2021-10-25

Early Evolution of Encounter Text

A number of times in recent episodes of the Wandering DMs D&D Sunday talk show we've wound up debating how much detail is best for published adventure text. So I wanted to share some of the possibilities running around my head from classic D&D products. In particular, for what follows I'll present a snippet of an adventure authored by Gary Gygax, graduated in 2-year intervals throughout his tenure as the "boss" of D&D (1974-1985). I've also intentionally tried to focus on his higher-level adventures, which composed most of his published output, so as to compare like-to-like in as many cases as possible (i.e., I intentionally skipped Keep on the Borderlands for this reason). Here we go:

1974 – Castle Greyhawk Notes


Okay, so this one wasn't exactly published, and it also isn't high-level. But we managed to get a look at Gary's notes for running Castle Greyhawk's first level when Matt Bogen (Eridanis) caught a photo at a special Gen Con 2007 event. As you can see, it's only a single curt bullet-line per encounter area, noting monster type, number range, and treasure (and absolutely nothing else). In fact, it's more stark than that, because multiple rooms on the map are denoted with any one given key number (often a half-dozen or so); and about three-quarters of the rooms have no key whatsoever, being simply empty (in line with the design advice from OD&D Vol-3).

Most everyone involved agrees the map layout is Gygax's original design for Greyhawk from the early 70's (could be 1972-1974 or so), developed in conjunction with the D&D rules; but there is some debate on whether the encounter text is original or not (maybe changed up for convention play?). Personally, I'm pretty confident that the text seen here is either original or in the same basic format -- if Gygax was so famous for improvising things mid-game, what would be the point to rewriting these minimalist text notes? Big thanks to Alan Grohe (grodog) & Zach Howard (Zenopus) for managing to decode that fuzzy photo taken from over Gary's shoulder.

1976 – Lost Caverns of Tsojconth Tournament

This is a snippet from the document Gygax wrote for DMs running the D&D tournament at Winter Con V, which occurred at Oakland University in 1976. Rather than bullet-point atoms, here we at least get full English sentences describing each area (usually 2-3 per encounter). But the emphasis is largely the same: the very first piece of text per area gives the monsters and hit points; this is followed by a small bit of descriptive text, and ends by denoting a treasure value (possibly none). Instead of a numeric range for monster number appearing (as in his Greyhawk notes), the monster numbers are now fixed, and their hit points are listed in advance. Arguably this might be motivated by the tournament situation where the strength of encounters should be as fixed as possible across different tables. We should also note that the back of the document has a table (one page per dungeon level) with summary statistics for all the monsters present, including number, hit points, move, attacks, and specials -- and even a "hit bonus" and a target number to hit each PC in the adventure, which is equivalent to ascending-AC in the d20 system (!).

 1978 – Steading of the Hill Giant Chief

By the time of the GDQ (Giants/Drow) series, Gygax now starts providing a name for each keyed area (instead of just identifying it by monster type). The narrative descriptions are somewhat more textured, and the paragraphs tend to lead with those sense-descriptions -- with identification of any monster, and their hit points, possibly occurring deeper in the block. (As a result, it's possibly easy for the DM to overlook the monster present in room 2. above, say). Again it tends to conclude with a discussion of treasure, with somewhat more detail given to that description as well. 

1980 – Expedition to the Barrier Peaks

AD&D module S3, Expedition to the Barrier Peaks, like most of Gygax's famous adventures, was birthed at a tournament event -- in this case at the Origins II game fair in Baltimore in 1976. Four years later, we got the elaborate published product, in full-color, and with a copious illustration booklet. Note that it exhibits similar DNA to the 1976 Tsojconth tournament text: areas are often identified only by the monster, the text is very mechanically terse, it concludes with a treasure, etc. (e.g., areas 5. and 6. above). But in some other cases we have a blossoming of more deeply detailed areas, even without any monsters or traps at all; for example area 7. above, the "Ship Commander's Quarters" -- which actually continues to sub-areas a., b., c., and d. (none of which have any monsters or unnatural contents), almost 500 words total for the one area. Also we see here an expansion of the monster statistics within the paragraph text -- not just hit points alone, but now also armor class, movement, hit dice, attacks, and damage. In some case these notational statistics take up a larger proportion of a paragraph, possibly making for a choppier read.

