Editorial: Looking at an old perspective
I recently came across this footage of a Fanny Vergne pre-vis concept (a game supposedly rejected at Ubisoft) that got me thinking. Why do games so frequently fail to push innovations in camera placement and movement? Take a look at this video and I’ll meet up with you after below.
I’ve rarely seen camera work like this before and I’m trying to figure out why. It looks great, it looks exciting and it looks immersive. I also think about the 3rd person games I’ve played and I realize most of the camera work feels the same. Most of them provide a full range of movement camera that just floats on behind. Isn’t there anything more we can do than just follow? Is that what we really want?
Ultimately, yes, the camera needs to show what we need to see. Prince of Persia would be near impossible if we didn’t have complete control of the camera. On the other hand, games often have moments where the subject of your actions is specific and the player doesn’t really need full control of the camera. An obvious example would be God of War, but first I’ll give mention to the camera lock on function in Spiderman 2. Whenever I got into a car chase sequence, I would immediately lock the camera onto the car before chasing it. This gave a whole different feeling to the gameplay because the camera would round corners and would take rather cinematic angles naturally that framed your character and the car very well. I enjoyed this camera behavior so much, I would typically chase the car for longer than I needed to. It especially felt great when you could aim Spiderman correctly and land on the car from the air. In this scenario, I didn’t need control of the camera, the camera was automatically taking care of pointing itself at the subject of my focus and was doing it with some style as well. All I needed to see was the car I was chasing. The controls were easy enough that mobility didn’t rely on looking at anything else. What’s amazing enough is that this was just a lock on camera, not predetermined movements or scripted events, just lock on targeting. In this instance, the stars aligned in such a way that it worked great in this open world game. Here’s some example footage of what I’m talking about (there’s no audio, so don’t freak out):
Assassin’s Creed is another example of a game where the player could angle the camera in odd directions, but still be able to navigate the environment without much trouble. The ease of control in this case comes from the fact that the character’s AI does a great deal of the movements that require pin point accuracy automatically. The player is generally just freely guiding the character where they want to go and the character is finding the footwork along the way.
Then there’s something like GTA 4. The game provides the player with an option to use a cinema style camera during car chases, switching between helicopter view, street view, side car view and a few others. It’s fun to play with this mode activated while generally messing around, but when it comes to story missions, when accuracy in driving really counts, it’s too difficult to use practically. The car physics were frequently unforgiving at high speeds so that, unlike Assassin’s Creed where the movement was easy even if the camera was not showing the player’s direct path, GTA 4 became too difficult to control. Clearly, in order to make a stylized view like this work, it needs to be paired with ease of control. It’s really a shame actually. Playing with a unique perspective like this really adds a new layer to the gameplay.
The key to a good camera is basically just this; the game should show the player what they need to react to in a clear and immediate way without disorienting the player. It should also do this while providing the player with controls fluid enough to easily navigate the game world at off angles. The camera work for this concept game above is far more interesting and immersive than just the general free movement camera in the majority of games today. Could something like this really work though? I would say yes. In this pre-vis it seems like the game is on rails, or uses auto lock on/targeting to direct the camera. It’s not a real game, but as presented the camera feels like it would be functionally either pre-determined, restricted, or automated rather than free movement. This is all assumption, but for the sake of the argument, lets assume that’s how it would work. The most important thing here is that, when it comes time to pick up the controller, the player can in fact see what they need to react to clearly and immediately. This title found a way to do it very uniquely. In this example, potentially, the player can still perform the movements they need to and at the same time be thrilled through cinematic camera work.
As far as I’m concerned, it’s perfectly acceptable in a game to lock down the camera, to take away the player’s control, so long as the camera is dedicated to showing the player what they need to see clearly and effectively as in the Spiderman 2 example above. Games do this all the time without us realizing it. Take a look at Real Time Strategy games. They have the most restrictive camera possible, but they are still 3rd person games (1st person if you consider yourself god). You have the ability to see what you need to and react accordingly, to send out your units and place your buildings. Any more freedom of camera movement is unnecessary. Resident Evil and Heavy Rain are some 3rd person games that completely lock down the camera at times. Both provide tank controls to make sure the player is able to control the character without being disoriented by camera shifts. Resident Evil occasionally falters at showing the player what they need to see because sometimes there’s a zombie off camera that the player is trying to shoot at. In this case the camera becomes a problem, the player placed at a disadvantage. Some might say this adds to the suspense, and it does in a way. It’s still a bit of a problem though.
