4-6​
Sorcerer (Wild Magic)
7-8​
Warlock (Archfey)
9-10​
Warlock (Fiend)
11-12​
Warlock (Great Old One)

Rogue (Thief/Arcane Trickster) subtable
d20
Archetype
1-19​
Thief
20​
Arcane Trickster

Monk/Bard subtable
d12
Class
1-3​
Bard (Lore)
4-6​
Bard (Valor)
7-8​
Monk (Four Elements)
9-10​
Monk (Open Hand)
11-12​
Monk (Shadow)

I've also come up with some tables for backgrounds, but I think I'll start a new thread for those.
 
Last edited:

Shiroiken

Adventurer
Random character creation would be an interesting campaign idea, assuming you have full player buy-in. You could also do it as a one-shot, but honestly I think just making pre-gens would be better.

If I was going to do a set of random tables, I'd go all out. I'd have tables, followed by sub-tables, followed by more and more sub-tables. I'd go with race, sub-race, background, personality, ideals, bonds, flaws, class, sub-class, and ended with XGtE "this is your life." It would be like Traveller without the random death in character creation. Race would impact everything after it, since it would adjust many cultural aspects. The goal of this would be for someone to test the player's roleplaying ability to adapt to something that may be completely different than anything they would ever choose.
 

Lanefan

Victoria Rules
Where the random tables come in really handy IME is when a player has some specific ideas for a character but no real ideas on how to flesh out the rest.

For example, someone might have a great idea for an Elf with some specific personality traits, but no clue what class it should be and so rolls for its class and some background details on the random class table.
 

Hriston

Explorer
@Hriston. is there an argument that Eldritch Knight probabilities should be closer to F/MU? And Arcane Trickster could be similar to MU/Th, F/MU/Th and Il/Th.
Yes, I think there is! Initially, I hadn’t given much thought to multi-classing and had forgotten that the concepts represented by Eldritch Knight and Arcane Trickster were already present in 1E’s multi-classing rules. I think this is complicated by the fact that in 5E you can have a multi-classed Fighter/Wizard as well as a Fighter (Eldritch Knight), and the two will have different abilities and will likely play differently, the same being true for Rogue/Wizard and Arcane Trickster.

Nevertheless, I think it’s the right approach to add these subclasses to the table at the frequencies at which F/M-U and M-U/Th would be produced using the multi-classing frequencies given in 1E. I’m going to edit the tables in my post up-thread to reflect this change.

To compensate for moving these two dual-class combinations into the base table, I would reduce the base frequency of multi-class, from 1 in 10, to 1 in 12 and increase the chance of triple-class, from 1 in 4, to 7 in 20.

If rolling for chance of multi-class based on race (as I’m inclined to do), I’d use the following:
Race
Chance of multi-class
Dwarf
11%​
Elf
61%​
Gnome
18%​
Half-Elf
61%​
Halfling
7%​
Half-Orc
36%​
Then if a character is determined to be multi-classed, they have a 35% chance of being triple-classed.
 

Hriston

Explorer
Remember that in AD&D only elves and half-elves can be triple-classed. Dwarves, gnomes, halflings and half-elves are confined to two classes.
I think this is a good point, but I don’t see the fact that only elves and half-elves can have three classes as the result of a general rule in the same way that only non-humans can multi-class is. I see it more as the result of there being only certain allowed class combinations and that only elves and half-elves (and humans) can be magic-users.

I’ve deliberately left racial class and multi-class combination restrictions out of this conversion because I feel they go against the more modular design ethos of 5E, where, for example. half-orc paladin and sorcerer/warlock are considered valid options. I don’t want to exclude results like that, which is why I don’t share @Lanefan's concern about paladin/assassins. I could always play 1E if I wanted those restrictions. So while my preference is to limit multi-classing to non-humans, I’m not going to limit specific class options based on race or limit what combinations of classes can multi-class, which may seem like a half measure, but it’s one I feel good about.

I edited the tables up-thread to reflect bringing in AT and EK at frequencies that I figured M-U/Th and F/M-U would appear if race-based restrictions on class are ignored. My goal is a set of tables that doesn’t assume multi-classing is being used but is compatible with multi-classing, which can be applied at a standard frequency to all characters regardless of race (about 8%), to all “demihumans” (about 38%), or at the frequencies for each demihuman race I gave up-thread.
 
Last edited:

Advertisement

Top