Figure 1. Characterising Perceptual Opportunities
(The material from this page has now been greatly expanded and published as chapters 7 and 8 of Game Invaders: The Theory and Understanding of Computer Games.)
It’s just an ordinary day. You’re walking down a street near the middle of town when suddenly bullets ricochet around you off the walls and sidewalk. First things first. You need to find cover: down some steps, behind a low wall, down an alley? Not too badly hurt, you use the path of the incoming bullets and tracers to find out where your attacker is. You can just about make out someone way up above you on a parapet. You check that you’ve got enough weapons and ammo. You also need a vantage point. Fighting back from down here is not going to be easy. You’ve got to plan your way out of this.
Later on in this ordinary day you are driving through town on the wrong side of the road way above the speed limit. You notice a line of police cars blocking the road in front of you and you also notice a psychotic looking police car in your rear view mirror. You can’t turn round because the police in your mirror will ram you and that will probably be the end of it. Your attention is drawn to a small gap between two of the police cars ahead. You might be able to get through without too much damage to your car. That’s the only option. Make for the gap and hope your driving is good enough to get you through. You just planned your way out of this.
It’s turning into quite a day!
No surprise, the shooting incident is a shoot-em-up. Its Sincity, a death match (DM) level from Ritual Entertainment’s Sin. Being a shoot-em-up, this is a first person, multi-player level with guns and ammo, health and power-ups floating around just above the floor. Being a DM level the setting is constrained, not to say claustrophobic. Bullets and ricochets naturally attract your attention, as do a few moving pixels which signify an opponent. Keeping well armed and up high are good tactics and the major activity is the firefights. The pace of play is determined by the players and in Sincity can be sometimes about taking your time and being stealthy while at others fast and furious.
The consequences of your less than sensible driving habits are provided by Driver from Reflections. This is a third (second?) person single player driving game with the whole big, bad city for you to play in. There are no weapons or power-ups to collect but you’ve got a dynamic local map to help you. Everything around you is pretty natural except the big red arrow you know is out there somewhere signalling you’ve reached the objective for the level. Keeping your car free from damage is a good tactic. Not getting noticed by police is also a very good idea though not very often possible given the time constraints you have to work with.
Two very different games: different genres, different game plays, and different perspectives. Yet somehow the situations are very much alike. You are in a bad situation and have to find a way out or start all over again. You can take pleasure in using your skills, both motor and intellectual, to keep failure at bay. These are some of the recognisable pleasures of games. On the one side, the pleasures, we have similarity while on the other, genre; we see almost no similarities. It would be useful to know in more detail why the situations are so similar aesthetically and if there are general structural similarities that we could apply to all games?
Perceptual Opportunities (POs) are a generic content model for games and virtual environments in general which aim to make such analyses both practical and useful. We will dissect Sincity using POs to illustrate the model and then perform a similar dissection on Driver and then compare and contrast the two.
This isn’t to say that looking at games in terms of genre is wrong, quite the opposite, genre will be a major selling point and will also tell us a lot about player expectations and de facto standards. There are many people who would play Driver but not Thief despite the fact that, beyond genre, POs show us they are very similar in many respects. In any case, it will turn out that you can’t ignore genre completely when analysing games but the model will highlight this as well.
But first a powerup.
We can observe that a variety of quite different objects to be found in quite different games serve similar purposes in terms of game play. Static police cars in a road block and ricocheting bullets serve the same basic purpose of attracting our attention to some problem or hazard which has to be solved or overcome in order to progress with the game. Having had our attention attracted we then have to formulate a plan to solve the problem, or avoid it, or whatever. Objects we perceive around us, or whose existence and whereabouts we remember, will take on new meanings in our quest to resolve the current problem.
We can identify two types of meanings that we give to objects. In the ‘real’ world objects are predominantly denotative in meaning. They are themselves and function as expected: a cup is a cup, a chair is a chair, a glass phial is a glass phial, a door is a door, and so on. Sometime objects also have connotative meaning. We use them to make associations not obviously connected with their everyday nature or function. Souvenirs and mementoes are examples of this. Clothes are another example of everyday objects with connotative meaning. We choose clothes not merely for functionality but for what we would like them to say about us in different situations. The words on this page have exactly this type of meaning. The little black marks ‘cup’ within the single quotes - have nothing to do with the object cup except by connotation. On the whole denotative world meaning the thing itself as itself dominates our perception of the real world because that is how we use objects to do our job, earn our living, feed ourselves, and generally survive and, hopefully prosper.
