Tomb Raider: Underworld Demo... Bump it up my Gamefly Q

Creating a great demo is hard. It needs to be accessible to all gamers yet it can't be so dumbed down that it bores the core market who have played lots of similar games and maybe even previous versions.

Tomb Raider: Underworld was underwhelming in some ways as a demo in terms of game play. But it was really impressive in terms of visual polish and the "feel" of exploration. The game is flat-out beautiful. The music, the attention to detail when it comes to animating Lara Croft's climbing techniques, and the beautiful and lush world make me want to just climb and explore for hours on end.

The combat, however, was a little shaky for an initial experience. The combat system is really designed for slower targets that use cover, not large and fast (and deadly) lions. If I could have changed that one aspect and the unsatisfying ending then I think the demo would have been outstanding. Some of the specific changes I'd have made:

• Combat vs. enemy humans. This would have been easier to balance and much more accessible. The game isn't really about free form running and gunning (well, at least the previous version wasn't) which is what the demo combat encounters were like.
• A more satisfying ending. It was clear that after I solved the puzzle I triggered a Quick Time Event (QTE). Unfortunately the way the demo ended and cut to the upsell screen left me feeling that I had failed the QTE and died before realizing that I wasn't even going to get a chance to check out one of the core features of the game. The absence of QTE events in the demo seems weird.
• More forgiving failure conditions during early platforming/exploration. The initial climb was long and reasonably complex for a early-middle game experience. Unfortunately this meant that as a first experience there was much falling and starting over. This got a little bit frustrating as the punitiveness of the fall tends to make the player less willing to experiment and try new things.

Other things I feel like commenting on...

Using a radial interface for the main game shell menu:

• I'm not sure this is a great idea. The icons aren't nearly as useful as text labels. Removing text labels has simply meant that more menu selections can be put on the screen at once -- which means more bouncing around because you need to toggle through all of them to find the one you want. Moreover, there isn't any room for context-sensitive help text, so it's hard to figure out what the different toggle options are.

The "learn-as-you-play" experience:

• I'd like to see the actual intro level to this game. The current "tutorial mode" was a standard series of text billboards that were sometimes contextually relevant (like the first time you stomp a spider) and sometimes way, way, way out of context (like the weapon aiming and firing instructions that were provided whilst I was climbing a huge pillar).
• The climbing puzzles were a bit too advanced for an initial experience and I'm assuming that they were extra frustrating for first time players. A number of off-camera "leap of faith" maneuvers were required and failure was punished by a significant (but not unreasonable) amount of backtracking.
• Combat was frustrating and did not feel cool. As I mentioned above, the ranged weapon system is much more suited for human-to-human combat where both sides are exploiting cover and other environmental devices. Being surrounded by multiple strong, fast enemies with little obvious way to get to cover and regroup made for several annoying player deaths.
• That said, the game does a really neat job of providing cool environmental visual and sound cues to help guide the player. I discuss this in a bit more detail in the next section.

The "where to go/what do I do next" systems:

• Both the Tomb Raider and Uncharted: Drake's Fortune series try to push realism and pull back user interface clutter in order to bring about an immersive experience in a beautiful world. Lighting, environmental design, character animation, music, and sound contribute to creating worlds that are fun to explore. 
• Both series also appreciate the difficulty facing the player when they are thrown into a 3-d game world from a 3rd person perspective camera. Not only is the moment-to-moment navigation often a challenge (how do I notice targets that are off-screen?) but it's also difficult to convey an overall map-level understanding of the level (where did I start? where have I been? where do I need to go?).
• The moment-to-moment navigation requires attention to subtle environmental details (like does Lara lean or not when you pull in a certain direction; are birds perching in a conspicuous place) and patience (letting the camera float to a "best" position if you're stuck and unsure of where to go next). You also get a reinforcing musical "ching" when you've hit the next way point, which assures you that you're making progress and not just backtracking. Of course, it would be nice to have a visual onscreen flash/text, too... But I understand why the game designers decided not to do this.
• In terms of map-level understanding, there are a few systems in place that reflect great intentions but that fall short in execution. First, there is the 3-d Sonar map. This is a very intimidating 3-d visualization of the game world. It's possible that the player could be gradually introduced to this tool over time, but it certainly seems overwhelming and not very useful. Second is the built in hint system. The basic idea is very clever: Have a spoiler-free method of obtaining hints on your PDA. You need to press A ("hear what to focus on" -- HINT) or Y (hear what to do -- TASK). On the surface this is cool and made me think of the great hint system in Bioshock (level 1 was an in-fiction reminder of your objective; level 2 was the practical "what do you need to do next" explanation). In practice, both hints were so vague as to be unhelpful when I truly was stuck.

I bumped it up on my Gamefly Q, but won't be purchasing it (even though I did purchase the previous version).