Even a game like Uncharted will have camera restrictive moments to heighten the gameplay. In both Uncharted 1 and 2, there are chase sequences where the character is running forward and looking behind. There’s no need to look in front of the character in these scenes, all the action is happening behind the player. I love these moments because they were focused on a particular threat and it was exciting to be staring in the face of the danger. The opening sequence of Uncharted 2 is a perfect example of this (skip to about 2 minutes into the video for the dangling train sequence):
God of War does exactly what I’ve suggested a restrictive camera should do, show the player what they need to see as well as framing the interesting things in the background, but I’ve always found the camera in God of War to be a little too hands off, a little too far away. It always made the character Kratos feel smaller to me because of it. People may disagree, but I feel like God of War got the idea correct, but didn’t add much immersion to the camera’s presence.
Games like Gears of War, or Split/Second have some examples of what most games typically do to provide some subtle immersive camera movements. In Gears, when you sprint, the camera takes a new path of movement, swaying with your every hard impact foot fall. This jostles the player a bit more, presenting them with a greater feeling of force in the sprint. In Split/Second when the car reaches maximum speed, the camera vibrates under the force of the air resistance. Many games shake the camera to coincide with explosions and call it a day in their camera’s immersion department. I’m always proud to see it when a game takes a greater chance at being more engrossing through its camera movements, something like The Bourne Conspiracy. It did a pretty good job, though somewhat unpolished, of providing some unique style and flavor to the camera placement, delivering a visually interesting and immersive view of the action (skip to 1 minute 11 seconds in the video to see the fight scene):
The player wants to feel like the game is real, that the danger is real. I think we underestimate how important the camera is in connecting us to that. I also feel like we’ve become complacent in what we expect the camera to do. It can do more that just show us the action. It has the power to really put us in the scene, it just frequently doesn’t. I personally believe we’re ready to see what else is possible in camera techniques and I hope that the games that attempt to expand the conventional placements find success. Then again, the game above was cancelled, so maybe it didn’t work. Let’s just hope this next one finds a release Ubisoft:
12 Comments
Leave a Reply
Good article on camera perspectives!
Great article! I think that you are definitely right. If done correctly, the camera can almost become a character in the game, providing information about the PC, the atmosphere of the world, and the status of the action. Have you played Kane and Lynch 2? They have employed a very interesting use of camera techniques in the game, and it makes the gameplay feel much more multi-layered and interesting. I’ve been playing the Mafia 2 demo and Kane and Lynch 2 demo back to back and it has been very illuminating. The K&L 2 demo has a much stronger sense of place and a more powerful atmosphere, even though both games have relatively similar play mechanics (3rd person shooter with cover system). Obviously this has a lot to do with level design, art assets, etc. but I think that the single greatest difference is K&L 2′s superior camera. K&L 2 specifically uses a shaky camera (which has obviously already been done) but with a manipulation of the focus and distortion of local lighting through lens glare. The experience of the camera is definitely greater than the sum of its parts, but I think that the individual decisions to make the metaphoric camera in the game respond to the world like a real camera explains much of the innovation that they achieved.
The problem with having an intelligent camera like that is that the player can be so unpredictable. As was especially evident in the last clip, such dramatic, sweeping camera angles are only possible when the game can be sure what the player is going to do and when. A lot of those camera angles would have been completely useless if it had been an actual player controlling the character. In some cases the camera began moving away from the character before she even jumped, while in others the character jumped blindly and the camera only caught up by the time she was in mid-air.
Unless the game restricts the character to an extremely closed environment (as in Uncharted) or if the details of the player’s environment aren’t very important (as in Spiderman), automatic camera control is difficult if not impossible to pull off. That’s why so many games these days are including full camera control as an option – so that reviewers stop slamming them for poor camera work.