In computer games objects also have the potential for both denotative and connotative meaning but I suggest it is their connotative meaning that is the most important. A gun is still a gun and has its denotative value as a weapon of offence or defence but many more objects than in the real world will have connotative meanings to do with game-play, strategy, and winning. A chair could be used to sit on but might, more likely perhaps, be used to stand on to reach an object that in itself has denotative meaning. The chair has a connotative meaning in the puzzle solving process. The glass phial on the shelf high above you might be a power-up. The key you can just reach after standing on the chair might unlock a door that gives you access to the treasure. Even such mundane objects as walls, floors and ceilings become not just corridors but objects in the game-play that restrict your choice and lead you on to confront the next situation.
The Perceptual Opportunities (PO) model is a characterisation of the roles objects are intended to play in establishing game play. The figure below shows how POs may be broken down into three principal forms, each of which focuses on different kinds of meaning that objects may offer. Sureties deliver denotative meaning and collectively try to establish basic believability. Surprises seek to deliver connotative meaning and thus collectively seek to deliver purpose. Shocks are perceptual bugs that tend to negate the other two forms by breaking the illusion. We will not pursue shocks in this article.
Figure 1. Characterising Perceptual Opportunities
The relationships between POs can be documented using perceptual maps, which are a sort of grammatical structuring that seeks to ensure that users construct an appropriate temporal ordering over their attentions and activities within the game.
Sureties should inform the game player of such things as; How big am I?, How fast am I moving? What do I look like? Have I been here before? And so on. Sureties also provide other reassuring information to do with such things as the physics of the world and the believability of other beings in terms of their avatars and behaviours. Furthermore, we are used to the real world being complex and cluttered so it helps if the virtual world of the game is as well. We call this redundant complexity perceptual noise. A useful aphorism is that in interacting with the real world we are trying to make sense of too much information whereas in games we are trying to make sense of too little.
In Sincity, such things as sidewalks, fire escapes, alleyways and doorways, roofs and parapets, and so on all reassure us. They tell us where we are, what sort of place we are in, how big we are, how fast we are moving, how high we are, how far away things are away, and so on and so on. The neon sign on top of the half-finished building doesn’t contribute to game play but certainly helps establish height and the believably ordinariness of the street scene. The general complexity of the buildings both finished and unfinished, the levels provided by roofs and parapets, the wire fences, ladders, doorways, etc. all contribute to perceptual noise and give our unconscious plenty to do.
Sureties succeed by not being noticed when they are there but would be missed if they weren’t. They are thus the basis on which the games designer seeks to achieve the willing suspension of disbelief in the mind of the player. If sureties are the basis of this then surprises are what really deliver the goods.
Surprises can be: implausible but beneficial or totally plausible but unexpected, and there are three basic types:
Attractors may be characterised according to the reasons they draw attention to themselves both denotationally and connotatively. First of fall the denotative or everyday forms of attractors:
However, although attractors rely on peoples' natural curiosity they are also directly related to the player’s emotional involvement with the game. A second and very useful characterisation concerns the connotative meaning players attach to attractors. Two of the most important for games are:
For the sake of simplicity we will assume only one opponent in this analysis. The following are some of the attractors to be found in Sincity.
Gun flashes aimed at you certainly focus the attention as active objects of fear. An extreme form of attractor that should stimulate an immediate response of the players part. The associated goal of find cover immediately would seem reasonable.
Above, the bottom two highlights are approaching bullets. Bullets and ricochets make great attractors, objects of fear, for obvious reasons but alignment also points out the top highlight which is the opponent, here just a few animated pixels, but a vital attractor all the same. This latter is an attractor because of its connotative meaning as an object of fear in this genre and would not be an attractor in all game genres. One goal here is similar to the attractor above find cover immediately but anther goal could frag opponent. We thus have an attractor as choice point here.
More attractors, on the left: a vantage point, an object of desire with the associated goal to take up an advantageous position. The two leftmost highlights are also vantage points. These are great attractors, objects of desire, in this level because you can ambush your opponent from them. Ammo, guns, health, and powerups are also great attractors, objects of desire.
The centre highlight is another attractor in the form of an opponent on the suspended girder, plus two more vantage points. The former could be both an object of desire and fear because opponents are your target and source of frags but they can also fight back.
