Spaceship Shenanigans

Summary

This game resembles the arcade game Missile Command, and tests distribution and shifting of attention, and the effect of multimodal (auditory and visual) stimuli.

Premise

The WSS White Giant is under attack! It is your job as Tactical Officer to guide the defense system to aid your fighters in shooting down incoming squadrons of enemy Wasp aircraft as well as colony against the attacks from the Wasp commander itself.

Objective

Protect the WSS White Giant, destroy all enemy Wasp aircraft, and defeat the Wasp commander.

Characters

Cheetah C-24 Fighter: See Meteor Madness

WSS White Giant: The colony that needs to be defended. It is armed with a Mobile Laser Defense Turret (MLDT) as its primary means of defense. It also gets support from the various C-24 fighters it carries.

Wasp Fighters: Basic Wasp aircraft. They attack WSI ships in waves and kamikazi into the colony to take out its defenses.

Cargo Box: This rewards resources and warps in right next to the colony.

Wasp Commander: Enemy commander who's impervious to almost all attacks. After the main Wasp force fails, directly assaults the colony.

Gameplay

Colony Defense

The player defends the WSS White Giant, the Colony, in the center of the screen, floating somewhere in space. This colony is armed with a mobile laser defense turret (MLDT) capable of firing charged lasers from the colony to the edge of the screen. The MLDT hovers around the edge of the colony and sweeps clockwise (right arrow) or anticlockwise (left arrow) along the circular defense perimeter.

One issue with placing the base in the centre of the display rather than at the bottom of the display is that it could complicate the detection of brain electrical potentials: stimuli from the top half of space are processed in the inferior occipital lobe whereas stimuli from the bottom half of space are processed in the superior occipital lobe, so we will be recording voltages with grossly different scalp distributions and will be unable to average them together. We could go this route if it distinctly improves game play, but the price that we pay for this flexibility is that we will need double the amount of game play in order to achieve sufficient numbers of trials for successful data analysis, since we'll have to analyse targets in the upper and lower visual fields separately.

If the requirement of twice as many tests being needed is the only issue with having the colony in the center rather than in at the bottom of the screen, then it won't be a problem. The design of the game has already been lengthened to included more phases to help isolate the various tests and due to the central design, about twice as many tests can be fit into the same time frame because of the shorter distance between the edge of the screen and the center of the screen than the distance between the top of the screen and the bottom of the screen.

Sure the centre placement gives twice the amount of space on the display, but it doesn't give twice the amount of time in which to present stimuli. So I'm not certain that twice as many stimuli can be fit in.

As game play progresses, various enemy alien aircraft, Wasps, appear from off screen and dive at the colony in the center of the screen in an attempt to destroy it. Later on in the game, friendly ships, C-24 fighters, will appear from off screen as well, though less frequently. They will then begin to assist in the attack on the enemy alien aircrafts, which will fight back against the C-24s.

In addition to enemies, cargo boxes appear. Cargo is warped in directly next to the colony. When this is about to happen, a cue signal is sent out by the colony letting the player know from what direction the cargo will appear.

Shooting

Every time the space bar is pressed, a laser is fired from the MLDT. There is no cooldown on the laser and the laser shoots to the edge of the screen piercing through all objects it hits.

When you write "stop the ship," to what ship are you referring?

In what way is the laser's coverage area increased? If the beam widens, could this widening interfere with the measurement of reaction time, for example in a case where a target warps in within the wider coverage area so that no movement of the turret is necessary?

The "stop the ship" was a typo which has now been fixed. As for the area of coverage for charging the laser, this has been changed after careful consideration to not allowing charging now but to just shoot a laser after every press of the space bar.

Cargo

In every stage, a cargo box materializes on the defense perimeter every 5 to 10 seconds since the last one was collected or warped out. It will appear either to the left, right, above, or below the colony. An early-warning system activates a stimulus in one of three different conditions of sensory modality:

Unisensory Visual

Display a large, 10 hz (50% duty cycle) flashing arrow at the center of the colony. The arrow appears either 100 ms (one flash) or 800 ms (eight flashes) prior to the materialization of the cargo and remains until the cargo materializes.

The early-warning system is fallible. 20% of the time, it indicates a sector 180º away from the cargo's actual sector.

This aspect of the task combines a variation of a centrally cued Posner attention paradigm (testing effects of cue validity and cue-target onset asynchrony) (Posner et al. 1987).

Multisensory Audiovisual

The arrow is accompanied by a klaxon (100 hz square wave modulated at 10 hz in synchrony with the arrow).

Unisensory Auditory

The klaxon alone.

If we were to go to two quadrants instead of four (i.e., placing the base at the bottom of the display instead of in the centre), we could perhaps use a monaural (left or right) klaxon to give some directional information, akin to the valid or invalid direction cue that's given by the arrow.

Would it be better to remove directional specificity from the test altogether to focus strictly on the multi-modal aspect?

I have thought about this, and on reflection, we can go with the plain, non-directional klaxon. It seems useful to "kill two birds with one stone," as it were, in getting both a Posner attention-shifting paradigm and a unimodal-multimodal paradigm out of the same single task, if we can pack both in.

The player has until 800 ms after the cargo materializes before it warps out. The player presses the up-arrow key to grab the cargo using the grapple beam.

