My strength as a DM is improvisation. The less work I do before a game, the better my sessions seem to run. Which works for me because I'm inherently lazy 
The problem with this method is that it doesn't gel very well with more recent systems. I learned to DM AD&D (1e & 2e... really a hybrid since I didn't differentiate between the two and used the rulebooks mostly as a DM screen or writing pads) where you HAD to improvise almost everything, at least until the Powers & So-Called Options books and plethora of Handbooks came out and turned everything into a rules-fest.
Anyway, point being is that 3e and 4e sort-of require you to know the rules. Now I KNOW that you can improvise in any system. The logical part of my brain tells me that. But I also truly believe that people are hard-wired to follow-the-leader and fall into lock-step. So regardless of knowing this, I still tend to follow the rules and use them as written (or with houserules). And in part, this is also because everyone at the table expects me to do so. They all came to play 4e, not AD&D 4e.
This is one of the reasons why I'm quite excited by one of the aspects of 5e that has been touted, ie. bringing back improvisation and supporting it within the fundamental mechanics of the system. So even though I know I could just run a 4e game using page 42, having it as an expectation within the system makes my life a lot easier as a DM. The players expect that type of game and I don't have to fight the urge to fall into step with the rules.
This all leads me to module design. One of the things that has always put me off using modules is the sheer amount of information you have to take in to run one properly. Worse are the little tidbits of information that the module turns on and that if you forget or muck-up, you can muck-up the entire module. Referencing modules in-game are big pet-peeve for me as both a player and DM.
In the end, I just end up not using them altogether. I don't want to have to read through thirty pages of snore-fest and memorise it all in order to just run a few sessions. And what's more, is that invariably I find that players don't follow the clues laid out before them anyway and since it's a module, I end up having to rail-road them towards the conclusions just to keep the game moving forward.
Most of the problems of modules, I feel, comes from providing answers instead of just questions. The fun and engagement of improvisation comes primarily from seeing just how players deal with problems. Trying to predict outcomes for situations that have so many variables it's impossible to do so, ultimately leads to a sense of forced conclusion and disbelief on the part of the player.
So why provide the answers at all? Hell, why even provide suggestions? Imagine a module that ONLY provided questions? Forget stat-blocks, just give a very brief description of an NPC's personality and let the chips fall where they may. Present problems, traps, situations, dangers, etc. but without answers the DM and players are then left to engage each other in providing the solutions.
Now of course, this may have already been done elsewhere and if it has, I don't really care. That's not the point of this thread. The point is to discuss the merits of doing things this way as opposed to the way I've seen Dungeon, and every other D&D module I've ever bothered to read through, do it.

The problem with this method is that it doesn't gel very well with more recent systems. I learned to DM AD&D (1e & 2e... really a hybrid since I didn't differentiate between the two and used the rulebooks mostly as a DM screen or writing pads) where you HAD to improvise almost everything, at least until the Powers & So-Called Options books and plethora of Handbooks came out and turned everything into a rules-fest.
Anyway, point being is that 3e and 4e sort-of require you to know the rules. Now I KNOW that you can improvise in any system. The logical part of my brain tells me that. But I also truly believe that people are hard-wired to follow-the-leader and fall into lock-step. So regardless of knowing this, I still tend to follow the rules and use them as written (or with houserules). And in part, this is also because everyone at the table expects me to do so. They all came to play 4e, not AD&D 4e.
This is one of the reasons why I'm quite excited by one of the aspects of 5e that has been touted, ie. bringing back improvisation and supporting it within the fundamental mechanics of the system. So even though I know I could just run a 4e game using page 42, having it as an expectation within the system makes my life a lot easier as a DM. The players expect that type of game and I don't have to fight the urge to fall into step with the rules.
This all leads me to module design. One of the things that has always put me off using modules is the sheer amount of information you have to take in to run one properly. Worse are the little tidbits of information that the module turns on and that if you forget or muck-up, you can muck-up the entire module. Referencing modules in-game are big pet-peeve for me as both a player and DM.
In the end, I just end up not using them altogether. I don't want to have to read through thirty pages of snore-fest and memorise it all in order to just run a few sessions. And what's more, is that invariably I find that players don't follow the clues laid out before them anyway and since it's a module, I end up having to rail-road them towards the conclusions just to keep the game moving forward.
Most of the problems of modules, I feel, comes from providing answers instead of just questions. The fun and engagement of improvisation comes primarily from seeing just how players deal with problems. Trying to predict outcomes for situations that have so many variables it's impossible to do so, ultimately leads to a sense of forced conclusion and disbelief on the part of the player.
So why provide the answers at all? Hell, why even provide suggestions? Imagine a module that ONLY provided questions? Forget stat-blocks, just give a very brief description of an NPC's personality and let the chips fall where they may. Present problems, traps, situations, dangers, etc. but without answers the DM and players are then left to engage each other in providing the solutions.
Now of course, this may have already been done elsewhere and if it has, I don't really care. That's not the point of this thread. The point is to discuss the merits of doing things this way as opposed to the way I've seen Dungeon, and every other D&D module I've ever bothered to read through, do it.