1982 – Lost Caverns of Tsojcanth Publication

The 1982 publication version of Lost Caverns of Tsojcanth gives us a golden opportunity to compare to the earlier tournament module (above). The maps in question closely match; from a quick survey they appear identical (despite lots of tricky, crooked, freehand-caverns). Areas 2., 3., and 4. in the 1982 publication (above) are the same caves as B., A., and D., respectively, in the earlier 1976 adventure (further up). We can see then that many of the contents were changed, moved, added, etc. Stirges in 2/B are basically the same. A flesh golem in D was changed to a clay golem in 3 (formerly A), the back-map removed. Blink dogs are removed entirely and the new mobat monster inserted. And so forth.

Aside from those content changes, the text is again more elaborate than in the tournament document. Note that the brand-new encounter with the mobats gets significantly more verbiage than some older encounters, say. We again have a name for each area, and a tendency to start the description with a monster, and end with a treasure. Monsters have expanded parenthetical stat blocks, including possibly extended description of special abilities. And hey, there's boxed text for the first time! 

Another aspect we should point out is that compared to the earlier version, S3 gets an added extensive wilderness adventure section through the mountains before the Lost Caverns can be explored. This runs 9 pages, including 3 pages devoted to a gnome vale (complete with large underground lair maze/complex) intended for use by the PCs as a secure base for rest & rehabilitation between forays. Two of the really interesting differences between this section and the dungeon text is that: (1) while the dungeon areas use boxed text, the wilderness does not, and (2) while the dungeon has parenthetical monster stats mid-paragraph, the newer wilderness areas remove them to separate stat-blocks outside the paragraphs. Personally, I find it really interesting to be able to pinpoint in this one document the moment when monster stat blocks grew too big to comfortably fit within the paragraph text, and obviously had to be moved to a separate dedicated place on each page.

1984 (c.) – Dungeonland & Temple of Elemental Evil


 

For my last entry, I'm hindered by the fact that there's no published adventure with Gygax as the principal author in the year of 1984. (An earlier version of this article presented a snippet from 1984's Mordenkainen's Fantastic Adventure, which credits Gygax & Kuntz, but we know that was originally a Kuntz creation, and it differs in some key ways from other modules of the same era.)

But if we look at adventures by EGG in 1983 and 1985, the overall style is pretty similar, so we can safely interpolate how one of his adventures in 1984 should have looked like. Above I've taken one encounter area from 1983's Dungeonland, and one from 1985's Temple of Elemental Evil, and compared them side-by-side. The Dungeonland adventure (a pastiche on Alice in Wonderland) is credited to Gary Gyax, and has no other credits listed that I can find, so it seems fair to assign responsibility wholly to him (and the same for the follow-up, The Land Beyond the Magic Mirror). Temple of Elemental Evil, of course, is credited to Gary Gyagx and Frank Mentzer (with editing platoon of Carmlen, Heard, Johnson, Russell, and Winter); and we know the adventure was primarily loose notes from Gygax's campaign, turned over to Mentzer, who made significant changes when drafting the published text. 