Why the 800 ms absolute time limit? This could lead to frustration. Is there a way to motivate the player to react quickly that would still provide the player something instead of nothing if (s)he takes longer than 800 ms? Also, regarding the 'G' key, requiring the player to move their finger to a small key distinct from the key that the player uses for other functions is going to increase reaction time very markedly. Is there a reason for not using the up-arrow key, as Maritime Defender does for the wormhole connections?

The 800 ms was arbitrary. In the visual spatial article, the test used 2000 ms as a time limit for stimuli; would this be more acceptable? Also, the 'G' key was considered to reference the term 'Grapple'. Using the up-arrow key is a better idea since it maintains consistency between mini-games.

As we discussed in conference, it's best to implement some continuous decrease in reward rather than an abrupt and frustrating vanishing of reward - perhaps this could be done by having the cargo slowly fade back into the warp whence it appeared.

This aspect of the task is a test of multimodal (audiovisual) integration. (Senkowski, Molholm, Gomez-Ramirez, Foxe)(auditory_visual_integration_2006.pdf)

Phases

Phase 1

Enemy ships fly in from off screen, toward the base. Enemy ships might vary their path slightly and move at a constant speed. Enemy ships aim to crash into the perimeter to weaken it, which will result in fewer bonus resources for the player due to the need to repair the damaged colony. (This amount of decrease will be determined along with the Colony Simulator design.)

If an interesting game can be implemented without path variation, that probably would be most desirable. Otherwise it may be difficult to ascertain what the player is reacting to at any particular time, for example if a meandering enemy moves from one defence sector to another.

The variance in path would be slight, such as moving along a sin wave path where the start point is at the edge of the screen and the end point is the center of the screen, where the colony is. This should not cause any issues with trying to ascertain what the player is reacting to since this kind of path variation should still restrict the enemy ship to its own sector.

Stage 1

1 enemy ship spawns 1850 ms to 2150 ms after the player clicks 'Start'. Enemy ships spawn one at a time 1850 ms to 2150 ms seconds after the previous ship is destroyed. A total of 10 enemy ships will spawn.

Stage 2

2 enemy ships spawn 1850 ms to 2150 ms after the player clicks 'Start'. Enemy ships spawn two at a time 1850 ms to 2150 ms after the last visible ship is destroyed. A total of 20 enemy ships will spawn.

Stage 3

1 enemy ship spawns 1850 ms to 2150 ms after the player clicks 'Start'. Enemy ships spawn one at a time every 2850 ms to 3150 ms. A total of 10 enemy ships will spawn.

Stage 4

2 enemy ships spawn 1850 ms to 2150 ms seconds after the player clicks 'Start'. Enemy ships spawn two at a time every 2850 ms to 3150 ms seconds. A total of 20 enemy ships will spawn.

Phase 2 (Distractor Session):

Friendly ships (distractors) fly in from off screen and orbit the base. Friendly ships fire at enemy ships every 3 to 4 seconds. Enemy ships fire at friendly ships every 3 to 4 seconds. When completing the game, the more friendly ships left alive, the more resources awarded. (This amount will later be determined along with the Colony Simulator design.) This is a positive incentive.

Reaction time within this paradigm measures the player's ability to distribute attention across a broad field in the presence or absence of distractors.

Phase 2 repeats the same stages as phase 1 in regards to enemy spawning.

Friendly ships spawn 20% of the time each enemy spawns. A friendly ship can randomly spawn in any of the four quadrants with equal chance.

Boss Battle

Based on the basic five-box test from J. Townsend et al. - Brain Cognitive Research II.

After enough(?) waves of Wasps have been destroyed, the Wasp commander appears above the colony and starts to attack it. The Wasp commander is a large floating head with 5 eyes, sharp pointy teeth, and antennas that the commander shoots lasers at the colony and MLDT from. The player must continue to use the MLDT in order to drive away the Wasp commander while dodging the laser blasts it emits.

We have to be mindful of how dodging the lasers interlaces with shooting the eyes -- since this is a focused-attention task, we want to avoid the laser blasts' drawing attention away from the target eye, even momentarily. So I'm uncertain whether we'll be able to keep the laser blasts. (And given that, does it make sense at all to frame it as a boss game, or would it make too boring a boss game?)

Rather than diverting some of the player's attention away from the eyes by having them avoid lasers, we could instead assign key presses to each eye and require that the correct key corresponding to the correct eye to be pressed when the eye is open in order to count as a hit, otherwise the open eye will close.

In order for the player to actually be able to damage the commander, they must shoot at its weak point: one of its eyes (the attended location). The vulnerable eye will be colored slightly different from the others which will let the player know that that is their target, all other eyes simply pose as distractors since a successful shot at one does nothing to the commander. Only one eye opens at a time (the stimulus). When the player fires the laser, it will automatically aim at the target eye. When no stimulus is visible, the MLDT fires at the impervious exoskeletal armor of the commander, essentially doing nothing.

Scoring

Cash is awarded for each Wasp destroyed in any phase. Bonus cash is awarded for collecting cargo boxes. A special cash bonus is awarded for defeating the Wasp commander at the end of the level. (The actual values of the cash will be determined along with the Colony Simulator design)