I really disagree, this actually annoys me quite a bit in games. The initial video and the beyond good + evil clip would be frustrating for me. I like to be in control of the camera all the time. For me the perfect camera is one that lets me zoom in and out, and rotate 360 degrees without it being reset or tugged away from me. What is it that makes games unique and appealing? The fact that the player is in control. Games are not films! Trying to make these dynamic cameras is not playing to the mediums strengths. You mention Assassins Creed as a good example, but that only works because the player has less control over their characters actions – its all automatic. Less control over the camera seems to necessitate less control over gameplay – i would much rather have precise control over both character and camera. If somethings chasing me through a city street, i’d like the choice to rotate the camera for a cinematic head-on view, or if i prefer, change it to a side-on view to help me jump obstacles. Thats where games should excel, in enabling player choice. The more games try to emulate cinema, the more i feel they’re missing the point…
Games aren’t and shouldn’t be movies. 100% undisputable. The fact is though film has gone through the trial and error exploring visual advances, both technical and stylistically, that games have no reason not to borrow from. The way we advance is by using the technology and theories of past generations and improving on them, expanding their context. I want to know how we can advance on cameras in gaming.
@Hmm -
I feel you completely here. I want to be able to see what I need to see immediately and the way I want to see it. The problem is that the player doesn’t have the physical capability to constantly control everything at all times. You can’t aim, swing the camera around, jump between platforms and use items or weapons all at the same time. The game needs to do some things for you, or at least aid you. I remember playing Prototype and I loved swinging the camera around for a chase mode, but this was only fun because the parkour was automatic. I would run away from things, bounding over cars with ease, concentrating all on the chase. I tried this again in Infamous and it wasn’t as fun because the bounding and mobility wasn’t automatic (though Infamous is a far superior game in my eyes).
Either way, this “chase” mode is a player determined restricted camera. You want the player to have the decision of what to restrict it to and I agree with this. There’s a time and a place for restricting the camera, and you’re right, it’s not all the time. That’s why I bring up the example of Spiderman 2. Gameplay wise it’s much easier to follow the car when the camera is locked to it. I had such a hard time controlling the character and the camera at the same time. It was much more fun and visually stimulating choosing to lock onto it. There are times in a game where the rest of the environment is irrelevant, so whether the designers or the player chooses to restrict the camera, it ends in the same way. I want that choice though. I’d like to be able to choose between restrictive cameras so that I can decide if I want a lock on, a chase mode, free mode or dolly mode. To contradict myself immediately though, I also want the game to choose this for me when my objective is very specific. I’m chasing a car in Spiderman 2, I don’t need an additional lock on button. All I need is a way to turn the lock on off when I choose to no longer chase the car.
Let’s look at your Assassin’s Creed example. You sometimes want the camera to the side of the character for platforming. Do you want the camera to remain in profile, dollying relative to the character, or do you want the camera locked relative to the environment? If the camera is rotating to remain profile to the character, then we’re talking assisted restrictive camera. If you want the camera to never auto correct, then you constantly have to be removing a finger from the mobility buttons to fix the camera when you turn corners. This can in some cases make the character control more difficult than you want it to be. The point is you want to have the freedom to pick where the camera should be for the situation.
@Max –
As for the Beyond Good and Evil clip, I agree completely. This looks impractical and impossible to use depending on the player’s interaction with it. I’d love to see it explored though just to see what parts of it can work, what the player can retain control of. I remember playing Prince of Persia and I wouldn’t always reposition the camera to look at where I was jumping. I would trust that the level design would be predictable and that the character would grab onto what I needed him to. In those cases, the placement wasn’t as important. Doesn’t work the same in Assassin’s Creed though which doesn’t have a linear setting. This concept demo seems to be more of the latter. Still, if there’s a “grab doorframe and quick turn” button, I’m ok with the camera jumping around as it does at 6 seconds in. When she starts climbing the walls at 18 seconds in, I like that the camera is leading her, assuming where she’s likely to go. So long as the player can still nudge the camera if it isn’t showing you where you want to be, then I’m fine with it. When you climb up a support column like that, 9 times out of 10, the environment above you is most relevant, you want to see where you’re climbing to. In the 1 out of 10th time, you want to be able to look down quickly. Still seems conceivably plausible to provide the player with that capability.
