DETAILS


COLUMNS


CONTRIBUTIONS



GAMING & GRAPHICS

Vol.33 No.3 August 1999
ACM SIGGRAPH


What’s Your Perspective?



Richard Rouse III
Paranoid Productions


August 99 Columns
Entertaining the Future Visfiles


Gaming & Graphics
Previous Gaming & Graphics Next Gaming & Graphics

Doom
Tomb Raider
Figures 1 & 2: In the first-person Doom, players see exactly what their character in the game world would see, including the weapon they’re currently firing. Tomb Raider offers the player a full view of their character, Lara Croft.


“First person is dead as a perspective.” So said one of the designers of one of the most famous first-person perspective PC action games of the ‘90s. I only heard this declaration second hand through a friend of said designer, and as such won’t mention who the developer was, or which game he designed. Nonetheless the statement is representative of a sentiment one could see throughout the industry following the runaway success of Tomb Raider, the first hit action game to use a real-time 3D engine with a third-person perspective. The recent success of such first person games as Half Life and Rainbow Six seems to have revitalized the popularity of that perspective in the gaming community, but still the debate between the viewpoints continues. What may seem to a developer as a chance to show off the cool moves of a well-defined main character, actually has deep ramifications on the type of game one creates, both in terms of game design and the resulting emotional attachment the player develops within the game.

It seems the key comparison to make in considering first and third person PC action games would be between two of the most popular games in each category: John Carmack and John Romero’s Doom and Toby Gard and Paul Douglas’ Tomb Raider. Doom established the real-time 3D first-person shooter style of game, using a design that largely survives in first-person shooters released today. In Doom, players storm through maze-like levels seeing the world exactly as if they were the character they are controlling - their surrogate in the game world. The Space Marine in Doom is a barely-defined character, a blank-slate onto which players can project whatever kind of personality they want. For Tomb Raider, the developers took related technology but changed the game’s design completely by making the player see their surrogate in the game world, witnessing the character’s deft footwork as she dodged around corners and jumped over pits. Here the central character is Lara Croft, who has a very specific style to her appearance and actions. Though she doesn’t talk much in the Tomb Raider games, Lara is infinitely more defined than the Space Marine from Doom.

Difference of Views

When thinking about the differences between a first-person and third-person game, foremost in many a player’s mind is the most obvious change: the camera. In a first-person game, the camera always looks wherever the player is looking. Since the character is always in a “valid” location in the game world, the camera is always in the same position relative to this character. In a third-person game, however, this is not the case. The preferred view of the player’s character in an action game seems to be an “over the shoulder” camera location, far enough back so that the player is able to see all of said character.

The problem then becomes what if the camera is in a position that isn’t in the game world; for instance if the player’s character is backed up against a wall. Since most real-time 3D engines can’t handle rendering from viewpoints at invalid locations in the game world, the camera must be moved to a valid location. Where should the camera go? Further up? Off to the side? Should it zoom in on the player’s shoulder? All the different options have their own strengths and weaknesses, with no solution sure to please all players, and the game is bound to be somewhat disorienting in certain situations as a result. Programming a camera which is acceptable to all players is a daunting task for third-person game programmers, a problem which is completely circumvented by first-person games.

The visibility of the player’s character was not the only thing that changed between Doom and Tomb Raider. Doom is much more a pure action experience, with puzzle solving taking a backseat to fast-action, trigger-finger exercise. Tomb Raider, on the other hand, is much more of an action adventure, with the player needing to figure out puzzles, and divine means of attaining difficult to reach positions. Though Tomb Raider has its fast-action moments, these are separated by long puzzle-solving passages. However, the switch from pure action to action adventure was not done merely because the developers felt like it; the removed view of a third-person game lends itself more to navigating the player’s surrogate through the world instead of aiming and shooting at its inhabitants. Indeed, Tomb Raider would be all but unplayable without its auto-aiming functionality; the player doesn’t need to have Lara Croft point directly at an eagle overhead. Simply pressing the fire button causes Lara to shoot at whatever target may be around. This simplification of shooting brings the game further away from an action-fest like Doom. In Doom much of the challenge is hitting your demonic adversaries, and the player’s ability to determine exactly what they’re shooting at is key to this style of gameplay. When the player is no longer seeing through the eyes of their game world character, aiming becomes much more difficult and unintuitive, and a change to a style of gameplay in which aiming is not so central becomes necessary.

