The Rise and Fall of Graphic Adventure Games
The Opposable Thumbs blog has a detailed retrospective on almost three decades of history in the graphic adventure genre. While this type of game has fallen from favor in recent times, many classic titles made indelible marks on the memories and preferences of an entire generation of gamers. If you played video games in the '80s and '90s, you'll probably see something you recognize. Quoting:
"In its sometimes-turbulent thirty-year history, the graphic-adventure genre has driven technology adoption, ridden at both the crest and trough of the graphics and audio waves, touched the lives of millions of people, and shaped the rise (and, in some cases, fall) of several big-name people and companies in the gaming industry. It's a genre that has often been held back by its own insularity, suffering from an unwillingness to adapt to changing market conditions or to further push the boundaries of interactivity. Adventure games certainly did these things, but the efforts to truly innovate seemed to peak in the mid-'90s, before rapidly falling off—with only a few exceptions. The improving fortunes of adventure game developers in recent years may at least in part be attributable to their efforts to innovate—Telltale with the episodic structure, Quantic Dream with a new control system (for better or worse), and Japanese developers such as Cing with Nintendo DS titles that introduce elements from visual novels.
Old Man Murray made a compelling argument explaining the decline of adventure games:
http://www.oldmanmurray.com/features/77.html
I would suspect that one factor in the death of adventure games as a genre(or at least their relegation to "smells funny" status) is that bad adventure games are absolutely fucking awful; but comparatively easy to make, while things like shooters, RPGs, and RTSes, tend to have a vast stretch of mediocrity to fall into.
Without tough-to-quantify-or-demonstrate-in-a-ten-minute-tech-demo-to-the-suits stuff like wit, good puzzle logic, and a dash of elegance, it is all to easy for an "adventure" title to fall into the morass of being a mixture of grindingly dull and unrealistic pixel hunts(You need a stick for reasons that make no sense. Go to the 'forest' area and move your cursor from right to left, line by line, until it changes to the 'action cursor' icon when you have found the one stick in the forest that is actually an inventory item, rather than painted background.) and dialog trees that read like the bastard spawn of a choose-your-own-adventure book and the worst tech support call ever endured by man. Extra credit for puzzles that make up for their childish simplicity by tacking on utterly arbitrary requirements that can only be fulfilled with fanatical inventory management and the prescience of the Kwisatz Haderach. The technical requirements of making such a game are minimal, so the barrier to entry is low; but the result is utterly unplayable dreck.
By contrast, with the exception of the "baby's first 3d engine" horrors that no sane human pulls out of the bargain bin(Extreme Paintbrawl anyone?), the world is full of utterly generic; but playable enough, Doom Clone N+1s, illegitimate children of either C&C or Warcraft, and Diablo clones of assorted stripes. Most are not good; but the more action-oriented genres seem to have a much wider band of playable adequacy. This both makes them lower risk to produce, and makes the average endurability of those genres higher. Ergo, more are churned out.
It's like humor vs. generic summer splatterfests. Humor well done is excellent. Humor ill done isn't simply dull, it is downright painful(I find this odd; but it seems to be the case). Your basic run-and-gun action fest or hyped horror vehicle, on the other hand, has to work much harder to be downright painful, even if its odds of being excellent are basically nil. For whatever curious reason, there is just a broad band of "OK" in some genres; but much sharper division between "superb" and "painfully worthless" in others...