Last night I was involved in a conversation with someone who wanted to run a Pathfinder game on an episodic basis: the PCs as 0-level characters, then third level, sixth, ninth, and so on. Each episode is set a few years after the previous one. There is continuity of character, but each may have things happen ‘between adventures’.
There was some thought about how the 0-level characters and play might be structured, and I suggested he look into Dungeon Crawl Classics… and especially the funnel.
The Funnel in DCC consists of each player bringing three or four quickly-developed (ability scores, roll on an occupation table for occupation and starting gear, go!) characters that are quite expendable. They go on an adventure they are poorly equipped for, woefully poorly in fact, in which they they find some fortune. Some is good (coins! jewels! wondrous devices!), some not so good (I had three drown in the same mud pit, dammit). ‘Wealth by attrition’ is a thing, though, so dead PCs let you concentrate your starting equipment. I’ve found playing the funnel to be a lot of fun because of how clever you need to be to survive, and that it really gives shape to the characters that survive. I don’t want a steady diet of it, but it works as a nice start to a campaign — especially since you can use up potential-PCs without feeling bad about it, while getting a feel for certain campaign elements.
Then I considered how a funnel might be constructed and applied in Pathfinder, and I came up with the following guidelines:
- Each player gets three or four
- As with DCC, roll stats (3d6 in order… Pathfinder doesn’t usually do it this way, but I’ll explain more below)
- Choose a race, but perhaps don’t apply any racial traits. Right now it’s just for show.
- Each character is a first-level commoner. Raid the DCC tables or similar for occupation (which would determine starting skills, though I’d probably identify them during play — “yeah, a scribe might have a rank in a knowledge skill” — rather than try to nail them down up front) and starting gear.
Now go on the first adventure, use up some of the fledgling adventurers (remember, wealth by attrition), and when you’re done you pick a survivor for your first-level character.
- For each ability score, roll another d6 and drop the lowest of the four dice (the three original ones plus the new one).
- Swap two scores if you want.
- Apply racial modifiers and traits.
- Choose your class.
- Roll hit points as usual (maximum at first level) and add the ‘commoner hit points’ as a one-time bonus. You’re replacing the Hit Die (you are not a second-level character, so you don’t have two Hit Dice).
- Choose your skills and feats. You can keep the skills and skill points previously gained from the commoner level, but (having only one Hit Die) do not get a higher number of ranks. I think I would allow you to keep those skills as class skills though, regardless of class, so you get +3 to all of them. You also gain proficiency regardless of class with any weapon you use and carry at the end of the funnel.
- Apply any goofy stuff you may have picked up during the funnel.
- Spend your loot and carry on as a blooded adventurer!
This should allow the funnel to be the charmingly amusing experience we enjoy in DCC while not shorting the PCs power (ability scores) once they are full-fledged adventurers. The PCs have ability scores roughly in line with 4d6 drop 1 — the loss of ‘arrange to taste’ is, I think, mitigated by being able to swap two scores and by having developed something of an image of the character that isn’t based on ability scores. Slightly unoptimized perhaps, especially compared to point buy (which I’ve honestly never much cared for), but workable.
There are some possibilities for streamlining this a bit more:
Roll each character’s starting race rather than picking it.
Either each occupation provides a fixed set of skill ranks, or else identify the character’s skills during play as you mention.
Generally, I prefer to avoid a situation where we’re making choices before play: funnel fodder are generated rather than constructed.
The purpose of a 4d6-pick-3 roll is just to introduce a skew towards the higher ability scores, but running the funnel fodder through a dungeon will have already introduced a skew by tending to weed out those with lower scores. On that basis I would skip the post-funnel adjustment to ability scores.
I would prefer not to swap any ability scores post-funnel either: it’s hard to see what this would mean in in-world terms. Again, any really unusable ability score distributions will tend to get weeded out during the funnel itself.
You might think better ability scores survive more, but I think it is not enough to bring the average up that far. In my experience nobody is safe, even if you try to insulate a promising-looking character. One “every make a save” can thin the herd pretty fiercely.
I think I’d keep the extra die on ability scores (you became stronger, gained confidence in your abilities, etc., whatever), but agree that the swap may be unneeded.
Of course nobody is completely safe, but then neither does 4d6-pick-3 completely prevent you starting with a Constitution of 3. But that’s just the nature of statistics: individually unlikely things can happen, but check the overall distribution.
The skew from 4d6-pick-3 isn’t actually all that big, only shifting the mean ability score by ~1.75. I checked it on a graph, and 3d6+2 has a pretty similar probability distribution, apart from the ends not matching up.
I guess I have an existing prejudice of having always seen 4d6-pick-3 as an unfortunately necessary hack to get more survivable characters. That means I’m more likely to favour using something like the DCC funnel to replace it (“you survived, statistically you’re more survivable!”).
Except it’s “statistically you’re more survivable because of lucky dice” (or clever play because you’re otherwise so unsurvivable, rather than honestly deserving it through natural ability (high scores).
It’s a bit like “stuff used to be made so much more durable back in the day”… it wasn’t, you just found the ones that didn’t get wrecked. They were probably not significantly different from the ones in the same manufacturing batch, they just spent their duration in a safer place.
Yeah, I just want a different way of doing things. It’s especially prominent here where you need to remember your lowest die roll for each ability score so that you can do the adjustment later. :-(
I’m not entirely happy by that bit myself, but still I want to move the ability scores more or less into where they might be in a normal Pathfinder game. It might be simpler to just add +2 to each with a maximum of 18 (3d6 mean is 10.5, 4d6c3 is 12.24, and I think the standard array has a mean of 12.17)… probably close enough.