In a movie, there’s (eventually) one camera. And the camera shows the audience exactly what the audience needs to see. However, while shooting, while the actors are emoting, and the set is looking epic, and the music swells, and the fog machine is pumping out atmosphere, the whole time the camera is pointed at the actors and the set. The camera is showing the audience what they need to see.
In games, there’s also (eventually) one camera. And there’s a scene in front of the player/camera where beautiful models with slick animations are emoting the hell out of the dialogue while the music swells, and the environment the artists put together looks great, and the VFX are swirling in the air, and the music swells, and, hopefully, the player/camera is looking at all of this. The player/camera is showing the audience/player what they need to see.
Now imagine, the movie camera is filming the scene, and the cameraman decouples the camera, turns around, walks over to the Kraft Services table, and starts playing with a soda can; filming the entire time. He picks it up, looks around, tosses it in the air, throws it at an intern. And then he walks over to a door and tries to leave.
This is crazy in terms of a movie, but this can easily happen in a game. Instead of watching the characters exposit story, the player turns around and plays with the game physics of a soda can for five minutes.
I think the challenge that Gabe was talking about was that of providing the player with enough interest and engagement as well as level design, level art, and general design tricks that the most compelling thing for the player to do is watch that scene, and not wander off in the middle of it. Or if they do, allow the game to respond to that player action in an intelligent way so as not to interrupt the suspension of disbelief.
O’Ladybrain; I think your first paragraph sums it up nicely, and I think your casting metaphor works fairly well. You might have an A-List, professional actor come in and perform perfectly, and you may have a player who’s invested in the game; the mechanics, the story, everything, and would sit through a performance by digital people and not wander over to the soda can. But one could also cast a no-experience ‘extra’ off the street as the main actor and have the shoot go horrifically awry. Just like having a player who got the game as a gift, but never plays this type of game, decides to give it a try anyways, but, for example, only likes the shooty parts; as B McP indicated in his post.
B McP; I think you may be casting people in narrower genre preferences than perhaps they actually are. Sure, I’ve met a dude who only ones an XBox360 so he can play COD:MW3 (he’s never bought a newer COD), but I’ve also met a guy who loves the Madden Games, but also played the entirety of ZenoClash. These two data points do not a conclusion make, but I’ve met many more of the latter people than the former. At the same time, however, I do agree with your premise that games that try to appeal to as large a purchasing group as possible have a tendency to have ‘less great mechanics and more average ones’ rather than games that contain ‘several great mechanics and only a few average ones’. And the more successful games generally succeed with narrower foci. But it’s a hard balance to strike; wide enough appeal to make (the publishers) lots of money, while remaining narrow enough to focus on quality. To me, this seems like one of those pesky ‘ineffable’ qualities of a game that only are proven after release, and before release, it’s just educated guessing.