[Update, I've been working on my own puzzle game, Bonnie's Bookstore.  It doesn't address everything listed below, but I'm pretty happy with it :)]

In the last few weeks, I’ve been playing a lot of puzzle games, such as Bejeweled, BookWorm, and Puzzle Pirates.  After a burst of popularity 10 years ago in Tetris/Myst era, puzzle games had seen steadily declining popularity.  But in the last 2 years or so, easy web access to downloadable Flash and Java games has led to resurgence for the category. 

I enjoy these games, along with their close cousins – web based card games (Bridge, Spades, and lately, Poker).  Some thoughts:

The Good (i.e. why this niche has grown considerably in the last 2-3 years):

1)   No video card driver issues - All games are 2D and run the same on any reasonably modern PC.

2)   Easy browsing – Portals such as RealArcade, MSN Games, PopCap, and Pogo assemble a variety of games, filter out the bad ones, and provide an easy access point to sample new games.

3)   Easy startup – These puzzle games generally involve a single click to launch, and nothing that truly needs to be ‘installed’ on your PC.  You’re looking at the game’s main menu, ready to play, 5-50 seconds after you select a game.

4)   Easy to learn – The designers of these games have done a good job of coming up with very simple game mechanics that allow a novice to jump right in.  After you finish a level or two, a new variant may come into play and make the game a bit deeper.  Overall, these games are very easy for even grandma to learn and play (and in fact, the biggest market for many of these games is the decisively non-core gamer market of women, ages 35-60).

The Bad (i.e. what this niche needs to do to keep growing and ultimately surpass traditional retail PC games):

1)   No feeling of ‘progression’ – Almost all of the puzzle games last 2-20 minutes, with no larger meta-game to encourage you to play it ‘just one more time’.  After you’ve mastered the basic mechanics, further play nets only small improvements in your best score – you never ‘win’ or advance. 

What these games need is some kind of linkage – a way that you can progress by playing and beating ever-higher levels of an ever-broader array of puzzles.  Myst used puzzles as a way to advance a story.  The obscure old Macintosh game ‘The Fool’s Errand’ used individual puzzle-solving as a means to solve a larger macro-puzzle.  MSN games give a taste of this – you build a rank for yourself by beating different puzzle games.  But their implementation is very thin.

Think of it this way – Diablo II is one of the most addictive games of all time.  But it’s core game mechanic (clicking on monsters to kill them, then clicking to pick up treasure), is extremely simple and repetitive.  What makes the game compelling is that the basic game mechanic translates to progression for the meta-game (making your character stronger and progressing through the adventure).  A puzzle game that does this right could be as explosively popular as Diablo II (or Myst, for that matter).

2)   Poor integration of group and solo play – Most puzzle games (Puzzle Pirates excepted) are focused purely on solo play.  Most card games are focused purely on group play, with, at best, bad A.I. bots filling out seats.  Focusing too much on either manner of play is a mistake. 

People want to use the internet to reach out to others.  MMORPGs (i.e. Everquest, World of Warcraft), are the clearest (and fastest growing) example of this.  But blogs, in their own way do this, as do web forums, instant messaging, and many of the other most popular newer bits of the ‘net.  If I play Bejeweled by myself for much more than an hour, I feel socially retarded.  But playing cards on the ‘net, in the right environment, can feel like a nice socially connected activity.  (Most card games don’t do as well as they could, but that’s a different story) 

Conversely, there are times when players want to play by themselves.  Perhaps they’re just learning the game in question, can’t find a group to play with (late at night, perhaps), or don’t want to make the time commitment (20 minutes or so), to play with a group.  Unfortunately, most games that allow group play in fact require it.  In addition to all the other problems, this can turn off new users, who don’t want to be ridiculed for their poor play by the experts. 

So the solution is clear.  To the greatest extent possible, design games in ways that allow group-play.  And in virtually every instance where group play is supported, solo-play (with decent A.I. bots filling in, if needed) should be supported.

3)   Copycat games – Puzzle games seem to be just as infected with copycat disease as other gaming genres.  Virtually every game involves some variation on grouping shapes/colors/words to remove them from the board – the formula first popularized by Tetris 15 years ago.  This is fun for a bit, but if new users keep finding nothing but variations on this same basic theme, they will quickly abandon the genre.  There are bursts of originality (a Sherlock Holmes puzzle inside a Victorian mansion), but they are too few.  In addition to more creative puzzle games, the industry needs to use the apparatus and techniques built for these games (i.e. the portals and downloading systems, etc.) with other genres.  Simple strategy, action and sports games seem like logical directions to explore. 

4)   Licensing needed – OK, I’m pretty reluctant to even suggest this, but the downloadable market needs to explore licensed content/stories/brands.  Yes, the move towards licensing has swamped conventional PC and console retail gaming.   I would never want to see licensed games consuming 50%+ of the market as is the case in retail gaming.  But there are benefits to licensing a Harry Potter, Spiderman, or Law & Order.  The familiar names and settings of these properties make games more accessible to casual fans.  For a market that is rather explicitly targeting a more casual fan base, it would seem a no-brainer to use licensed content as the draw.  Yes, the revenue in downloadables is small (for now), and so negotiations with content owners would be tricky at first.  But computer games were once a niche market too, before E.A. started using brand name endorsers (John Madden, Earl Weaver) and others rolled out successful licensed games (Star Wars, James Bond).  Anyways, the properties that might have the most appeal within the downloadable/puzzle market are not necessarily big hits for retail gaming, and might be flexible – TV shows in particular (CSI, Extreme Makeover – Home Edition, Survivor, American Idol)