Does a game's perspective serve to set your expectations in terms of content and overall quality? Mention a game with a first person perspective, be it a shooter or a space sim or an RPG, and it seems that the content and interactivity expectations are astronomical when compared to a more removed perspective (like top down or distant 3rd person). If it's FPS, you'll practically hear people talk about expecting the bad guy's nose hairs to be animated. If it's a space sim, every rock and pebble needs to be visited from light years away (uh, and without loading screens thank you very much). Yet if you keep the same gameplay and change the perspective, the level of detail expectations dramatically drop. For instance, in Civ IV you can build dozens of cities but I don't know anyone who expects to be able to see every building, let alone travel inside them (yet I've heard this complaint about games like Mass Effect, GTA and Crack Down). Have you seen this effect? Do you expect more when you see through the eyes of the avatar as opposed to seeing them in the distance? If so, what do you think is at work in creating it? In a space empire game, for instance, why don't we demand to be able to visit every blade of grass as we now seem to in space sims?

I think it's because people make comparisons within the subset of games that most closely resemble the example game in question, using their own experiences and preconceptions as a guide.

I too have noticed that this is extremely prevalent amongst first person viewpoint games, since the beginning of the FPS genre. Magic Carpet was compared against Doom, even though the games have nothing in common save for the first person viewpoint. The Dark engine used in Thief got negative reviews when it was compared by action gamers who wanted something like Quake: while the Dark engine wasn't as fluid as Quake and had some collision issues, it had much better sound support and could host levels many, many times larger than Quake, making it better suited to Thief.

I think if you wanted to avoid this with a first person viewpoint game, you'd need to strongly differentiate your engine from the mainstream in the field. For example, if your game has a highly abstracted, cell shaded cartoon look, I don't think people will be comparing it against the latest Unreal engine.

I have definitely noticed this trend. I by no means have any sort of formal education to support this, but I have a guess.

When you put people in a first-person perspective, the game is trying to imitate something we are always working with as humans having eyes. We all know what the world looks like from that perspective, and we know what to expect. Zoom out the camera to a third-person follow or isometric camera, and people are viewing things in a way they don't normally. They do not have the ever waking moments of previous decades to compare with. It is easier to hold that 'Suspension of Disbelief' because it is abstracted a little more.

Just a guess.

Interesting question. I don't think I have an answer to it just couple of thoughts. I think I agree with Robsykes: it's about abstraction. In a first person perspective the game sort of implies that there is no abstraction - that this is just like the real world. So you expect realism. You expect a world that is not different from your own in terms of detail. In another perspective say the birds eye view of a RTS game the game abstracts a lot for you and you are ok with it because it doesn't imply that this is real or is an as real as possible simulation of reality (or parallel universe or future past or distant reality).

I think there are games where you can have both. Think of Command and Conquer. You're on the battlefield and control your army, kill the enemy. I don't think you imagine that you sit in a bunker somewhere near the battlefield or in your countries strategic command center and control your units via satellite uplink and a computer interface that looks an awful lot like video game. Yet still you might think that you have to stop the enemy from building that nuclear silo or otherwise innocent people will die. You will have the missions objective in your mind and take it for real within the boundaries of the abstractions of the game. But then the mission is over. And you're lucky and get a cutscene - an audience with Stalin or Kane or the President. Then you're in a first person perspective - they talk to you refer to you directly. I think you would expect that to be real! You want real actors and good acting and a nice set so you believe that you're actually there.

interesting question definitely.

Another reason is that the further away from the character(s) you are the more stuff you usually have to render and thus requires you to reduce detail because of a triangle or man(artist) power limit.

Its also harder to make out detail if you are in 3rd person because you are further away. It would require the player to take time out from the game to zoom into the detail.

On a reverse note some fps's would benefit from simpler graphics because the gameplay is so fast and your attention is on winning, not gawking at your surroundings. UT3 is an exellent example, the detail/bloom/crap gets in the way. UT04 with higher res textures / dynamic lighting would have been perfect, the detail in the levels did not need increasing.

I found most of the detail in black & white 2 was a waste (i.e. grass).

A good example of exellent use of detail is Dawn Of War, a technically challenged game (DX8) but still managed to pull of engaging/interesting visuals with correct placement of detail.

From a strictly artistic perspective, I liked the visuals in Command & Conquer 3:Tiberium Wars much better than those of Supreme Commander. But CNC3's zoom level seemed cramped, forced; many times I wanted to zoom out.

Supreme Commander has strategic mouse zoom, which probably presents some technical challenges in level-of-detail. I almost never play the game zoomed in to CNC3 altitude. Instead, I zoom out to get as much territory visible as I can (most of the map), but stop zooming out when it becomes difficult to differentiate/pick the unit icons because they are so close together.