Player Immersion

One of the key points brought up in favor of the third-person perspective is that the player gets to witness their game world surrogate performing “all those cool moves.” To be fair, there is a certain thrill to watching your keyboard commands translated into your character performing balletic flips while firing a pair of pistols. But this oft-cited advantage to third-person play actually plays a crucial role in one of the fundamental challenges of game design: player immersion. By being able to view their surrogate performing the actions they command, players intuitively realize that the character doing all those cool moves is very much not them. In a first-person game, the player sees their actions carried out by the movement of the camera through the world or the changing state of that world as viewed by their character. Thus the player is more drawn into the game and might - for brief moments in time - even think they actually are in the game world.

In its October 1998 issue, Next Generation magazine ran an article about famous characters from gaming history. It included an interview with Toby Gard, one of the creators of Tomb Raider. The magazine was kind enough to ask a question I would have asked myself: “Do gamers watch the lead character or take on the role of the lead character?” Gard’s response was: “I think it depends on the game. Generally speaking, if it’s third-person, then you’re watching and controlling a character external to yourself. This allows us to give that character more personality of their own, and the player, suitably distanced, doesn’t find it disconcerting when the character does things of its own accord. In a first-person game you can’t do that because you’re meant to be taking on that role, and as a player you expect to put all the personality of that character in yourself.”

Gard nails the issue dead-on: third-person games provide “suitable distance” between not only the player and their character, but also between the player and the game world. Hence the player’s sense of immersion is significantly less. I’ve always found one of the greatest appeals of computer games as an art form is that they allow the player to be in control, to see the consequences of the actions they choose instead of, say, the consequences of the actions chosen by the characters in a book. Immersing the player in the game world as much as possible is important to this “ownership” of choices made in the game world. As the player gets “suitably distanced” from the game world, the experience becomes less about making their own correct decisions in the game world, and more about prodding another character to make the right choices. In the latter option, it seems the player inherently has much less of an “ownership role” in the whole process. Whatever choices the player makes, it will appear to them that someone other than themselves is performing these actions, and hence the actions aren’t really theirs at all.

Duke Nukem 3D
Figure 3: Duke Nukem 3D used a first-person perspective with gameplay very similar to Doom. But the game went to great pains (mainly through dialog) to define the personality of the character the player controls, Duke Nukem himself.


Making a game third-person instead of first-person allows the game designers to give the player a much stronger, fully-defined character. Witness the difference in definition between the Space Marine in Doom and Lara Croft in Tomb Raider. Duke Nukem 3D provides an interesting mix between the two games in terms of player-character definition. A first-person game, the game leaves no doubt in the player’s mind what character they’re playing and just what sort of a fellow he is. Mostly communicated through snippets of conversation (or “Duke-speak”) during the gameplay, Duke comes off as a crass, death-loving, misogynistic action hero, and the player controls this character, love him or hate him. In many ways Duke is more fully defined than Lara Croft. At the same time the player is still immersed in the 3D world via the first-person perspective. This allows the game to be more of a straight ahead action game in the spirit of Doom while incorporating a strong player-character. Duke Nukem 3D did offer a third-person camera option, and this was no doubt added in order to satiate the feeling that one wanted to see the actions of this talkative, fully-formed character they were controlling. Due to the nature of the game’s gameplay and technical weaknesses in the implementation of the camera, however, this third-person camera is not really practical for playing the game. It seems like many fewer people (if any) had the desire to play Doom from the third-person, probably because the player-character was so unformed and amorphous that players felt they were that character. Barring narcissism, why would they want to look at themselves running around the Doom world any more than they would want to watch themselves going through life?

