What I don't want to do is take away any player agency from the process. But choices take time. Do I want to be fighty or sneaky or spell-castery, and anyway why can't I be an Elf and a Paladin as well? What do you mean Elves can't have classes? What, there's no Paladins anyway? OK, an Elf then. Shall I buy leather armour, or can I afford chainmail? Does that mean I can't afford a lantern? Shall I get a 10' pole? What are my spells? What languages should I speak? And so on.
Players in some ways get agency in the 'wrong' places. We don't as real humans in meat-space chose how we are born and brought up. Character creation from stats up is not like 'growing up'. Something closer to 'reality' would be if players were given character sheets with Race (and sex and handedness if anyone cares) already filled in, 2d6 for all stats marked in the boxes (2d6 actually rolled, in order, rather than just saying '7' for each or whatever) and starting gold already calculated.
The stats could represent (in a somewhat schematic way) the interaction between genetic determinism (OK for the first round of rolls you were STR 2 INT 2 WIS 4 DEX 3 CON 2 CHA 6, that's your basic 'this is what I got from the genes' result when you're born), and the environmental determinism of your background (STR +1 INT +3 WIS +5 DEX +2 CON +1 and CHA+3 means that your upbringing emphasised empathy and perception over brute force; you also had a bit of cleverness and persuasion in there).
The player could then roll 6d6, and chose which roll goes with which stat. Currently they're on STR 3 (max 9) INT 5 (max 11) WIS 9 (max 15) DEX 5 (max 11) CON 3 (max 9) and CHA 9 (max 15). They're never going to make a Fighter and they're going to be a mediocre Magic User or Thief this way, but they might make a decent Cleric.
So you roll 1, 4, 2, 2, 4, 3 (16 in total)
This last round of rolls would represent the choices you made when you were growing up - what you prioritised which might be the same or different to the way your parents pushed you (maybe you don't want to be a Priest after all and you put a 4 on INT because you really want to be a mediocre Mage).
These are genuine rolls. They're also terrible. Average of 18d6 is 63 - this is the number that is the centre of the bell-shaped curve for the total of all stats. This character actually has a stat total of 50. They've had a really unfortunate life to this point I would say. Their genetics gave them 19 (average at each stage is 21), their environment gave them 16 and their choices gave them 15, so at each stage they're below average.
STR 4 INT 15 WIS 10 DEX 6 CON 4 CHA 10 (on the basis that each stat must have 1 point assigned and the any remaining points can be assigned to a single stat, up to 18)
- which could indeed represent someone from an (underprivileged) background who's being pushed to the priesthood rejecting it and dedicating all of their (not earth-shattering) talents to becoming a Magic User. Incidentally swap INT and DEX (both were 5 after stage 2) and change the narrative to 'joined a criminal gang' and you have a thief. Not a very tough one but adequately sneaky.
Anyway; this doesn't really help speed things up, because any kind of point-buy is going to be longer than just determining numbers with dice. Choices take time as I said earlier. But if you want a longer creation-system that gives some idea of the interplay of inheritance in terms of genetic background, environmental influences on development, and agency related to 'teenage choices (insert equivalents for Elves, Dwarves and Halflings)', then splitting the rolls and recording them separately might be a good way to go.
In the mean time, with the system as is, players get to assign agency to aspects of character creation that they don't have in reality, but then again, this is just a game and forcing players to play characters if they don't have some control over the creation process can result in players playing characters that they really don't want. That, I would think, would somewhat impact on the fun.
So is there a way to make all this quicker? Perhaps pre-gens would be useful: not necessarily 'here's your character with name, eye colour, sex, armour, weapons and favourite pudding filled in' but more like 'you want to play a Cleric? Here's a sheet for a Cleric with stats, HP and the 1st-level Clerical abilities already filled in'. More work for the DM but who counts prep-time anyway? It's playing time that's important. Anyway, DMs have little else to do but mess around rolling dice (or writing blog-posts about rolling dice).
So in short for the stats side of things I'm not sure there's a way of shortening the process without cutting down player agency, which even if it's not particularly realistic is probably best avoided.
Is it any easier for equipment, spells and languages though? Let's say we somehow have a Magic User with an Intelligence of 18 (perhaps she assigned everything she could to INT and put two other stats up by 1 each); she speaks 3 extra languages, has one Lvl 1 spell slot, and needs to buy equipment (she has 110GP). That's a lot of choices to be made.
The language table in B/X can be randomised. Roll a d20 and look at the table (then do it twice more for your 3 languages). Is that quicker than just picking one (or three)? Not sure. Does it take away agency? Yeah, maybe. What about saying 'if a MU knows one extra language, it's always Elvish, if they know a second, it's always Dragon'? Does that make things faster? Sure, but it looks like it's taking away agency.
But what is agency anyway in this instance? Agency depends on informed choice. Picking options if you don't have an idea of consequence isn't actually agency. How do the players know what languages will be important? If the DM knows half of all scrolls are written in Elvish, a quarter are written in Draconic and the rest are split between all the other languages randomly, then maybe it makes sense for a Magic User to learn Elvish, Dragon and another tongue, in that order. On the other hand, maybe they pick Goblin and Orc (handy to talk to the foes after all) and Gargoyle (on a whim/long shot) but find out that the first dungeon they go to is overrun with Kobolds enslaved by Gnolls led by a Doppelganger.