Connectors are POs that help players by supporting planing to achieve goals stimulated by attractors. Connectors are thus the means by which players make connections, both mental and 'physical', between attractors and retainers which allow players to achieve their goals and deliver objectives specific to the purpose of the game. The actual objective of a retainer might well be hidden or not clear from the point of view of its attractor(s) but lower level goal formation should lead players into situations where objectives can be realised
In Sincity we can identify a number of connectors. Similar view but here the highlights pinpoint connectors such as: a ladder, suspended girder and ledge, which all make physical connectors supporting goals such as, take up an advantageous position. Players put connectors together to achieve goals. Other goals, such as collect health or ammo, might well be form part of a larger goal such as the one just identified above. So here the plan might be to get to the parapet above the highlighted ladder using that ladder and another hidden from view so as to satisfy a goal to gain a position of advantage. There might also be a sub-goal - collect some ammo - which can also be planned for and attained on the way. On the parapet is another gun and access to the movable girder to get to the unfinished building furthest from view.
Now we’re into a firefight, a retainer or mini-mission, with the opponent in our sights and maybe the element of surprise.
In Sincity we can see attractors, connectors and retainers all functioning together to deliver the aesthetics pleasures of a death match level. Surprises should work together in patterns to form possible temporal orders on retainers and thus the coherent set of purposive experiences that are intended to deliver the purpose of the world. We call these patterns Perceptual maps and they can be seen as a sort of invisible but comprehensible labyrinth that gamers will want to discover through the game play. Attractors should draw attention to sites of retainers and, if properly designed, lead players around the world in a meaningful way using connectors. Attractors may also themselves be rewards. Seen from a distance an animated object may act as an attractor but when experienced close up the object may be some sort of vehicle to ride in and control thus becoming a retainer. Retainers are actually localised patterns of patterns of attractors, connectors and rewards. Early computer games, such as Viper and Breakout, can be viewed as retainers in this sense.
Since surprises are of three forms namely attractors, retainers and deflectors they will require different kinds of information to specify them. However, some general rules apply:
For Sincity a table of surprises looks something like this:
|
|
|
|
| Ricochets
(Dynamic objects of fear) Goal is find cover |
Plan is make for cover
Uses doorways, walls, alleyway, etc. |
Activity is take cover
(Local) Reward is time to think, plan, etc. |
| Movement of opponent(s)
(Dynamic object(s) of fear and desire – your opponent can fight back) Goal is find cover |
Plan is make for cover
Uses doorways, walls, alleyway, etc. |
Activity is take cover
(Local) Reward is time to think, plan, etc. |
| Movement of opponent(s)
(Dynamic object(s) of fear and desire – your opponent can fight back) Goal is frag opponent |
Plan is take opponent by surprise
Uses guns and ammo and maybe cover. |
Activity is firefight
(dynamic, peripatetic) Reward is fun + increase frag count |
| Parapet
(Object of desire) Goal is take opponent by surprise |
Plan is take up position on
parapet
Uses ladders, wire fence, ledges, suspended girder, etc. NB. Can also collect ammo and weapons, health, powerups etc. |
Activity is ambush opponent
(dynamic, peripatetic) Reward is fun + increase frag count NB. Firefights are mini-missions in their own right and could be documented with their own perceptual maps. |
| Health, indicator
(Objects of fear) Goal is increase health count NB. Similarly for weapons, ammo and powerups, etc.) |
Plan is get to location of
health
Uses road, sidewalk, ladders and rooftops etc. |
Activity is walk over health
(static) Reward is increase health count NB. Mini-mission. |
| Illuminated sign over unfinished
building
(Awesome object) Goal is destroy sign |
Plan is select weapon and shoot
sign
Uses guns and ammo. |
Activity is take pot
shots at sign
(peripatetic) Reward is fun until you realise you can’t destroy the sign NB. A red herring because shooting the sign does not help you achieve the main objective – gain points by killing opponents) |
Of course, players might have a number of plans at any one time and will be taking notice of a host of attractors as they execute one or more of their current plans.
Perceptual maps have much in common with the way painters arrange the composition of a work so as to catch the viewers attention and lead it around the canvas in a particular way. Although it is not possible to tell a story in a game in the same way as in a film or TV program, there is never the less an important narrative potential in games which needs to be designed for. This refers to the purposive accumulation of experience, which allows players to exchange stories about their way through the game. A perceptual map is the basis upon which such narrative potentials can be designed.
Sureties and surprises in games work together much in the way jokes do:
First of all a few brief notes on sureties. Street furniture, building fronts, all give realistic sensation of appropriate speed. Buildings at a distance good, people and vehicles at various relative distances give both distance and scale sureties. There is a lot of detail in the city, buildings, moving cars and people, petrol stations, underground car parks, etc. all of which generate good levels of perceptual noise. Good self image. Being a third person driver you can only see car you are driving but that behaves very realistically, wheels turn to steer, suspension, collision damage etc. Car horns sounding at you give a good sense of other people around. Sureties for the past are provided by crashed vehicles, skid marks in grass verges etc. the remaining damage to your car and the police car. In terms of physics sureties you can crash through tables, chairs, parking meters etc. but not street light poles, and trees etc. Its very easy to leave the ground going over humps and bumps but this is fun so still provides good sureties. All in all we see a rich set of sureties.
