In general I'm not going to bother to review games here. If you want a review of GRAW, go here. What I will do, though, is try to talk a bit (more) about what I learn specifically from specific games that I play.
Here I'm gonna talk about what I learned about Camera and HUD from Ghost Recon: Advanced Warfighter, which I played on the X360.
First, let me say that about half of what started me thinking about this was a discussion with my friend Pat, the Narrative Designer on my project, who pointed out a few brilliant things about the HUD design in GRAW that tied in in a nice way with my own currently developing thoughts on Camera in games in general.
Let me start by giving you my (current) definition of Camera.
Camera is the metaphor we use to partition the state-space of a game for display and parsability in the moment-to-moment experience.
I like this definition because it is broad. It's so broad in fact that you could fairly easily stretch it a bit to encompass non-computer-based games, but I'm not going to bother to stretch it to include something like chess or Monopoly. I'll constrain myself to talking about computer-based games which generally need to be displayed in some form or another.
GRAW is unusual in the sense that it uses what I have started to called a Second Person Camera metaphor. We obviously all know what a First Person Camera metaphor is - I see through the eyes of my playing piece or avatar. We also know what a Third Person Camera metaphor is - I see the world immediately surrounding my playing piece or avatar as well as my avatar itself.
A First Person Camera metaphor is hard to beat for immersion. You are seeing through the eyes of the avatar. A Third Person Camera metaphor is better for situational and physical awareness (which in some cases provides superior immersion). 3rd Person is a very useful metaphor if you have an avatar like Sam Fisher or Lara Croft or the Prince of Persia who performs a robust set of physical interactions with their environment.
A Second Person Camera metaphor is what I call a Camera that is situated sort of 'psychologically between' the 1st and 3rd Person Camera. It should not be mistaken for the literary or linguistic notion of 'Second Person' (this potential confusion is why it's a terrible name, but I prefer it to '3-1 Hybrid Camera metaphor' or something similar). A Second Person Camera metaphor works when a First Person metaphor is too limiting of the players situational awareness, but when a Third Person metaphor provides too much useless, boring or repetitive information. A First Person metaphor is good for a straight shooter. A Third Person metaphor is good for a platformer with an athletic character. A Second Person camera is good for a tactical simulation where much of your interaction with the world is shooter-like, but the physical specifics of your avatars stance and body position are important tactical elements, as in games like Full Spectrum Warrior, GRAW, the combat elements of Splinter Cell, and the much anticipated Gears of War (note here that I have recently heard Cliff Blezinski refer to the Second Person Camera metaphor used in Gears as the 'Parrot Cam', which works, but I prefer '2nd Person'... I'll save 'Parrot Cam' for when I get to make a game about pirates.)
GRAW uses a Second Person Camera metaphor. Most of the time my basic interactions are very shooter-like. When I'm moving around the world, I largely tune out the presence of my avatar on screen. However, when I get involved in the nitty-gritty details of maneuvering from cover point to cover point and firing around corners to suppress an enemy while ordering my squad to advance, I become acutely aware of the physical stance and position of my body. Correct position of my avatar in these intense situations is critical to success.
That's Camera covered... on to HUD.
Since the very first FPS, which probably be called Wolfenstein 3D (if we exclude the carnival shooters like Duck Hunt), the designers apparently realized that there was a hell of a lot of information missing in the 'just what you see through the eyes' 1st Person Camera metaphor and they added a HUD to tell you what level of the game you were on, what your health and armor levels were, what gun you had equipped, it even told you your SCORE. Some of this information was important, some was useless or redundant. Essentially it was telling you a bunch of stuff about the state-space that many other games had told you before to a greater or lesser degree. Over the next 5 years or so, designers refined the HUD to get rid of useless stuff like 'number of lives remaining', and 'score' and to reduce redundant stuff like telling you which weapon was equipped (after all, you should be able to know that from looking at the thing in your hand), and the big bar at the bottom of the screen eventually developed its own metaphor. It was now called a HUD.
Now there are some problems here. In my definition above, the HUD is technically a component of the Camera as it is part of the game that is displaying the state-space for the player. At the same time, a 'HUD' - if you want to think of it realistically, is a pretty high-tech concept. A HUD - in reality - is a piece of equipment connected to a computer that provides the wearer with information about the world imprinted somehow into his field of view. A HUD is great for an Apache Helicopter simulator - but how does a HUD hold up in a game set in the old west? Well - the answer is - it doesn't.
By default - the HUD was forced to develop its own metaphor. Usually it was an extension of the design metaphor used in the games more general menu and interface system. Generally, the HUD was designed to be something fairly non-obtrusive, but the pendulum has certainly swung back and forth on this. (Hal Barwood talked around the edges of this topic at GDC 2004 in his presentation about Cognitive Dissonance in games - which you can listen to for 5 bucks here).