It’s no surprise that when the Duke Nukem character appeared on a console in Duke Nukem: Time To Kill, the game was played from a third-person perspective. It only made sense; in addition to the fact that nearly all console games - be they 2D or 3D - are played from a third-person perspective, Duke Nukem was already such a strong character that it follows logically that the player would want to see him firing his weapons and charging into combat. I was never as big a fan of Duke Nukem 3D as I was of Doom or Doom-a-likes such as the Quake series or the Marathon cycle. I believe I can trace this to the fact that I never liked Duke Nukem as a character, and his constant chatter, though amusing on one level, simultaneously distanced me from the game world. And since I didn’t really care for Duke, my desire to “guide” him through this game world - even in the first-person - was decreased by the fact that he was defined so strongly.

Viewing Adventure and Role-Playing Games (RPGs)

Myst
Leisure Suit Larry
Figures 4 & 5: Myst was certainly not the first adventure game to use the first-person perspective, but it’s probably the most famous. Leisure Suit Larry wasn’t the first third-person adventure game, but it may be the most infamous.


The Last Express
Figure 6: Jordan Mechner’s The Last Express offered many clever innovations to adventure gaming, one of which was the alternating first and third-person viewpoints found in the game.


Stepping outside the world of 3D action games, one finds that the first versus third-person war has been fought in other gaming genres as well. One such genre would be adventure games, where one can compare the perennially-popular Myst with the whole Sierra or LucasArts style of adventure games, such as the Leisure Suit Larry or Monkey Island series. In Myst, the player sees exactly what they would see if they were wandering around the game world, and the game seems to be designed around the concept of allowing the player to “step into another world.” Myst’s lineage can be traced all the way back to the days of the text adventure and the reign of Infocom, where players were presented with text descriptions of their environments. One might read: “You see an old house with a dog out front. The door is ajar, and the dog is barking at you.” The logical extension of this description into a graphical world is to draw a picture of what the player sees, as did many of the early graphical adventure games such as those made by Penguin Software or early Sierra titles. It was many years before Sierra hit upon the idea of making third-person adventure games, where the player doesn’t see what their characters in the game-world see, but instead sees a view of their surrogate navigating the game world. A game like Al Lowe’s Leisure Suit Larry, for instance, has the player guiding a small Larry around various environments. This mode of adventure gaming allowed the Sierra designers to bestow the player surrogates with more character, eventually allowing designers to have these characters carry on conversations with people in the game world without it seeming all that strange. Of course one hears of people getting “sucked in” to the immersive, first-person world of Myst much more than they do into the world of Leisure Suit Larry. In the latter, one is only trying to save Larry, whereas in a game like Myst one is trying to save oneself. At the same time, players can think back at what a funny guy Larry is, while their memories of Myst have much more to do with the world they encountered than the character they controlled.

Jordan Mechner’s The Last Express adventure game offers an interesting hybrid between first and third-person perspective gaming. When the player has control of their game world surrogate - Cath - the action takes place in first-person, with the player seeing exactly what Cath sees. When Cath walks up to someone who is interested in talking to him, the action suddenly switches to third-person, and the player watches a conversation between Cath and the other character unfold. In these cutaways, the player comes to see Cath as a character separate from themselves, due in part to the distancing effect of the third-person perspective, and also because the player is unable to interact with the game during the conversations. But once the conversation is over, the gameplay switches back to first-person, which allows the player to again become immersed in the game world. When I interviewed Mechner several years ago, he explained how the effect is not unlike that used in some films which attempt to have the player empathize with the main character through use of the first-person perspective. For instance, in Alfred Hitchcock’s Rear Window, the viewpoint continually switches between first-person shots of what Jimmy Stewart’s character, L.B. Jeffries, sees out his window, and back to a third-person perspective when Jeffries is arguing with his girlfriend or talking with other characters in the movie. By switching between an immersive, first-person world exploration mode and a distancing third-person view during scripted conversations, The Last Express, like Rear Window, attempts to give the player the best of both worlds, and it succeeds to some extent, creating a unique hybrid others designers would do well to explore further.

