Chesterton’s d20

AD&D as Crunchy OD&D Case Law

If you read Peterson’s Game Wizards you see one of Arneson’s many failed projects was an index for D&D. This was necessary because within a few years finding rules spread across the white box and the three supplements was difficult and confusing. Unfortunately for Arneson, the publication of Advanced Dungeons & Dragons rendered the effort both superfluous and obsolete as soon it was published. A big part of what AD&D did was take all the stuff Arneson indexed and put it in a coherently organized fashion. Well, coherently organized relative to original D&D and its supplements if not relative to such later revisions of AD&D as AD&D 2e, OSRIC, or Hyperborea. Of course, AD&D was more than reorganization but also included a lot of fiddly bits, which is why Swords & Wizardry: Complete (retroclone of OD&D + supplements) and OSRIC (retroclone of AD&D) are similar but different games.

Those fiddly bits AD&D added to OD&D took two broad forms: simulationist tinkering and what I think of as D&D case law to balance overpowered game elements. This raises the issue of whether AD&D’s crunch makes for a better game. If you’re such a hardcore simulationist that it drives you to nerd rage if maces aren’t better than longswords against plate armor because that’s how … then that’s one thing but I am thinking more about fiddly rules that don’t seem to make sense which those of us who played in the 1980s mostly ignored. Do you need those? Or put differently, what style of gameplay does keeping all that crap in encourage and is it a fun style of gameplay?

I recently joined an AD&D game that attempts to run the game as closely as possible to rules as written, including the parts most people skipped back when AD&D was the current edition. The premise of the game is that maybe all that weird crunchy stuff is in there for a reason and the only way to find out is to run it RAW and find out what the gameplay is like when you keep casting segments and weapon modifiers vs specific AC and all that. Just as you shouldn’t tear down a fence until you learn why it exists, you can’t discard casting segments or disease tables until you learn what they’re for. I’m skeptical, in part because by all accounts Gygax himself ignored most of that stuff, but I respect the experiment and the DM does a good job with keeping track of the system crunch and explaining it as necessary.

The Problem of Training Costs

One of the rules that’s proven to be very salient is training costs. Training costs seem like a game balance mechanic. One of the notorious problems of GP=XP is once the PCs scoop up enough XP pellets to level up, they have enough money to last them years (at least assuming they have a diet low in garlic), so if they’re really murder hobos why do they keep heading into the dungeon looking for treasure instead of retiring to a villa where dancing girls feed them peeled grapes? This implies the need for some kind of mechanic that bleeds the PCs dry. The modern OSR solution takes a page from the sword and sorcery source material and introduces a carousing mechanic: you can keep the money if you want but you level up faster if you squander it drinking, gambling, and, whoring.

The Gygax approach in AD&D was to nickel and dime the PCs. There is a bunch of stuff about taxes (DMG p. 90) but the two most routine things are “expenses” and training. Living expenses are (100gp x level)/month, which implies a standard of living 20x-100x more opulent than an ordinary laborer (DMG p. 25 vs 28). Training is level x 1500gp and that’s assuming you don’t have a penalty multiplier for being a crappy role-player (DMG p. 86).

Now let’s think through the implications of that. Suppose there is a party of four level one characters: Tessie the thief, Calvin the cleric, Fergus the fighter, and Marv the magic user. They sneak their way through level one of the dungeon and prudently avoid both traps and combat (and therefore earn no monster XP) and via cunning combat as war play, they draw the monsters away from the treasure and yoink. They roll really well on the treasure table and it turns out the treasure therein is worth 5600 gp. The whole way back to town, Tessie can’t shut up about “woo hoo, we’re on a GP=XP standard and I’m a thief which means even with my mediocre dexterity I have enough gold to level up.” They get back to town, a bona fide safe haven and Tessie can barely fall asleep she’s so excited at the thought that come morning she’ll have an extra d6 hit points and a 15% chance to hide in shadows!