Anyway, the format of these 1983-1985 adventures is very similar. Encounter areas tend to be several paragraphs each. Both adventures make use of boxed text, as introduced in the 1982 Tsojcanth dungeon areas. (Interestingly, Temple has a strong tendency to start with some non-boxed info to the DM, and then give a boxed description fairly deep in the encounter text; as shown above, this isn't universal.) Similarly, monster stat blocks are now uniformly removed from the paragraph text, and given as separate lines after the paragraph they're mentioned in, as was done in the Tsojcanth wilderness areas. Dungeonland has a 2-column format, while the Temple uses 3-columns per page. While both keep the tradition of noting treasure near the end of the text, Dungeonland had a formal annotation of a boldfaced "Treasure" label, whereas the later Temple abandoned that technique. (This is a trend I've seen a few times over the years: the more formal one tries to make the structure, the more brittle it is, and the less likely to stand the test of time.)

Also worthy of consideration: 1985's Isle of the Ape, which had Gary's name alone on the cover (and he'd mentioned it as a level in his Greyhawk campaign back in the 1979 DMG), with a credit inside for "Development: Bruce A. Heard". The interesting wrinkle here is that monster stats are entirely eliminated from the encounter text pages, and instead removed to a summary table on an extended flap of the module cover. 

And another common point among all the adventures of this time frame: The author/designer/publisher feels free (or maybe compelled?) to include some areas with very long descriptive text indeed. Each of the modules considered here have a number of areas with descriptions spanning multiple whole pages of text. Dungeonland sees this with The Long Hall (p. 4-5), Lawn and House (p. 14 and 19-21), Park (p. 21-23), and The Palace (p. 25-27). Isle of the Ape has an introductory boxed-speech by Tenser that is runs 2 full pages of text to be read verbatim (some 1,600 words!); with multi-page encounters such as Kawibusas' Ambush (p. 10-13), Plateau Area (p. 31-33), The Spheres of Thought (p. 38-39), and The Glowing Pearl Chamber (p. 43-45). The Gygax/Mentzer Temple of Elemental Evil is similar, with an introductory boxed text section that runs more than a full page (about 1,300 words), and area descriptions that commonly run many paragraphs or a whole page; e.g.: room 404 is ironically a 20'×30' room with 1,200 words of content text, room 417 has the same dimensions and 1,400 words, and so forth.

Open Questions

If you could set a dial for preferred extent of descriptive text in an adventure you'd pick up and run now, at what year would you set it: 1974, 1976, 1980, 1982, or 1984?

Where do you like your monster stats -- mid-text, after the paragraph, or removed to a separate table or section at the end of the module?

And did you like the boxed text innovation, or not?

2021-10-04

Random Wizard's Online Arena

Verona arena lit at night

You may be aware that a while back I developed a set of tools called Arena/Athena to investigate things in OD&D like implied combat efficiency, level demographics, balancing monster threat levels, generating bands of bandits & pirates by playing out their full career arcs, etc. You can see a lot of results from the tool in prior blog posts here. 

Problem is, those tools run on the command-line locally and assume you have a Java installation and some programmer knowledge, with no fancy GUI controls. Because I'm hardcore like that, obv.

Thankfully, our friend Random Wizard made a proof-of-concept conversion from my Arena Java code to an online Javascript version that runs in a web browser, so it's a bit easier to test it out. Give it a minute when you click the link, because the simulator immediately starts running on page-load and it's a bit CPU-intensive while it thinks. After that you can enter other command-line arguments for different behavior, as per the original program. These include the following:


-a apply aging effects
-b base type of armor (=0-3, default 3)
-e report every encounter
-f fights per year (default =12)
-m magic per level chance (default =15)
-n number of men fighting (default =100)
-p play-by-play reporting
-r reporting types
      s summary statistics    y year-end info
      d detailed data         k monster kills
      t  total monster kills   x xp award ratios
-s start level for fighters (default =0)
-t treasure awards by monster (default by dungeon)
-u create matrix of win percentages
-v man-vs-monster (default man-vs-man)
-w use fighter sweep attacks (by level vs. 1 HD)
-x use revised XP award table (from Sup-I)
-y number of years to simulate (default =50)
-z fighter party size (default =1)

Thanks to Random Wizard for making this possible! Any interesting results you've discovered with it?

Random Wizard's Online Arena