Introduction how the goedle.io plugin for Unity can be implemented. The basic setup just takes about 2 minutes.
To get an APP and API key please reach out to us support@goedle.io.
In order to install the Unity SDK for tracking telemetric data from Unity, the SDK has to be downloaded. https://github.com/goedleIO/sdk_unity/blob/master/goedle_io_sdk.unitypackage
The SDK comes as UnityPackage, therefore it can directly be imported into a Unity project as Asset. To add the SDK as Asset, it has to be imported as custom packages.
Afterward selecting the downloaded package, the Scripts folder has to be checked. The Plugin and Materials folder is automatically imported.
Once the package is available as Asset, a loader object and two manager objects have to be added. These objects are necessary so that the SDK is available over the whole runtime of the Unity application.
Adding of the objects can be simply done via drag and drop. Depending at which position in the lifecycle of the Unity application the SDK should be available, the objects have to be dragged to the corresponding scene. For instance, if the SDK should be accessible in the start scene, the loader and the manager objects have to be dragged in this scene. The objects are pre configured, so one has only to add credentials to the GoedleManager object to authorize at the GIO backend.
The loader object and manager objects are shown, these object can be found in the goedle_io asset, under goedle_io -> Scripts
.
The following object have to be dragged to a scene:
- GoedleLoader - The loader initializes the GoedleManager, this object is has to be available to track data. Furthermore, the GoedleHttpClient is initialized to send data to the GIO backend.
- GoedleManager - This object holds all tracking functions and the authorization interface. As well as a check box to run the SDK in staging mode or to disable the tracking. Additionally one can add a Universal Analytics ID to send data to Google Analytics. The GoedleManager loads the GoedleAnalytics script which contains a wrapper and the interface to initialize and programmatically control and configure the GIO SDK.
- GoedleHttpClient - The GoedleHttpClient client, is called after executing a tracking function to send data to the GIO backend. The GoedleHttpClient client also sends data via coroutines to reduce the load on the main thread.
Once the objects are added to the scene. The credentials for the GIO backend have to be added. These are namely the APP key and the API key. As well as additional meta information to further describe the Unity application and the Universal Analytics configuration. Figure 1.10 shows the interface which can be accessed by selecting the GoedleManger.
The configuration options for the SDK:
Type | Default Value | Importance | Description |
---|---|---|---|
STAGING | unselected | optional | This checkbox is on default unselected, it can be used to switch between the live API of the GIO backend and and an text based output of the tracking. If the box is selected, one can see the tracking events in the console and they are not sent to the GIO backend. |
APP KEY | empty | mandatory | This key is needed to authorize at the GIO backend. |
API KEY | empty | mandatory | This key is needed to authorize at the GIO backend. |
APP VERSION | empty | optional | The app version is used to distinguish between different version of an Unity application. |
APP NAME | empty | optional | This app name should be a human readable string to identify the Unity application. |
GA Tracking ID | empty | optional | The universal Analytics Tracking ID. The id has to be taken from a Google Analytics property. The property has to be created in advance. |
GA CD EVENT | empty | optional | Google Analytics allows to make use of custom dimensions, which help to further segment tracked data. If one once to use a custom dimension, the name has to be set. This allows the SDK to handle a tracking call with the used event name as custom dimension. The GIO SDK supports up to two custom dimensions. |
GA CD 1 | empty | optional | The first custom dimension, which can be later accessed through Google Analytics. |
GA CD 2 | empty | optional | The second custom dimension, which can be later accessed through Google Analytics. |
TRACKING DISABLE | unselected | optional | This checkbox allows to disable the tracking, this can be used for instance to opt out a user. |
Beside the graphical interface, all fields of the interface from can be added programmatically.
After the (mandatory) fields are set in the interface or programmatically, the SDK is ready for usage.
To track data from a Unity application with the goedle.io SDK, the tracking points have to be set in the source code of the Unity application. Therefore, a certain tracking notation is recommended. Besides the event tracking functionality, the SDK holds additional functions, to respect the privacy and to identify a user.
There are two main variables to track an event. The first is the action, e.g., view, share, like these variables are the specific actions which are done by a user. The second variable specifies the object of the action e.g., task, question, scene. These two variables are concatenated and delimited with a ".". Note however that a specifier is not mandatory and we also support hierarchies with more than two levels. The tracking origins in the Object-Notation framework, which also used by Segment or Google Analytics.
