First of all, let's be honest, the AR technology on mobile devices is very far from its true potential. Because of obvious technical limitations, it takes only baby steps toward the full expression of its talents. In some cases, the name of the technology itself has a greater affect on the audience than the real application's functionality. This means that to provide a truly viable product, rather than a single-purpose, technical gimmick, you should be less-ambitious, and the goals should be chosen very carefully and maturely.

Remember that there is no truly reliable way to generate an accurate model of reality on the screen of a smartphone camera yet. Modern devices can scan and recognize only a small part of the real world, since the screen representation is only an illusion and you have do your best to make it very convincing and attractive. Modern AR solutions should be used on iOS devices only when there's a creative idea and it is clear that a game will gain some advantage if it were to utilize the advantages of the AR environment. In other cases, it is better to skip the idea and use more traditional approaches. Remember that an exciting demonstration of a technology is not yet a game; games are not about an attractive technology, but about interesting playability.

To plan an AR game, it's better to start with the weak points of current technology implementations that are based on graphic-based markers and the standard camera (one the most popular methods today).

Markerless AR games are not ideal either. They try to track reference points from the surroundings. This job is not always perfect so virtual elements shift in wrong directions very often, and the process is not smooth enough. The games that generate a 3D landscape of the space around are conceptually interesting, but they require players to make special preparations and adjustments, and some people may be confused or get unsatisfactory results.

You should remember the psychosocial aspects of behavior invoked by AR games. In some cases, a person playing an AR game can look very strange, moving more actively, especially if they are involved in a gameplay of AR shooters. Many games force players to spin around pointing at devices that are on the walls or buildings around. This may be uncomfortable both for the players and the people around them.

Now you see that there are some problems in AR gaming. Many of them cannot be resolved easily without changing some technical paradigms. However, you can try to develop rules and conditions so that players will accept that these disadvantages are not noticeable. The audience should be guided by special directions, showing only the bright side and not the troubles backstage. Primarily, it should be clear that an AR game is not a universal product for the masses as many traditional mobile games are. It is a specific application for a specific gaming experience. This should be explained to future players so they are prepared for this fact.

In the beginning, it is necessary to invite the audience to a proper environment where the game will feel comfortable and can be expressed well. In its name and description, appropriate guidance words may be included, such as top-table fight, table battle, and living room race. Then, photos of ideal spaces to set up games can be included in the promotional materials and the tutorial of the game. It is as though the product is honestly admitting that there are some conditions. The model of proper behavior can be additionally implemented by assigning the game to a reasonable gaming genre and suitable mechanics.

Some of the most popular environments for AR games are horizontal surfaces such as tabletops or floors. A fiducial marker is placed somewhere on them and the game generates virtual fragments around it. The coordinate system of the scene is strongly defined by the marker. Such a paradigm can be referred to as the pole AR game. A game scene has a fixed center and all events take place around this point. In some senses, this concept is very similar to early single-screen arcades and fixed-shooters games. The players cannot leave their specified location and must fight all their battles centered around that spot. It looks like technological postmodernism—classical mechanics rethought using modern principles.

The motion of characters and other game pieces can be organized based on several basic schemes:

The pole AR games have a small creative issue; as long as an AR game scene has no strict screen frames and a player may see everything constantly changing in their perspective and vision angle, some procedures cannot be hidden. I'm talking about the spawning of characters. In a traditional game, enemies may be generated out of the current window (behind the back of a player, somewhere around the corner of a 3D map) and then successfully introduced into a main scene. But the AR scene in most cases cannot use such tricks. Therefore, some other tricks should be invented on the fly. Usually, authors simply demonstrate characters that come out of the ground (this is very good for stories about zombies) or dropping down from above (this is suitable for war games). Additionally, doors or hatchways can be used as if they were on a surface. Games with a science-fiction background may utilize a metaphor about space portals.

Another interesting idea is special, flying transportation vehicles that land troops who then disappear from the camera's view so that players cannot easily track where they went.

By its appearance, the concept of the pole AR game is very appropriate for various types of board games. As a rule, they are played sitting at a table, so seasoned players will be familiar with the requirement that a tabletop is needed for the process. In AR, you can create a fully virtual game board or add some additional functionality to traditional game pieces made of paper or plastic. For example, special playing cards can have fiducial markers so by placing them on a game board, players can interact with a virtual elements of the game. Mixed reality is very suitable for various race games, in particular those where flying vehicles are presented, as long as such a concept does not require very accurate tracking of the surface where the game is transpiring. It is better to use aircraft with propellers to animate the scene further. Some strategy-based games are good as well; a tabletop with a lot of tiny units and real-time combat makes a great scene!

You could imagine an AR game concept where some scroll effects are used since the fiducial marker cannot be left behind, the realistic movement of characters in space is replaced with a treadmill-like solution. A character is fixed at one position, but obstacles (or enemies) are moving on a virtual conveyor belt toward him; for instance, in a simple SteepleChase game. Such mechanics are called AR-projected scrolling, and are not very realistic since they simulate a moving surface on something that is stood still. Nevertheless, that can be resolved using various creative tricks. For example, the game may use very minimalistic artwork and the obstacles may be designed as simple geometric shapes, so the behavior of such obstacles won't appear that unnatural.

A panoramic AR game is another popular type of game mechanics. Its major advantage is its lack of graphic fiducials. A player is in the center of a so-called sphere of events, a space where all game elements are concentrated. AR in such games is not as accurate as models with graphic markers; this disadvantage is usually compensated with faster gameplay and simpler game rules. Usually, this is used in shoot 'em up arcade games or shooting-range games, in which enemies appear around players and they have to move their mobile devices quickly, while aiming and shooting. The most reliable concept is that of fighting flying antagonists, such as alien spaceships or hostile aircrafts, since flying elements look very natural in AR. An interesting concept is to use some ideas from paratrooper-type arcade games, such as tiny airplanes flying above a player's character that may drop some troops, so a lot of angry parachutists are.

A screenshot from AR Earth Invasion shooter game

However, there are some issues with the panoramic games. First of all, the basic concept requires players to spin around to find and fight protagonists; the gaming process looks like simple physical physical jerks and that may sound pretty positive. However, it may become tedious and cause a little dizziness if the gameplay is very fast. Therefore, it is better to include an option that turns the sphere of events into a semisphere, so instead of 360-degree range, a 180-degree range would be used by a game. Another problem deals with the question of game fragments situated outside of the camera frame, for example there can be several enemies which are placed behind the back of the player. They cannot be seen until they do some actions. There must be a signaling system to indicate that an enemy is hidden somewhere. This can be a mini-map or radar on where characters are marked or special red arrows may show the direction of enemy fire so a player can quickly re-orient his mobile device. By the way, the radar can be turned into a general element of the gameplay; a game can hide something in a random position within the sphere of events. The element is very tiny so it cannot be easily noticed from a distance, but the radar can hint at its location by changing the tone of its sonar sound or using special animation. That can be an improvised adventure game.

Remember that for an AR game to be perceived as a reliable product should be very observant and honest. If the system notices any trouble with the AR environment, for instance, the fiducial is lost or it is too dark, it must pause the game and display an alert so a player knows the current conditions are bad and should be adjusted. The interruption of the game process in the case of technical difficulties is necessary; otherwise, it can be very frustrating. The game scene disappears from the screen, but events continue somewhere within the application, causing a loss of game progress. The image-recognition algorithms can report some problems with the light. Additionally, it may track a fiducial and pause the game if a very sharp and dramatic change of values occurs, signaling that the marker is lost. Hence, it is good to have a special support system in a game to avoid misunderstandings with players.