Morning comes and Tessie realizes that she’s not just mostly useless for anything but climbing walls but utterly useless for anything but climbing walls. Still 1d6 hit points and 10% chance to hide in shadows! What gives? She heads down to the common room of the tavern and finds an old man in a pointy hat sipping his breakfast ale and says “hey NPC, you look like you’d understand this, why haven’t I leveled up given that my share of the loot was 1400 GP which means I should be level 2 and 1/8 of my way to level 3?” The NPC laughs and explains to Tessie that the thief XP table on page 27 of Player’s Handbook just means that she is eligible to level up but she doesn’t actually level up until she pays for training which will be 1500 GP this time and 3000 GP next time and 4500 GP the time after that. Oh, and by the way, you don’t earn any more XP until you pay the training. Around this time Calvin the cleric is wandering down to the common room and has caught enough of the conversation to regret putting down a deposit at the armorer on plate mail and tossing platinum pieces to the local beggars. Tessie and Calvin look at each other and realize that there is no way that Tessie can level up faster than Calvin and Calvin himself can only hope to reach level 2 faster than Fergus the fighter if he doesn’t spend a single copper piece on armor, hirelings, horses, or holy water, but rather he saves it all for training. And for reaching level 3 it doesn’t matter how tight-fisted Calvin is, he ain’t reaching it faster than Fergus. Moreover, both Tessie and Calvin better hope they get that treasure fast because come first of the month they each have to pay 100 GP in rent or feasting or alms or wizard cocaine or whatever it is that constitutes their monthly “expenses.”

My view of this is that it is simply bad design. Old school D&D has variable XP and presumably the reason for that is to balance classes. Go to any OSR discussion board and you won’t have to look hard to find some version of “old school thieves suck but the good news is they level up quickly.” Apparently people who say that have never run AD&D RAW because if they had they’d know that the 1250/2500/5000/… thief XP table is a joke and the real thief XP table is at least 1500 for level 2, then at least 4500 for level 3, then at least 9000 for level 4. I say “at least” because I’m assuming no monster XP, no prime requisite XP bonus, and no living expenses. If you’re running AD&D RAW and you factor all that crap in, a thief or cleric’s real XP table is closer to that of a fighter or a magic user than it is to their nominal XP table.

I checked two notable attempts to refine and reorganize AD&D and noticed that both of them greatly relax the AD&D training rule. AD&D 2e DMG explicitly flags training as an optional rule then says if you do use it, the training fee is at DM’s discretion but defaults to 100 GP x level per week where number of weeks follows a complicated algorithm but is probably about 7-10 weeks. The AD&D 2e training rule is expensive, but manageably expensive. The downside is that it implies it’s a huge pain in the ass to find an appropriate trainer and you’re supposed to role-play the whole thing (which doesn’t sound like as much fun as fighting ghouls). The other system I checked was Astonishing Swordsmen and Sorcerers of Hyperborea (2nd ed). Hyperborea gives the PC the choice between 100 GP/week where # of weeks equals the new level or leveling for free through independent training at two weeks/level. Thus in AD&D 2e it would probably cost Tessie the thief about 700-1000 GP to train up to 2nd level and in Hyperborea it would either cost her 200 GP and a week of downtime or no money plus two weeks of downtime. The AD&D 2e rule might still make it tricky for Tessie the thief to reach 3rd level before Calvin the cleric but both of them will be consistently leveling up faster than Fergus the fighter or Marv the magic user. All this suggests to me that both Zeb Cook and Jeff Talanian considered the AD&D rule and decided to relax it so as to make training a wealth drain, but not an obstacle to the extent that it effectively supersedes the XP tables.

I also searched the blogs and found surprisingly little discussion of a rule that, if run as written would effectively supersede the XP tables which suggests to me that most people ignore this rule. @John_Cyrano on Twitter noticed the same problem I did with thieves. The best sustained argument for the training rule I saw was from The Blue Bard but he also suggests changing the requirement from GP to SP which certainly solves the problem of thieves and clerics aren’t really leveling faster than fighters and magic users. On Fantasy Grounds, Leo Zelig said his group runs training RAW and they make it work via horse-trading between the PCs.

