In the beginning, new DMs ran games for new players, and neither was particularly skilled or experienced. At low levels, the game was simpler, and as they grew more capable and skilled the game got harder. By high levels, it was a mighty challenge.
Then the campaign ended.
Then a new campaign started. But now, experienced DMs ran games for experienced players, and the low level game was found too easy and lacking challenge. So the experienced DM added new challenges, not just tougher or more subtle but also some trials created specifically in response to the established ethos of skillful play. Listening at doors was practically free intel, everyone did it and it was smart — so now, add earworms that might eat your brains if you listen at a door. Challenge is restored, and the game is deepened.
But what happens when that second-generation dungeon gets run for new players? A dungeon that’s merely brutally difficult can be overcome by determined players, it’s a trial by fire that forces them to learn and adapt (Caverns of Thracia proved this sort of crucible). But a problem can emerge when those elements that were included to add depth to accepted best practices are encountered by players who have yet to internalize those best practices — instead of learning that listening at doors can be dangerous, and therefore should be used with caution, they can end up taking away that listening at doors is dangerous, and therefore not worth it.
Some further concrete examples of this second-generation design:
Layouts that are intentionally hard to map.
“Time-waster” rooms that appear worth searching but aren’t.
Construct zombies immune to turning, along with most other monsters-disguised-as-other-monsters.
Enemies that use military oil, explode on death, or otherwise undermine the dungeon-phalanx and use of henchmen.
This is not to say that these are bad elements and shouldn’t be used. They can be great fun as a challenge to experienced players. They’re just not really beneficial for new players, inasmuch as they undermine the core procedures that should be developed.
“Isn’t this just an extreme case of trial by fire as discussed above?” Not in the same way. The key qualitative difference is that a hard-but-classic dungeon rewards the same basic skills as are generally useful to adventurers and develops proficiency in using them well, whereas second-generation design actively disincentivizes those basic skills in order to force experienced players to innovate.
This phenomenon is I think at the heart of the divide so clearly visible in OSR discussions between many AD&D grognards and many newer players. Those who have been playing D&D for 40-50 years want novel scenarios and more nuanced challenges, and that is right and proper, but it must be understood as the result of extensive experience with the system. Some of those advanced challenges may be appropriate for newer groups (even if they end up as “too hard” areas that serve a different purpose), but some aren’t, because they were created to undermine fundamentals that such groups haven’t developed yet.
There are some permutations of this that I’m not sure should exist at all, e.g. secret doors that are invisible to the spell detect secret doors. Given the high incidence of such in modules, it seems frankly better for the spell to simply not exist, but that’s a problem for another day.
My own table is getting to the point where more use of second generation design elements could be really interesting, and we’ve had some good fun lately with anti-phalanx tactics and disguised monsters. I frankly don’t think they are ready for time-waster traps, nor would I be as a player (I would like to think my DMing skills have outclassed my more mediocre player skills).
If we were playing at a table where some members had been playing much longer, I expect more of that would be passed down culturally. I got a little taste of that experience — my first foray into roleplaying games was rolling up a character for a high school teacher’s run through his original set of the GDQ modules — but my players have not, and our journey overall has been characterized by a lot of trial and error. Thus, I’ve found it a valuable tool to keep this distinction in mind, between classic and second generation challenges. It is a valuable tool towards seeing players get to that high level of skill to be able to to appreciate those more nuanced elements in the first place.
Great observation on a subtle problem. This is one of those places where designers, refs, and players can lose perspective, becoming unable to identify whether they are practicing first- or second-generation thinking. Luckily, simulationism can provide an anchor that roots all three groups into the same patch of ground, preventing anyone from straying too far from the basic "truths" of a system.
Seems like a very useful lens; one which I will keep in mind with my brand-new players. I had/have been making an effort to keep things simple but not quite in these terms.