We now move on to some notes on surprises, which we will represent directly as a table of surprises.
|
|
|
|
| Junction, turning or open space
ahead also indicated by traffic moving across current path
(Active, mystery objects) Goal is negotiate junction, move closer to objective, and avoid damage to car. |
Plan is to see if a turn is
required
Uses 2D map, roadways,
|
Activity is make a turn at
junction, alleyway, or open space etc.
(local) Reward is step closer to achieving the level’s objective and (hopefully) no damage to car |
| As above.
NB. This attractor and the one above together are a good example of choice point. |
As above. | Activity is cross junction
and ignore change of direction opportunity
(local) Reward is step closer to achieving the level’s objective and (hopefully) no damage to car |
| Police in rear view mirror
(Dynamic, active, Object of fear) Goal is get rid of police tail |
Plan is out-run police car
or make it crash
Uses traffic, street lights and trees, side turns |
Activity is swerving to
get police to crash into vehicles or street furniture behind you,
speeding and taking side turns to get away from a collision situation (dynamic) Reward is continue to attain the level’s finish location, the fun of the chase |
| Police roadblock ahead
(Dynamic, local, object of fear) Goal is avoid road block |
Plan is crash through road
block
Uses largest gap in the road block |
Activity is attempt to steer
through gap if there is one or handbrake turn to reverse direction
(dynamic) Reward is continue to attain the level’s finish location |
| As above
NB. Choice point. |
Plan is change course
Uses road behind etc. |
Activity is 180 degree handbrake
turn etc.
(dynamic) Reward is continue to attain the level’s finish location and fun |
| Police roadblock ahead and
police in rear view mirror (Dynamic, local, objects of fear) Goal is avoid roadblock and loose police tail |
Plan is crash through road
block
Uses largest gap in the road block |
Activity is attempt to steer
through gap if there is one or handbrake turn to reverse direction
(dynamic) Reward is continue to attain the level’s finish location |
| No attractor
(you should have seen this coming and avoided it) |
No plan
(you don’t plan to reach this retainer) |
Activity is extricate yourself
from collision situation with as little damage to your car as possible
(dynamic) Reward is continue to attain the level’s finish location NB. Mini-mission. |
| Pointer on dynamic 2D map
(Object of desire and mystery) Plan is get to level’s finish location NB. Offers a whole set of choice points. |
Plan is work out and follow
route to end point
Uses roads and junctions on the map plus road, junctions and traffic ahead of you on the street |
Activity is navigate to location
pointed to on map.
(Peripatetic) Reward is progress towards finish point and fun |
| Damage indicator high
(Object of fear) Goal is sustain no more damage |
Plan is drive carefully
Uses good driving skills |
No particular activity
(dynamic) Reward is prolong attempt to finish level. |
| Clock running down
(Object of fear) Goal is finish level faster |
Plan is drive faster
Uses advanced driving skills |
No particular activity
(dynamic) Reward is finish level before time runs out. |
| Police radio messages
(Sensational object of fear) Goal is avoid police cars. NB. This attractor also serves to heighten tension. |
Plan is keep a look out and
avoid police cars.
Uses rear view mirror, road ahead, 2D map. |
No particular action but switch
to new attractor if and when police cars are sighted.
Reward is anticipation of future problems |
| Café furniture, park benches,
garden fences etc.
(object of desire) Goal is destruction |
Plan is steer course for said
objects.
Uses said objects. |
Activity is collide with said
objects.