As you said, the player can be very unpredictable, but you also just described my point “unless the game restricts the character to an extremely closed environment…or if the details of the player’s environment aren’t very important”. There’s a time a place to restrict the camera. It’s conceivable that a game can create a priority system for the camera to soft auto lock onto. In Batman Arkham Asylum, the camera would frequently auto correct itself on the largest and closest cluster of enemies. This is perfect because that’s exactly what I want to see in the fight sequence. I want to see who I can attack and where they are so I can pick my strategy accordingly. Don’t wait for me to grab the camera control stick, show me the enemies immediately. If I choose to run away after I see them, then I’ll shift the angle. I need to know what I’m dealing with first and foremost.
Games include the free camera all the time because that’s what gamers are used to. Unfortunately innovation doesn’t always sell, no matter what the game is.
@aaron –
I haven’t played the Kane and Lynch 2 demo yet, but with your reactions to it, it’s the first thing I’ll do when I get to my console.
So, the final question for this comment is: What do you think the ideal time and place for restricting the camera is, and how restricted should it be in those cases?
Guys! This is an interesting discussion, but why the focus on third person action adventures?
“Boris”, do I have a game for YOU to play. It is a game which decided to make camera movement, restriction, and (kinda) control into key features of the challenge of the game. It’s called Space Giraffe. No, close that youtube tab, all the videos of it are regrettably shit. Just take it from me. Space Giraffe may look like a standard Shmup with a trippy skin, but what you’re actually looking at is the first game where the appearance actually *matters*. The challenge has been best described as “obfuscation based” as in, levels will frequently seem like visual nightmares, twisting, turning, flashing, colour-changing maelstroms. But with some patience, your eyeballs will eventually start dancing to the game’s beat. There are a lot of people who give up on the game too easily, or play the (admittedly misleading) demo and dismiss it. But seriously, dude, it’s 400 microsoft points, and by the end of the game your only problem will be a worry that Space Giraffe explores camera-centred gameplay SO comprehensively that there is really nothing left for other developers to do.
Wheesh, that was a long paragraph. I’ll also add that the rail shooters Panzer Dragoon Zwei (the temple level) and Sin and Punishment (the… aircraft level) have camerawork that is both challenging and interesting. And Flower.
Other than that, I’ll echo Max and say that coding a bog-standard “stick behind the player, don’t go into any walls, be controlled by the right analogue stick” is hard enough. Add “be interesting” to that, and you’re looking at something *tremendously* difficult. I suspect that game demo at the top was born of the idealism that you exhibit in this post, Boris, and I guarantee that if that game had been released then it would look NOTHING like that. The best camerawork in Uncharted 2 is available only in the heavily scripted sequences like the truck chasing you down the alleyway. Three of the things I just recommended are Shmups/rail shooters, where the camera is decided for you.
That leaves Flower and Spider Man. Both those games are focused on fast and cinematic movement. This is to say, if you want an interesting camera, you’re pretty much going to need to be thinking about it from the very beginning. I’m not sure whether that is conducive to good game design! Personally, I want to see strategic, frictitious-feeling games. Games like God Hand, posessor of the World’s Ugliest Camera. That game’s design is so intense and detailed that it NEEDS the camera in the place at all times, otherwise you wouldn’t stand a chance. It has fun with it when it can, but mostly the camera has to be focused DISPASSIONATELY on The Task At Hand. And when you think about it, if the gameplay is really good enough to stand on its own, why would you have anything else?
(Alternative perspective: Cinematographers know that if you want to draw attention to, say, someone handing an object to someone else, you ought to move the camera along with the thing, at least slightly: that is how the human eye and the human mind work together. In other words, a significant-feeling game interaction should probably give the camera a punch, as in Pac-Man Championship Edition or Burnout 3. Yeah, you know what, little bites like that are probably the place to start)
God Hand video: note the camera at 0:15 http://www.youtube.com/watch?v=MyfbtSyX3mc
@Hamish -
I unfortunately and shamefully do not have an Xbox, but I have instructed another GameCrasher, Brian, to purchase and play the game and report back. I alway love to look at games that have a unique take on perspective, so hopefully I can get an idea of this one by hearsay for the time being.