What Play Styles Are Encouraged by Training Costs

Entirely in the spirit of “hey, we’re supposed to be testing what works if we run RAW so we can keep doing it but you can also go ahead and put this mechanic on the broken list” rather than “change this so I can level faster,” I raised this issue with the AD&D RAW group and got some “wait a minute, maybe it works in ways you’re not considering.” The other players suggested ways that the “LOL, you think you get to level now” nature of the training rule can create emergent properties of gameplay.

Patrons. There’s a blink and you miss it reference in DMG (p. 86) to the possibility that trainers “might possibly accept some combination of gold and service, at the DM’s option.” This means that training costs are effectively a patron play mechanic, especially for clerics and thieves as they are the classes for whom the rule is not just a wealth sink but a leveling brake. And this works with the fiction since churches and thieves guilds are both established factions that can serve as patrons to the players. This also implies that either breaking from or being isolated from the patron will carry a heavy cost in terms of advancement which may be a bug or a feature depending on what type of play you’re trying to encourage.

Intra-party bargaining and strategizing. The party can decide who levels up first either by consensus (“we all agree we need the thief to be more survivable so she can scout”) or by a market mechanism (“I, the thief, agree to let the fighter keep the magic item if he lets me have his share of gold”). Closely related to this is that the magic user doesn’t need to spend all his money on special scroll ink like in B/X because in AD&D the magic user probably gets extra spells from an intelligence bonus.

A darker version of intra-party trading is PVP play where the thief either steals outright from other party members or filches dungeon treasure before it can be divided up by the party. The mechanic definitely encourages this and it might work for the fiction but it’s usually agreed that PVP thievery is among the more destructive forms of play.

Another version of intra-party trade relies on the fact that training costs mostly increase at a slower rate than do XP demands. I made a little spreadsheet showing training costs, cumulative training costs, XP table, and the difference between the XP table and the cumulative training cost. You can think of this last bit as “how much is training slowing you down if all your XP is from treasure, you don’t get a patron play discount on training, you don’t get a prime attribute XP bonus, you don’t have monthly expenses, etc.” The upshot is that training costs most meaningfully inhibit leveling below about level 7 and after that it’s an annoyance, but not an obstacle. This implies that if you have a mixed level party (presumably because it’s a long-running game and both new players and replacement PCs start at level one) then the high level PCs don’t need the money as much and can effectively serve as patron to the lower level PC, possibly by training them personally but more likely by loaning them the money for training. In my opinion this kind of intra-party patronage mitigates the problem of the training rule but doesn’t actually make a positive case for it. Moreover it introduces issues like what if the experienced PC doesn’t trust the noob to stick around? Paying to level up a thief or cleric might be a great deal if they stick around but if there’s a high turnover open table experienced PCs might rather deal with a more reliable NPC henchman.

The Verdict

I’m glad my DM is running the experiment of running AD&D RAW since it’s very interesting to see what the emergent properties of play are when you try them. However my personal reading of the experiment’s results so far is that it’s a fun game despite, not because, of the crunchy bits you find in AD&D that aren’t in Swords & Wizardry or B/X. Gygax was a good DM and a pioneering designer but not a designer who fine-tuned things. Much as 3rd edition D&D simplified some aspects of TSR D&D (a unified d20 resolution mechanic and unified XP table) but added crunch in other ways (feats), Gygax’s own second effort vs his first (AD&D vs. 1974 OD&D) both made some aspects of the game simpler (not presuming wargamer tacit knowledge) but made other aspects more complicated (spell components, casting segments, long-ass spell descriptions). Ad fontes is laudable as gamer experimental archaeology, but the modern OSR really has improved on the original through such mechanics as carousing rather than training. My opinion is only reaffirmed that the OSR consensus is right to use B/X as the framework and treat AD&D (especially DMG) as a great source of inspiration rather than something to run RAW. If the experiment shows that, I dunno, weapon speed rules, provides really interesting emergent gameplay, I’ll keep you posted.