Wish there were some quality detail at that level of zoom.
Example zoom levels

I remember imagining how wonderful it would be to have games that were scope agnostic, where everything relevant would be modeled down to the most believable possible degree, so I'm a bit biased in that I haven't naturally come to expect games to do things in stereotyped ways. I've had to grow accustomed to it.

I'm not twelve anymore, but I still do think that designers stick to expected genre conventions in terms of the level of detail in game simulations instead of trying to consider what would make a game fun. The aim, for example, may be to develop a fun FPS, instead of trying to develop a fun game and then deciding what features to add based on that. The logic seems to go genre->concept->fun, instead of concept->fun and then, at some point far down the road, a reviewer somewhere will arbitrarily decide what box to pigeonhole your creation into.

It seems increasingly popular for developers to pigeonhole themselves preemptively. It's probably some combination of marketing putting pressure on to deliver a game that will sell, a lack of designer creativity, and a fear of leaving the comfort zone provided by adhering to genre based standards. Much of it seems to be attempts at imitation of popular titles and trying to make the proverbial lightning strike twice. Beyond this, designers often have niche areas of interest: they have done FPS games forever. They don't know anything else. I find it confusing, no genre holds my interest that well without being an amalgam of various pieces of other genres.

That said, it would be nice to see games that did away with most genre stereotypes without themselves getting lost in trying not to conform. I wouldn't mind a game that was a good FPS that had a layer of strategy that would be considered quality in a turn based war game. I wouldn't mind an RTS that judged combat based on combat simulations less simplistic than distance and a few random variables, and I can't see any reasons that such a game wouldn't be fun - I'd wager to bet the limitations aren't in the viability of the concept, they're probably more tied to implementation difficulties.

@Nytehauq, I think you're underestimating the vast amount of work that goes into a highly specialized genre specific game. Modeling everything down to the most believable possible degree, especially before you decide what kind of game you are making, would be a huuuge expenditure without any measureable profit. (nothing to base your predicted success on... publishers aren't going to go for that)

Quote:

Original post by Nytehauq It seems increasingly popular for developers to pigeonhole themselves preemptively. It's probably some combination of marketing putting pressure on to deliver a game that will sell, a lack of designer creativity, and a fear of leaving the comfort zone provided by adhering to genre based standards. Much of it seems to be attempts at imitation of popular titles and trying to make the proverbial lightning strike twice. Beyond this, designers often have niche areas of interest: they have done FPS games forever. They don't know anything else. I find it confusing, no genre holds my interest that well without being an amalgam of various pieces of other genres.

Some of this is just aging of hobby in general.

But I think there's a lot to be said for playing it safe, at least in terms of trying to sell your game. Some of the most innovative games out there are fall into what Gamespot and PC Gamer have called "best game nobody ever played."

One of the things I find really frustrating is how genres have become shorthand for communicating concepts. Saying, "it's like Ghost Recon, but with monobikes" gets the idea across fast. But now let's say destructable environments come into play. Well, what game has that? Oh, and you want to add arena based competition and Madden style franchise management? Well now you're stuck no matter how good the idea is, because you're so outside the understood genre boundaries that people have a difficult time even imagining it (and you get the response, "well, we'd have to see it to know how we feel about it")

If you try to discuss the nuts and bolts (not just general concept) of such an idea enough, framing things in terms of genres becomes awfully tempting.

I especially agree with trapper and his (or her) notion that a lot of what the buyer expects is based on what the buyer is used to or what the genre tends to dish out. If a single CnC tank had millions of polygons, the users would expect millions of polygons on CnC 2 tanks as well as the tanks from CnC competitors.

Fortunately, at the zoom level of most RTS games, you can construct things effectively and beautifully with limited details because in real life we would not be able to see many details when looking down at things from such a distance. FPS games on the other hand, will continue to suffer from an increasing demand on details because the objects the in the users' view are so close. Fortunately at those distances the users wouldn't see as many things. A character model can be well developed because most of the other character models are culled out.

One of the wonderful things about GTA III and IV is that the user can barely ever really zoom in too far. There is usually much on the screen at a distance from the character. Even when zooming with the sniper rifle or camera, it seems like there is a bit of distance and the view mode is more concerned with steadying the reticule than showing a close-up. As a result, textures and polys need not be (and are not) as developed as games like Gears of War 2 and only certain characters and places in the game needed a very detailed texture set for cut scenes. I never found that "lola" character btw.

[Edited by - CucciKutta on February 19, 2009 11:04:21 AM]