POSSIBLE TEST 1 - "THE MERCY TEST"
OVERVIEW:
Player and NPC (Villain) are seperated by an unbreakable glass wall at all times. As the player pursues the Villain the Villain releases AI enemies to harrass and defeat the player. Upon reaching the final room, however, the level diverges. The villain finds himself in a dead-end, and buttons in front of the player determine his fate. His reactions are as follows (numbers represent test configures):
1. Repentant - The Villain begs for the player's mercy and surrenders.
2. Arrogant - The Villain taunts the player.
The player then has the choice between the two buttons- Kill or Save. Pressing the 'Kill' button causes the Villain to be killed by the room. A set time after the villain dies, the game ends automatically. The 'Save' button causes the Villain to thank the player for his kindness, and the game ends.
TESTING CONSIDERATIONS:
AIMS:
Players will either kill or spare our villain based on the characterisation we give them. This is a quick method of demonstrating whether or not players found our character believable, and will be backed up by the survey data they provide us.
PROS:
Players make a decision based on a judgement of an NPC they've had some manner of interaction with.
A life-and-death decision like this offers the player plenty of room for consideration.
CONS:
Would certain players be morally stronger? Thus, would they be less likely to even consider killing the villain?
Possibility of players for example sparing the Villain in the 2nd scenario, then killing him in the first simply to see the scene.
OTHER:
Playing the level more than once, the player knows what to expect- but they are then surprised by a different outcome. This test uses player choices as a barometer for testing the believability of our NPC and how engaging he is. This test also gathers certain ‘player choice’ data which can be an interesting field to consider in addition to how well our NPC worked.
MODIFICATIONS:
This test could be easily modified due to its simplistic nature. For example, a third option for the level could be to have the ‘Villain’ instead be a heroic character, who causes enemy-related problems for the player purely by accident. Would players make a snap decision regarding the value of the character’s life if they knew their troubles were simply accidental?
EXAMPLE TESTS:
POSSIBLE TEST 1 - "THE MERCY TEST" [PDF]
POSSIBLE TEST 2 - "THE FRIENDSHIP TEST" [PDF]
POSSIBLE TEST 3 -" THE SINISTER TWIN TEST" [PDF]
Showing posts with label framework. Show all posts
Showing posts with label framework. Show all posts
Thursday, February 18
Monday, February 15
Week II - Design Document Framework
The Title:
Creating interesting non-playable characters.
(Being a study into the contributing factors required to create interesting and believable characters within videogames.)
The Question:
What methods can we use to create realistic and engaging NPCs?
The Hypothesis:
If certain characteristics are considered more 'believable', then NPCs who have them will be more engaging to players.
Project Goals:
The ultimate goal of this project is to create some form of rule set for creating interesting Non Playable Characters (NPCs), and a clear explanation of the context these rules apply in. Through testing pre-defined characteristics that we theorise have positive or negative effects upon players, we should be able to create NPCs players consider 'realistic'. In order to reach this goal we will need to examine current thinking on the subject, case study games with interesting NPCs to look for common devices that we think make them interesting, and possibly distribute questionnaires/surveys to establish what type of person feels these characters are interesting, and how much they are in majority. Obviously it would be impossible to actually cover every possible device for making a character interesting and believable, so eventually it will be a requirement that we define the genre and subjects that will be tested (male 18-24).
The Test:
Our test will run a group of 'traditional videogamers' (males, 18-24) through a basic level and record their reaction to our NPC characters. We will gather qualitative and quantative data on their reactions through entry and exit questionnaires. To avoid an excessive number of personality variables, we must narrow down certain major traits. If we make characters too realistic (ie. totally multi-faceted) we will have far too many perspectives to test from. Our test framework must be such that variables can be changed without affecting the overall structure of the test. For example, making the character kind or callous- does the framework support this? By testing certain personality variables but leaving scope to modify them for repeat tests, we allow for the possibility of similar-conditions repeat testing.
Creating interesting non-playable characters.
(Being a study into the contributing factors required to create interesting and believable characters within videogames.)
The Question:
What methods can we use to create realistic and engaging NPCs?
The Hypothesis:
If certain characteristics are considered more 'believable', then NPCs who have them will be more engaging to players.
Project Goals:
The ultimate goal of this project is to create some form of rule set for creating interesting Non Playable Characters (NPCs), and a clear explanation of the context these rules apply in. Through testing pre-defined characteristics that we theorise have positive or negative effects upon players, we should be able to create NPCs players consider 'realistic'. In order to reach this goal we will need to examine current thinking on the subject, case study games with interesting NPCs to look for common devices that we think make them interesting, and possibly distribute questionnaires/surveys to establish what type of person feels these characters are interesting, and how much they are in majority. Obviously it would be impossible to actually cover every possible device for making a character interesting and believable, so eventually it will be a requirement that we define the genre and subjects that will be tested (male 18-24).
The Test:
Our test will run a group of 'traditional videogamers' (males, 18-24) through a basic level and record their reaction to our NPC characters. We will gather qualitative and quantative data on their reactions through entry and exit questionnaires. To avoid an excessive number of personality variables, we must narrow down certain major traits. If we make characters too realistic (ie. totally multi-faceted) we will have far too many perspectives to test from. Our test framework must be such that variables can be changed without affecting the overall structure of the test. For example, making the character kind or callous- does the framework support this? By testing certain personality variables but leaving scope to modify them for repeat tests, we allow for the possibility of similar-conditions repeat testing.