Reward is destruction of said objects. |
| Large floating red arrow
(Alien attractor, object of desire) Goal is complete level |
Plan is get to finish point
Uses road ahead to finish point |
Activity is stop under sign
if its an arrow,
(static) Reward is complete level and, watch pre-rendered sequence, success, a new mission NB. It might not be the end of the level but the completion of a section in the level. The rest is much the same. |
| Large floating red full stop
(Alien attractor, object of desire and disappointment) Goal is complete level |
Plan is waste some time
Uses local road plan |
Activity is keep going – you’re
too early
(static) Reward is you can still finish the level from this situation |
| Large floating red exclamation
mark
(Alien attractor, object of fear) Goal is lose police tail |
Plan is out-run police car
or make it crash
Uses traffic, street lights and trees, side turns |
Activity is swerving to get
police to crash into vehicles or street furniture behind you,
speeding and taking side turns to get away from a collision situation (dynamic) Reward is you can still finish the level from this situation |
If we compare the table of surprises for Sincity with that for Driver we see that, despite their differing genres and the fact that we are comparing a DM level with a complete SP game, the PO model does capture important insights that allow us to compare the two games. This is because we are dealing here with the underlying dynamics of game play. We can see now that the two original situations described in the opening paragraphs of this article were actually examples of attractor/connector/reward triples. Despite their superficial differences the model identifies them as similar structures. Furthermore, comparison of the two tables highlights a number of interesting things. The table for Driver is more than twice the size of that for Sincity. This is surely because the latter is a DM level and relies very much on player interaction while the former is a SP game and needs greater complexity to maintain player interest.
The attractors for Sincity display an even balance of objects of desire and objects of fear. This indicates that reward and risk balance out. In Driver the attractors are almost exclusively objects of fear with a few being objects of both desire and fear. There are only two objects solely of desire and one of these makes no contribution to possible success. Altogether this indicates that there is an unremitting level of risk in Driver. Both games use alien attractors though in different ways. In Sincity all guns, ammo, health and powerups are essentially alien because of the way they float just above ground. Health is directly connotative in that the object health represents a concept the individual’s physical well being. In Driver the three end of level markers are all alien but appropriately so. This is because it would be difficult to find a building or location at the best of times but when trying to complete a level in a damaged car pursued by manic police cars it would be nearly impossible. Also the added tension of almost but not quite having completed a level adds to the excitement of the game.
If we look at the rewards we see that the rewards in Sincity are concerned with both high level objectives such as increasing the frag count but also lower level objectives such as increasing health, weapons and ammo. The rewards in Driver are all focused on the overall objective reaching the specified location to complete the level. There are no rewards, which decrease damage or give the player extra time. From this, and the observations on attractors above, we can see that Driver is a game of attrition and requires more or less complete focus on the main objective. Sincity is very much about rising and declining fortunes and the player’s control over them.
This brings us on to the patterns of goal setting and planning in the two games. In Sincity it is the normal pattern of game play to have a number of goals and their connectors, plans etc., active at any one time. There are always the overall objectives of ambushing opponents and keeping a look out for opponents looking to do the same. However, players will also have lower level objectives, which they will operate concurrently, to get more ammo, weapons, and health etc. They will switch objectives quickly if a lower level objective become obtainable. Attaining a lower level objective does not remove it from the current list. In such games players will keep many such objectives active so they can satisfy them pragmatically as the opportunity arises. In Driver the high level objective of the level runs concurrently with the currently lower level objective which supports it and is to do with the current situation on the street. When the latter is satisfied a new lower level objective is activated to do with the next street situation on the way to the main objective. This pattern is only interrupted by the arrival of police on the scene at which point other short term objectives to do with immediate survival come into play and temporarily replace both the high level objective and its current lower level supporter.
We can thus see that POs and their perceptual maps can tell us much about the underlying structure of games and allow us to see something of why games are as are. We could also single out such things as the use of alien attractors to provide emphasis to be a common and useful design feature. This analysis also highlights the importance of understanding the connotative nature of even everyday objects in games.
POs are a means of analysing games in terms of the underlying dynamics of gameplay rather than in terms of genre, which can often group together games, which are quite different in nature. For instance, when analysed in terms of POs, Driver turns out to have more in common with Thief than some other games in the driving genre.
POs don’t at the moment give us the complete picture. They don’t tell us about the pace of gameplay or the nature of the virtual space in which the player(s) have to function. They don’t tell us about the look and feel of the environment and the way this affects players. More emphasis could be placed on skills and skill levels needed to form goals and execute plans. POs don’t offer insights into the complexities of the physical interface the mapping between mouse and keyboard controls and the perceptual interface the perceived weapons and their usage.
This said, POs do allow us to think about why certain game features are successful and how they might be translated to other genres. POs have also been used to teach game design to both graduate and undergraduate students on games design and programming degrees at the University of Teesside. Interestingly they have also been used to teach virtual reality design and allow the comparison of computer games and virtual environments and to thus apply lessons learnt in one discipline to the other. We can also use POs to investigate the aesthetics of games as well larger structures from which games are composed and can also be investigated. In the words of Doug Church what we are aiming for here is a way "to build on past discoveries, share concepts behind successes, and apply lessons learned in one domain or genre to another".