It's the same thing with equipment. Will our Magic User need Wolfsbane, Garlic and a Silver Dagger? Depends on how many Undead and Lycanthropes they're going to meet. But they don't know that when they start, so where is the agency? They haven't scouted out the terrain and found out if there are werewolves and vampires in the area, they're just picking blind. Does she need a mirror? The player doesn't know if Medusae and Basilisks are going to feature. A 10' Pole? How can the player tell?
Is it of any benefit it the player to second-guess the DM like this? It seems to me that this is merely the illusion of agency. "Haha!" thinks the player, "I have bought Wolfsbane, Garlic, a Silver Dagger and a Mirror, I'm been very crafty here and I am going to be fine whatever the DM does!" but spends the next 10 sessions plundering Gryphons and Minotaurs and Harpies but unable to carry all the loot home because they didn't buy enough Large Sacks.
I don't know what the answer is. I'm going to experiment with a variety of 'adventuring packs' that contain some useful tools. But I might just be reproducing the complexity at a different level of granularity. "Good news! You don't have to make 10 random choices from 50 possibilities, now you just have to pick one random possibility out of five! What's that? The detail? Let's unpack those five possibilities... gosh, it would have just been easier to get you to make your own pack wouldn't it?"
Last weekend in my new Rift City campaign, the players talked a lot about rope. "But the DM knows rope comes in 50' lengths, I bet he's put a 60' drop in somewhere, we should get 100'." One of them actually said this right in front of me. All I'm going to say here is that a 60' descent would probably take them to level 4. I don't need the fall to kill them in that case, just being on Level 4 would be enough. They're 1st Level for Monkey's sake, most inhabitants of Level 4 would have them as a midmorning snack then go looking for some real food.
To be honest, I can't really see why they need more than 15' of rope at the moment, probably the worst that might happen is one of them falls 10' into a pit (if they're not probing with the 10' Pole they might not have).
Maybe the way to do it is to split equipment into 'recommended' and 'optional' lists. So food, light and carrying capacity feature heavily, Wolfsbane and Mirrors and whatnot are nice to have but not as important as say a lantern and oil, some rations and a backpack. But how to do this without blowing the surprise? If I tell them that they don't need Wolfsbane, that pretty clearly says there's no werewolves. If there are werewolves, I'm a big fat liar, because they do need wolfsbane (or at least, it's bloody useful). So there'd better not be any werewolves. But why would I want to tell them there's no werewolves? Anyway, I want there to be werewolves, so I should recommend wolfsbane, Now I've pretty much told them there are werewolves, because if I get them to spend money on wolfsbane but there aren't werewolves, I'm just being an asshat.
I don't know, I'm doing a lot of thinking out loud here but not getting very far with the notion of making character creation easier while still maintaining agency (or increasing it! Meaningless choices are not agency!).
Rumours. Maybe they're the way to go. Give the players their rumours before they buy their equipment or make choices about optional abilities. They've heard there are werewolves in the northern forests. If they wanna go to the northern forests, it makes sense to get wolfsbane. If they're planning on going to the southern plains, they might not want to bother. They know there are lots of undead in the Dungeon of Doom, so maybe 'Sleep' isn't the best choice of spell after all if they're planning on heading there. They know that the Wizard Krufelfnuffer was a prolific maker of scrolls and fond of writing in Draconic, so they chose that language, rather than Gargoyle or Bugbear.
Maybe I just need to combine all of this. If I can come up with a 'Dungeoneer's Pack' (lantern, flasks of oil, rope, iron spikes, hammer, 10' pole) for those on spelunking missions and a 'Monster-hunter's Pack' (Wolfsbane, Garlic, Mirror, Silver Dagger - Clerics get a Holy Symbol and a 5GP donation to the Temple instead - and Holy Water) and a few others based on what sort of missions they think they might go for, that might help I suppose. Knowing in advance what common monsters might be will also help ("Orcs? I heard of 'em, over Thrarp way, but we ain't never not seen none round 'ere, no we never bain't have done, it all be Skellingtons and Zombees round here... yup, you heard, giant undead bees."). Some legends about history will also be useful (such as stories of the Monks of Gargoyle Mountain or the Lizard-Wizard of Wonky Swamp or Kronge the Infernal Trapmaster) to help give an idea of the hazards and opportunities they could meet... and this is before they even pick spells or languages. I'm reminded of a post from 3 years ago now (have I really been doing this that long?) where I try to make sense of how to pattern information in campaigns: it's obviously important to me how players know things but again it doesn't help speed things up.
I dunno, maybe I'm overcomplicating things. Perhaps I should just put up with it all. Character creation is long because that's the best way to give players a sense of ownership and that's about it. So what if they speak Gargoyle and never meet any Gargoyles? That means they have an unusual skill - that's a feature, not a bug, in role-playing terms. Not everything should be optimised or even immediately 'useful' to a character.
I'll stop this here, I'm not sure it's helping me at all to make things simpler...