Trespasser made an early ill-fated attempt to get rid of HUD entirely and collapse all the crucial information into the First Person Camera metaphor. It resulted in a lot of cleavage jokes I still occasionally hear today.
Currently, I would say that there is an occasional title that pushes a HUD-heavy experience (Star Wars: Republic Commando) but the trend is generally toward radical HUD reduction.
Compare Halo vs Halo 2, compare Call of Duty vs Call of Duty 2. Look at Condemned: Criminal Origins, or Peter Jackson's King Kong the Official Game of the Movie - a game who may be in sole possession of the claim that its HUD is less obtrusive than its title.
Now that I've talked about Camera and HUD, let's look at GRAW and what it does differently, what it does well, and what it does not so well and see what we can learn about Camera metaphors and how HUD figures in there...
Working on the original Splinter Cell, we encountered the weird problem of giving the player character night vision goggles. The weird thing about it was that when Sam - your avatar on screen - puts on his goggles, we dropped a visual filter layer over the screen that you as the player were looking at. Who is wearing the goggles? The player or the Avatar? There is a strange 'dissonant' element here, because now the visual information being given to the Avatar is being passed along through HUD to me as the player and is becoming a de facto part of the camera metaphor. Very strange stuff. It didn't seem to bother players though. I'm still not exactly sure why.
GRAW goes one further. They take their HUD - which is now a full-on 21st century special forces battlefield management computer that provides realtime communications, satellite feeds, and target recognition - along with a host of other features - and stick that on the head of the avatar. Of course, even though you can see your avatar on-screen wearing it, the visual information provided in his simulated HUD is bumped up to the level of the player and is made a kind of hybrid HUD/Camera metaphor.
What do I see when I am playing GRAW? I see my Avatar through the Second Person Camera with the overlay of the simulated HUD that my avatar is using imprinted on my own screen as a game HUD. Essentially what is happening, I think, is that the simultaneous simulation of the high-tech HUD being used by the soldier is being matched 1-for-1 with the game HUD. The effect of this is that it strengthens (tremendously, I would argue) the impact of experiencing the game as a hybrid player-avatar entity.
In many ways, I think this notion of experiencing a world as a 'hybrid entity' is a really compelling aspect of what games can do. I talked a little bit about the power of this in my own GDC 2005 presentation, Deconstructing Sam: Narrative in Splinter Cell (see page 17 of the paper). I think it is potentially more powerful than the capability of a film-viewer to empathize with an on-screen character. I think that the player, in navigating the psychological space of the relationship between the himself and his avatar, is offered a much richer perspective to the situation that the two joined entities must deal with together.
In my Splinter Cell talk, I was much more interested in actual character development, but in GRAW, this hybrid nature of the player-avatar entity is omnipresent and is embedded firmly in low-level play. The player is constantly experiencing this psychological space, and as a consequence is much more sensitive to challenges that 'tickle' the edges of it.
GRAW does something really smart with this space. They have a few missions and a few sections of missions where the enemy in the game is using some kind of interference device to jam the signal in the player-avatar HUD-Camera. Sadly, in most cases, this jamming device is used to make the game harder and add gameplay challenge at the mechanical level, and it is never used in such a way as to underscore a potential psychological conflict between the two invested entities. I wish the story, overall, had touched more on the idea that the character was being ordered to do things that the player would (probably) find distasteful, and that the jamming effect was used in those instances. That could have been very powerful. Unfortunately - they did not - but I am well aware of the kinds of constraints encountered in the writing of a 'Clancy-esque' script, and guys who question or disobey orders tend not to work well in the genre, so I understand and empathize.
Regardless, they use the mechanical technique surprisingly well. In any other game where the designers went 'surprise - now we're turning off your HUD!' I would put down the controller and that would be the end of it. When HUD is solely a game metaphor it is (generally speaking) out of bounds. But when it exists in this weird space 'between' it can be used as part of the game - which is interesting. Among other things, they use the HUD to inform you of when you have reached the boundaries of the world - which is in my opinion a startlingly elegant way to solve a simulation boundary problem. In other cases, when the jammers are being used as a weapon against you, they do not use it unfairly and punish you with some tedious and frustrating gameplay. Disabling the jammers becomes and objective, and it is pretty clear why you want to complete that objective, as your HUD is all screwed up and you are feeling kind of psychologically broken just as much as you are feeling mechanically impaired.
On the whole, I think the makers of GRAW did some very interesting and innovative things with their Camera and their HUD, and I hope that other designers and developers working on other games will learn from some of the things they did so well.