The camera work in rail shooters like Dead Space Extraction (I haven’t played Panzer Dragoon Zwei or Sin and Punishment) is awesome, but feel free to take a look at my review of the game because I feel like the interactive elements of Dead Space Extraction specifically detracted from the gameplay experience for me somewhat, like I almost would have rather just watched it as a movie. And odd paradox to me considering my stance.
As for Flower, the camera is more-so first person perspective, so there’s not terribly much that can be done other than change the focal depth when the player uses the boost move, which it does. I really like Flower, but you are basically the camera, save for the moments when the camera backs out of the world to show what you’ve done unfold around you. It’s always satisfying to have that moment of “sit back and watch everything come together”.
And yes, making interesting camera work is difficult, but so is making realism, a good story, realistic animation, realistic characterization and interaction. Basically everything is difficult to do really well, so I don’t really buy that argument that “it’s tough”. Professional game developers should be paving the way with innovations like this all the time. Innovators do it because it’s tough.
It’s difficult to debate the necessity of an “interesting camera” versus a “practical camera” without context. Honestly, when I think about doing things with the camera, it’s in context to something specific, like in Spiderman chasing a car or the Batman Arkham Asylum example I mentioned above: “In Batman Arkham Asylum, the camera would frequently auto correct itself on the largest and closest cluster of enemies. This is perfect because that’s exactly what I want to see in the fight sequence…”
If you have nothing specific to draw the camera’s attention to, then making “interesting camera” moves is irrelevant. At that point the cinematographer just wants to present the best view of the area. There’s still a great difference between good and bad cinematography though. I could give a camera to a great photographer and he would give me back a photo of a trailing camera perspective far better than I could shoot it.
What I’d like to see is more practical use of objective based, priority based camera functions and control. These can be something the designers are thinking about from the beginning based on the objectives throughout the game, and it doesn’t always have to be chase cameras like the Uncharted examples. I illustrated a concept of this with the Beyond Good and Evil 2 tech demo: “at 18 seconds in, I like that the camera is leading her, assuming where she’s likely to go…When you climb up a support column like that, 9 times out of 10, the environment above you is most relevant.” Here, momentum itself can be the context trigger. So long as the player has the capability to re-direct the camera freely without the game bogarting the camera again every 5 seconds, I don’t see why this couldn’t be possible. It would most certainly take time, but so do all innovations.
I can’t tell you how many times I cursed at the game Prototype, wishing the camera would auto correct itself when I was at the top of a parabolic arc, coming down on a building. I was pressing so many other buttons all at the same time, I couldn’t get to the camera stick with any confident ease. This is another example of what would be a very useful momentum trigger for automatic camera correction.
As with your God Hand video (it is quite ugly), while the camera transitions are poor, I like that the camera compensates to try to show you the whole picture. I’d say it could be fun to occasionally track the camera along with the flying villain into the wall behind them.
@aaron -
I tried the Kane and Lynch 2 demo, and I have to say it’s a start. It has a style that’s definitely unique. It’s presentation of unique camera work is mainly a shaky handheld cam feel, but especially while sprinting the player has a dynamic camera that jostles around (a little exaggerated in my book. I think it could be toned down a touch) but doesn’t feel like it inhibits the player’s movement at all. The game also feels like a Michael Mann movie like “Collateral” with a digital video look. Conceptually I like the visual presentation, but I don’t know how I’d feel after 10 hours of it. The game is a bit too dark for video seeing as video isn’t supposed to look good in low light. Also, it feels like the jostling camera would get sickening after a certain amount of time. Maybe that’s what it’s going for though. That or I should just stop sitting 3 feet from the screen… I’ll definitely give this one a try when it comes out regardless. Thanks for the tip.
Judging by the Uncharted video, it seems like the ideal situation is to use advanced camera techniques when the player is not necessarily in direct control. It could be for a cutscene, or less obviously, for very brief moments when the character is performing an animation or otherwise out of the player’s direct control, even momentarily (thus preventing making a wrong move on the controller).