In role-playing games, the conflict between first and third-person has been going on nearly as long as in adventure games. In the mid-eighties, my choice to play The Bard’s Tale series of games instead of the Ultima series was influenced in no small way by the fact that the former presented me with a first-person perspective and the latter with a third-person view. The conflict continues to this day with new Ultima games (in both on-line and single-player forms) still being third-person games, while the Might and Magic series holds strong with its first-person perspective. Interestingly, despite presenting a view of the world from a third-person vantage point, the Ultima series still downplayed the character of the game world surrogate the player was viewing. Perhaps this was a hold over from the pencil and paper role playing games from which Ultima was spawned; in these games - such as Dungeons and Dragons - players are very much encouraged to create their own characters and make them behave exactly as they want them to, not the way the Dungeon Master tells them they behave. Thereby, even though the early Ultima games allowed the player to view the action from a distant view point, they still kept the characterization of the viewed character up to the player’s imagination. To me, The Bard’s Tale and its first-person view offered more immersion in the game world, and hence was my computer role playing game of choice.

Different Views for Different Agendas

It’s important to realize that the shift from first to third-person in any computer game represents not just a switch in what the player is allowed to view of the world, but also a transformation in the type of game being played. Certain games designs will all but cease to function when viewed from any viewpoint other than the one they were designed to use. My most recent game, Centipede 3D, is one such game. Try playing the game in first-person and it becomes completely impossible. The goal of Centipede 3D is to provide the player with a 2D arcade experience - like the one found in the original Centipede - in a 3D world, and a 2D arcade experience dictates that the viewpoint of the game must be from the third-person. On the other hand, some games seem to actually work in a variety of view modes. Jedi Knight, though primarily designed as a first-person shooter, seems to have a good enough camera implementation that it can be played from a third-person view as well. In fact the designers of Jedi Knight encourage the player to switch to the third-person view while engaging in light saber duels, a view mode in which those duels actually play better. With such games as Jedi Knight and The Last Express encouraging and demanding (respectively) a switch between first-person and third-person perspectives, one must come to realize that neither view is necessarily superior to the other. Different views will allow different games to accomplish different artistic goals. Multiple views at different points within the same game may even allow that game to accomplish different artistic goals at those different points.

Many players of intense, well-designed first-person shooters find themselves physically leaping out of their real world chairs in order to dodge game world projectiles. This can only imply a level of immersion within the game world where the boundaries of reality and fantasy get blurred for the player, at least for as long as they’re playing the game. It would be interesting to do a minor psychological study of game players motivations to see if they felt they were “saving themselves” or “saving Lara” in a third-person game such as Tomb Raider. How removed does the third-person perspective make the player, and is this so removed that one should switch to first-person views to properly suck them in? As a designer sets out to create a game, they must consider what their artistic goals are in the creation of the game, and weigh all that is being lost and gained depending on the type of perspective they decide to use.



Richard Rouse III is Lead Designer and President of Paranoid Productions and has published two games to date: Odyssey - The Legend of Nemesis and Damage Incorporated. He recently functioned as Lead Designer and AI Programmer at Leaping Lizard Software on Centipede 3D, published by Hasbro Interactive. Having at long last completed work on the PlayStation version of that game, Rouse has opted for a change of view and has moved on to a new project at Surreal Software.

Richard Rouse III
716 11th Ave E.
Seattle, WA 98102


The copyright of articles and images printed remains with the author unless otherwise indicated.