Jeffro's Space Gaming Blog
Microgames, Monster Games, and Role Playing Games
The Unbearable Lightness of Binomial Distributions
Well, Santa brought a delectable vintage game book for me this year: the Cook/Marsh Expert D&D rules! It is insanely good, from the notes on how to use it with the Holmes edition, to the suggestion to Game Masters of the day to just roll their own rules for handling levels 15 to 36! Yes, it is indeed a time of peace on earth and good will to all men: Cook/Marsh uses the six mile hex for the wilderness map that places The Haunted Keep in among other goodies.
Brendan is taking some of the suggestions for extrapolating out levels 15-36 and applying them to his own rendition of the Thief class. As usual, my impulse to put in my two cents has spun out of control into a veritable orgy of bullet points:
- First off… +n% per level is inherently lame. It should immediately remind anyone of the Palladium system.
- Secondly… if you’re going to go to the trouble of a point-buy system for thief skills, just know that you are squarely into GURPS territory.
- There are three levels of skillfulness that matter at B/X D&D’s level of granularity. “Smarter than the average bear,” where you use the same mechanic as everyone else but have a +1 bonus or some other edge. “Good enough for government work” grants you a new ability… usually with a 60% +/- 5% chance of success. After that, there is “game breakingly awesome,” where skillfulness lies in the 90% +/- 5% range or else there is some other epic effect involved.
- Let the table note that non-human classes pay for their mixed bag of bonuses and abilities with severe level limits and painfully slow advancement. Might I suggest that this is the reason for the thief’s 1d4 hit dice…? Isn’t leveling up before everyone else worth something?
- Anyways… I said all of that so I could tell you this: those ten thief skills that Brendan has worked out…? I suggest borrowing a page from calithena’s Dragon Blooded over-class from Fight On! #2 and breaking the ten thief skills out into two separate tiers of awesomeness. Each time the thief levels up, he will roll 1d10 to either gain a new ability, move an existing ability to the second tier, or gain an extra hit point beyond the usual crappy +2 per level beyond level nine.
- I haven’t checked the math or run the Monte Carlo’s for this, but the idea is to keep things interesting all the way up to level 36– not that anybody necessarily gets that far. Brendan’s current draft has you (in effect) rolling d10, then d9, then d8, then d7 and so on. A binomial approach like mine would be cleaner… and also force you to wait a while longer for the game-breakingly awesome stuff.
- Also, under Brendan’s system, you know exactly where you’ll stand when you get to level ten. I prefer a little more uncertainty: the chance of getting to tier two in something, balanced by the chance of crapping out with a measly hit point if you’ve already done that before.
Jeffro's Space Gaming Blog 
Thanks for the feedback; this is all really useful.
On 1d4 hit dice: yeah, I can see that. Perhaps this is my 2E origin showing. 1d4 hit dice for anything other than magic-users just feels strange. I should probably use 1d4 though to keep divergences from B/X to a minimum, since that is the explicit baseline.
I like the idea of the “prestige abilities” and that would allow me to work in things like spider climbing on ceilings. The downside is that the explanation of the abilities becomes a bit more complicated and will take up more space on the page (I was hoping to fit the entire class on one sheet). I may try it though. It’s also not clear that all abilities would have a natural progression.
If I were ever to run a game that might get much higher than level 10, I would agree with your last point. Really, I tend to consider the Expert Rules level cap of 14 to be just about ideal. Clerics have access to their most powerful spells by 7th level, magic-users by 11th (I prefer the OD&D and B/X cap on spell levels: 5th level spells are the highest for clerics and 6th level are the highest for magic-users). Halflings, dwarves, and elves have level caps of 8, 10, and 12 respectively. In a campaign that might reach to levels in the 20s or 30s, playing a demi-human would never be an attractive choice. I do prefer a humanocentric game, but if I’m going to offer the demi-humans as options, I don’t want them to be completely outclassed. Everyone stops accumulating full hit dice at level 9. So having all the thief abilities at level 10 does not seem that bad. The only potential problem I see is that of homogeneity: will all 10th level thieves seem too similar? Fighters are going to have different equipment, magic users different spells. All clerics of level N, however, are more or less the same in terms of game mechanics (or, at least as similar as thieves would be under my rendition) so this does not seem to be that bad.
Another idea that just came to me: what about offering a reversed version of each thief ability as well? This would allow more individual flavor. The reverse of climb walls might be tumbling (falling without taking damage). The reverse of poison brewing would be antidote brewing. The reverse of move silently would be super-hearing. Etc. Also allows an interesting variation between offensive and defensive abilities.
Also, once characters have reached name level, I expect them to settle down a bit. They probably are in the process of establishing a stronghold. In Traveller (as in, the sci-fi game) terms, upon reaching name level, PCs have finished the character creation process; at that point, they mostly improve by accumulating social status, domain relationships, and gizmos.