See the part where the boulder falls past the train for an example. If this had happened while the player was trying to navigate carefully up the train (instead of in a brief moment when the character was jumping) it could have been distracting and could have caused the player to make a move that made the character fall. Here, it didn’t, and the camera move heightened the impact of the scene without compromising control.
As for GTA’s cinematic camera, I don’t think it was ever meant to be used during gameplay. But what if you could view a replay of every mission you undertook in GTA (maybe starting from a checkpoint where the action picks up) and watch your moves through this kind of A.I. director presentation? Seems like it would be an interesting use of the camera, because you maximize cinematography without needing to worry about gameplay control. I’d watch the replays.
Or take it a step further with placeable cameras and limited edit controls, as Driver did with its user-editable replays.
I loved the replay format of Driver. I really wished Split/Second allowed for user placed cameras in a replay mode.
“Basically everything is difficult to do really well, so I don’t really buy that argument that “it’s tough””
Well, I was just trying to explain why few people do it. The games I mentioned above are some of the best I’ve ever played, so if it didn’t come across: I completely agree with you when you say it would be interesting to see camera movement explored more! But, you know, don’t let that Ubisoft prototype get your hopes up is what I’m trying to say. The fact is that Zwei, S&P, and Space Giraffe all sold *terribly*. It’s no coincidence that you’ve not played them.
I think you’re wrong about flower. It’s not a first person perspective – when you go straight upward, for example, the camera is directed perpendicular to the direction of travel. I noticed the fun they have with the camera only when I saw it played by someone else – I recommend you try that. It swoops and swishes and ties itself in knots. It has the “objective” at its heart, i.e. showing the player what they need to know. But believe me, curves and skews like what you get there don’t happen on their own. They had a sense of viewport aesthetics, a sense of reflecting your actions by jogging your perspective.
And sorry, I only just watched the BG&E2 trailer. Dude: that is not going to happen. Game systems do not work that way. Like I say, I think that 3D games could do with more discretion applied to the camera. But there are certain directions that videogames *cannot* go in. That trailer is pre-rendered, just like the first one you posted. It may look like a game; like a really kickass game even. But it’s a spoof-game: a conception so idealized that it has nothing to do with games. If you want to know why, start thinking about it in terms of real game experiences. How would the player know that they could climb that pillar? That they should take that left turn through the building instead of continuing to run? That jumping over those guards and kicking them in their faces is a good idea? I don’t know about you, but when I was expected to jump onto the helicopter in Mirror’s Edge I stood there like an idiot for quite a while, and here is that exact sequence again.
And how is that a reflection on the camera? Well, think about it: that sequence looked awesome because we were looking at an idealized player who knew exactly what to do. Therefore, we could have an idealized camera that can have all the Paul Greengrass-mimicking fun it likes. Imagine what happens when the player stops on that roof and thinks “ok, so what now?” Not everyone is going to go for the helicopter: some people are going to look around the roof, try to find a weapon to defend themselves with or something, and that shakey camera is going to piss them off like you-don’t-know-what. I hate to break it to you (I genuinely do! It is a badass trailer), but if that sequence makes it into the final game, it is either going to be:
1) changed to a point beyond recognition
2) a cutscene
3) a quick time event
(PS: Panzer Dragoon Zwei is available on Gametap [though it doesn't really work on Vista] and S&P is on Wiiware [though you need a classic controller/wavebird])
@Hamish –
You’re not breaking any bad news to me. The point of a pre-vis is to organize the team’s efforts in exploring what is feasible. Beyond Good and Evil’s pre-vis is no different. To go with either three of your options though is to admit defeat before the first step.
Ultimately, the question isn’t “what will the player do or not do”. Things like “How would the player know that they could climb that pillar?” relate to gameplay and moment to moment decisions (you can climb pillars in many games) that cannot really be discussed with accuracy in a forum like this. Gameplay must be created to account for any player action. Some players will jump for the helicopter, some find a weapon, some run away, some will jump off the roof and die, some will put down the controller and make a sandwich, etc, etc to infinity (though gameplay should also dictate the importance of individual actions, i.e. if weapons aren’t found on the ground, there’s no need to search for one).
