As passionate gamers, we at GameSpy Tech (and I as a product manager) spend time each week thinking about which dynamics and technologies enable great gaming experiences, which don’t, and which just plain suck.
Recently, a friend forwarded me a new blog post from the smart Danc of Spry Fox titled, Steambirds: Survival: Goodbye Handcrafted Levels, in which he argues for killing the idea of levels and handcrafted-level design in favor of additional game modes which yield deeper game play.
(Don’t worry, I’ll wait for you go to read it and come back.)
But, to disagree with part of Danc’s post, I think it will be difficult (and unnecessary for a reason I’ll get to) to completely drop handcrafted-content in games.
Danc provides the following depiction of “deep play”, which he says results from pursuing new game modes instead of static levels.

I agree with Danc that this kind of expanded gameplay usually results from adding more play dynamics to a game.
For example, if a character in your game needs to reach a platform and your game affords only stackable boxes, all solutions will involve stacking the boxes (what Danc calls ‘shallow play’).
However, if you add ropes, pulleys, fasteners, and a bit of physics to your game, you now have a large number of potential solutions to reach a platform.
But here’s the tricky part about expanding a game this way: as you add more game dynamics that interact with and potentially build on one another in unpredictable ways, you are more at risk of accidentally introducing a “min-max” into your game, particularly if your game dynamics are already complex.
A “min-max” is a hidden solution or strategy that is either trivial or applicable to too wide an array of challenges in your game. It’s an overpowered gun in a shooter or the “Stasis” deck in old school Magic: The Gathering. It’s an unforeseen overpowered or near optimal strategy that erodes fun because it short-circuits the interesting tradeoffs or challenges inherent in your gameplay.
There are steps one can take against min-maxes, such as more QA and patching, but these can be development-cost-intensive and can’t guarantee all exploits will be resolved.
This doesn’t mean you shouldn’t add interesting dynamics to games, but perhaps they shouldn’t be considered an exclusive replacement for handcrafted content-generation.
Instead, one of the goals of game dynamics, in my opinion, is what Jenova Chen calls “Flow” in his excellent Flow in Games paper from 2006:

“Flow” happens when a game gives you new abilities or dynamics as you progress AND takes into account both your increasing skill and the interactions of the new dynamics, YET still manages to keep the challenge level in that sweet spot between boredom from overmastering or overpowering the game and frustration from overaggressive difficulty increase.
So, instead of eliminating handcrafted game-content and replacing it with more and more gameplay dynamics, I think a contrarian strategy to achieve deep player engagement is to build a fun but minmax-free set of gameplay dynamics and then open up the content generation to the player community.
Players relish opportunities for creative expression, to tell their own stories, and to add on to worlds they love.
And players who master a game are ideally-suited to produce the levels or scenarios that will provide both challenge and enjoyment for players like themselves.
One of my fondest childhood game memories is Lode Runner on C64, since it was the only game I knew of as a kid that came with a level editor. Once my friends and I had mastered Lode Runner, we invested endless fun hours building, sharing (on floppy disk), and challenging one another: first with more difficult levels than those provided with the game, then with themed levels that tested the limits of what could be built, and finally with “trick” levels, which looked possible, but weren’t!
Opening up creative content building and sharing to a player-community makes the enjoyment and content produced for a game potentially infinite.
And, since you’re reading the GameSpy Technology blog, I should probably point out that user-generated content is a feature you can readily add to most any game project using our cloud data storage offering, Sake, and our soon-to-be-released user-generated-content features, which add content ratings, flagging, moderation, virus-scanning, and a browser-based admin panel for moderators to monitor and manage all your game’s UGC.