My preference in roleplaying games is not towards storytelling, but deductive reasoning. That is what is fun for me and what I understand as central to roleplaying: the mental reasoning out of categorical patterns and their relations to each other in my imagination as being related to me an impartial referee, the pattern giver.
I think this is a very important perspective. Over at the OD&D boards I remember you using the analogy of a classroom, and I agree that playing D&D is like being a student: part of the fun is mastering the knowledge that the teacher has and you don't. Let's imagine it's a history classroom, and that the goal is "making the past come alive in the students' minds" (for better analogy with the D&D experience).
By that analogy, a sit-back teaching style might involve creating problems in historical cause-and-effect, such that in learning to solve these problems the students will develop the skills they need to understand the historical period and imagine it for themselves. A lean-forward teaching style might choose some key moments in history to describe in an exciting lecture.
"Immediate and vivid" is best delivered by the lecture, "deep and abiding" by problemsolving. Too much of the former risks the students not being able to imagine history for themselves once the lecture is over, too much of the latter risks an abstract and lifeless conceptualization. Good teachers certainly use some of both (the sit-back prof moves around the classroom monitoring and guiding the process of the groups' problemsolving, the lean-forward lecturer involves students in a Socratic dialogue) but I've certainly been in (and enjoyed and learned from) classes that felt more like one or another.
To argue with my earlier statements about old-schoolers not always being old-school:
- My experience with guys who were there in the '70s like Tim Kask is that they are in fact very strongly "old school" sit-back; the convention game of his I played in at Gary Con last year had like 100 rooms spread over 6 levels, giving an enormously powerful impression of "here is a world in miniature, go forth and explore it as you see fit" instead of "here's just enough flavorful bite-sized chunks to deliver a thrilling story in a four-hour time slot." Arguably, by the time Chzbro and I started playing in '80 this style was already being lost, either because players didn't understand it (I certainly didn't when I was 10, or even 30) or wanted something different.
- It's true that folks who were old-schoolers do change over time. Paul Jaquays' advice in Shick's
Heroic Worlds (1991), like "Create stories. Think of the adventure as if it were a piece of fiction... The plot is the arrangement of story elements, describing the tasks the players must perform to overcome the major obstacle and achieve the goal", is very different from the message I get from his
Caverns of Thracia (1979), which has lessons like "the sense of wonder and achievement that the players get from finding hidden sub-areas entirely on their own, with absolutely no guidance from the DM or adventure, is often sweeter than the triumph of beating the obvious boss that the main dungeon is funneling them towards." At the same time, when I first read Caverns I thought "that's crazy that there are these secret areas that the players have no way of finding!" but actually playing it showed me that OD&D gives players lots of tools, like intelligent swords that detect secret doors, that aid their job of exploration.
- I think a major reason that the general evolution of D&D has been away from sit-back and discover it yourself (with a lot of rules for how you find secret doors and track the time and hazards spent in exploration) and towards lean forwards and tell a story (with rules for how to balance challenges so that you can string lots of 'em together in a slam-bang action sequence and minimize "downtime" spent resting or getting from here to there) is that the old-school rules were in fact designed for an old-school style. It seems to me that the push for a new ruleset isn't from people who say "OD&D is really good at exploration and discovery, but we'll make it even better in this new edition" - those folks are going to be happy playing it as is. Newer editions become more new-school specifically because people want a lean-forward storytelling style that older editions left room for, but weren't ideally designed to facilitate.
EDIT:
Umbran's post while I was typing mine is spot on. I'd add, though, that a virtue of letting the session play out the decision about how to go down the mine is that sometimes what the players come up with isn't what the GM would have whisked them towards. Sitting back gives you more opportunities to be surprised, both by the ultimate outcome and by the peculiar details of the players' plan which often generates things you can bring into the action that you wouldn't have invented by yourself. A good lean-forwards GM needs to leave space for this player-generated stuff, just like a good sit-back GM needs not to leave players room for endless dithering.
Also, I meant to link to my Mule Abides post about
nudges, which talks about techniques for doing what Piratecat was talking about here: "The nice thing is that you as DM can stack the odds. Want them to take the mine cart because it's the far more cinematic approach? Make it worth their while - hint that it would deliver them into a more advantageous part of the mine, that it would be fast enough that they couldn't be easily ambushed, or what have you. They may still pick rappelling, but you can stack the cards in favor of what will end up being the most fun for your players."