The real question is, what can the camera do to 1. react to player’s actions, and 2. aid the player with automated perspective. There are ways to make almost anything in the pre-vis work, but the question is how. Some of it is practical and fun, some of it wouldn’t be. The important part is to take something that most people would look at and say “that’s impossible”, place it in the hands of a designer and allow them to figure out how it could be possible. I’m saying it’s more important to explore new ideas rather than constantly retreat to old ideas.
The reason I’ve mentioned a few times the idea of the camera leading the player at 18 seconds into the Beyond Good and Evil video as they climb up the pillar is because it’s a conceivable way to make that particular camera work possible and practical. Now let’s look at the section you mentioned with the guards. A way to make stepping on the guard’s head possible and practical is to provide a sort of Prototype-esque automated or triggered parkour off an enemy. Now press the attack button in the air to trigger a targeted strike. Now we have a description of how to make that section work without full scripting. The challenge then is making animations that flow.
We could go through the entire video and pick it apart to decide what is necessary to make each piece work, which is exactly what I hope designers will do, then decide what is practical to include in the game. We can’t know what’s possible until the limitations are pushed, until an idea has been created and examined for its conceivability and tried. You think anyone designing a watch phone doesn’t feel like Dick Tracy now and then? Most anything invented started as fiction at one point or another.
I don’t want to sound patronizing, but I think your view of game development might be somewhat idealized. You say that “There are ways to make almost anything in the pre-vis work”, but this is only tenuously true. Even a pretty cynically written design document can find some things impossible to implement. This is not a case of a lack of creativity or manpower, it’s just a case of “this thing doesn’t really work in games”. That may sound pessimistic. It is not. Think of it this way: there exists a language of video games. In theory, anything can be said in that language. But some sentences, though they may get a message across, will be garbled and ungracious, because we’re not exactly sure how to speak the language and to best use it.
To put it another way, Games Are Not Films. It’s admirable to see you made so optimistic by that trailer, but… seriously, games do not work that way. That camera, I’m sure you’ve noticed, responds to her movements before she makes them. That’s an example of something unfeasable, but it runs so much deeper. When we play exciting games and we are playing *well*, it is possible for them to feel similar to that video. That is what I meant by an “idealized player”. The question that, I must insist, needs to be asked, is “how did the player get this good at playing the game?” The answer is: there is no way they could have done! The camera, the character’s abilities, the area-layout… everything in that video would conspire against them.
Saying these things, I feel acutely aware that they are not really sufficient… I honestly have difficulty trying to translate that video into a playable game. To try and take a step back from it, tension in games comes from difficulty. Ideally, it comes from strategy. Strategy means trying different things, getting some things right and some wrong, and winning or losing as a result. And that’s pretty cool! It means you have to use your head! And if you’re not using your head to achieve the thing on the screen, then you’re watching a cutscene.
That is what good gameplay is structured around. Another thing gameplay can be structured around is how cool something would look when it is done really well. They try, in that video, to make you think you’re looking at someone who is being challenged (with the stumble at 0:17 and the flail at 0:22 – isn’t that a bit of a giveaway?), but you’re not.
Argh… seriously, I’m being so inadequate. Would you believe I rewrote a lot of this several times? I can only hope that it’ll come to you by osmosis. I’ll round it off: it is *dangerously* easy, to make something work in your head. The most idealizing soul with all the manpower in the world might be able to make a game that looks like the *film* in your head, but it won’t stand up to playtesters. This world has Cinematic Platformers: “Another World”, “heart of darkness”, “Ico”, “Shadow of the Colossus”. I’ll concede that these games are great! Possibly because their makers are able to feel as though they are showing you something beautiful. BUT when they’re difficult, they’re annoying (the solution involves stepping back a screen or whatever), and when they’re easy, they’re unconvincing (if button inputs require no precision, why ask us to press them at all?). So what do people remember about SotC and Ico? Mostly they remember the *behaviour* of Yorda and Aggro and the colossi. That video gave you no behaviour; it only filled you with enough dreams of looking that cool yourself to make you buy the game.