Furthermore, a event can be specified with an event id. As an example, if one wants to track the view-event of a scene, the event would be view, the specifier would be the scene and the event id the scene id X235123
If one has to track the information of the duration of a certain task or the result of a answered question or if the event has a certain value, an additional parameter can be passed. The event_value. The event name would be finished.task, the event id could be task-12 and the event value 69. Where 69 would be the duration in seconds.
For certain behavioral analysis passive game changes have to be tracked. These can be seen as a stream of data which represent the value state of the Unity application. A passive game change is the adjusting of values which are done by the application automatically, depending on the game design. To identify a game state, the event should be called “game.state”, the event id is in this case, is the passively changed variable. And the event value, the value of the variable.
To call the track function, the goedle.io SDK has to be integrated in the header of a Unity/ C# class file. This can be achieved through:
using goedle_sdk;
There are three different tracking signatures which cover the in Section 4.1 described tracking notations, while the game.state
has the same signature as the event value tracking call.
Tracking a single event
GoedleAnalytics.track ("<event_name>");
Example
GoedleAnalytics.track ("start.game");
Tracking a single event with an identifier
GoedleAnalytics.track ("<event_name>", "<event_id>");
Example
GoedleAnalytics.track("reached.level", "level-2");
Tracking a single event with an identifier and a value
GoedleAnalytics.track("<event_name>","<event_id>","<event_value>");
Example
GoedleAnalytics.track("finish.level", "level-12", "65");
Tracking the game state
To track a continuous stream of events, the code of the Unity application needs additional preparation. This is the polling interval, which can be implemented with a counter (nextActionTime), that is increased after every action and the duration of one interval (period).
An example configuration with a 15 seconds interval would look like this:
// Game state config
private float nextActionTime = 0.0f; // 15.0f means a period of 15 seconds, this is complete sufficient to track a game state.
public float period = 15.0f;
Afterwards the tracking has to be periodically called. This can be achieved by using the update function of a Unity class.
void Update () {
// Your code ...
if (Time.time > nextActionTime ) {
nextActionTime += period;
GoedleAnalytics.track("game.state","health", "100");
GoedleAnalytics.track("game.state","mana","400");
}
}
If a one wants to track information about a class, school or a team, the SDK supports so called groups. A group consists of a type, for instance class
and a group name 1A
. Here is the complete call:
GoedleAnalytics.group ("<group_type>", "<group_name>");
Example
GoedleAnalytics.group ("school", "EA");
To have the opportunity to further personalize the learning experience, additional information about a user can be helpful. Currently the goedle.io backend supports three different kinds of personal user information. Namely the last- / first name and the email address. It is important that these kind of information are only transmitted when user explicitly gave the permission to track these kind of information. The SDK only offers this function to track the information, but the developer should responsible use the functionality. It is important that correct naming convention is used, otherwise the information can not be stored and processed.
first_name
GoedleAnalytics.trackTraits("first_name","<play_first_name>");
Example
GoedleAnalytics.trackTraits("first_name","Hans");
last_name
GoedleAnalytics.trackTraits("last_name","<player_last_name>");
Example
GoedleAnalytics.trackTraits("last_name", "Smith");
GoedleAnalytics.trackTraits("email", "<email_address>");
Example
GoedleAnalytics.trackTraits("email", "hans@gmail.com");
goedle.io is GDPR ready and respects the privacy. Therefore, the SDK hold functions to control the data which is transmitted to the goedle.io infrastructure.
To respect the privacy, it is necessary to add an opt-out function. This can be done via the interface or via a function call.
GoedleAnalytics.disableTracking ();
This function disables the goedle.io SDK, and even if tracking points are set in the code, nothing is transmitted to the goedle.io backend or to Google Analytics.
If a user id is not available the goedle.io SDK creates automatically in the background a user id. But, the automatically created user ids are only session based. This is done, because of privacy reasons. Otherwise, the developer has to set explicit the user id in the code. For instance, if the user id is created on personal user information. To set a user id the function setUserId() can be called:
GoedleAnalytics.setUserId ("<unique_user_id>");