Learn the definitions of key terms used in Game Overrides.
Read time 4 minutesLast updated 8 months ago
A/B testing
A/B testing is a feature within Game Overrides that splits your Override into variants you set to see what impact they make on your game. For example, you might believe starting players are given too many coins and are progressing through your game too quickly or stop playing too early, missing IAP (in-app purchases) opportunities. You’d set two (or more) starting balances: the control, and the variant.
Analytics
An end-to-end data and analysis solution designed to support your entire studio. Analytics lets studios easily understand game performance and player behaviors. Use Analytics with Game Overrides to report on the impact of your changes, and target players with Audiences.
ARPDAU
Average revenue per daily active user.
Audience
Players that are grouped by criteria such as player behavior or location. Audiences can be targeted to personalize their game journey.
CCD
By using Unity's Cloud Content Delivery to write stateless server-side code on a fully managed infrastructure, you can focus more on developing your game logic. Cloud Code automatically provisions server capacity based on load so you can ensure that your players receive a good experience without any lag or downtime.
Content
Content in-game, such as lives, weapons, and power ups.
Control variant
A control variant is the original version of the game which new versions (with changes) are tested against. For example, if you try out a new level design, the control variant is the existing level players are used to.
Currency
A currency in Economy defines virtual money that exists within your game. When defining currencies, there is the option to set an initial balance (the amount a player receives when you initialize the currency), and a maximum currency balance (the limit to how much of that currency a player can have). For each defined currency, Economy stores a balance against each player account, which the game manages.
Daily play time per DAU
Average daily play time per daily active user. For example, a daily playtime per DAU of 3 minutes means your active players, on average, play 3 minutes per day.
Dashboard
Up-to-date bar and line charts that show you how well your Overrides perform at a glance.
Economy
Unity’s Economy service provides a way to create, manage and publish an economy system to be used in your game. Includes Currency, Inventory Item, Configuration, and Publication.
Group
Player groups fall into predetermined categories based on criteria. Also known as variants.
IAP
In-app purchases are items in-game that are bought with real money, such as gems, lives, or weapons.
Inventory item
A virtual item the player has in their inventory, for example, a weapon.
JEXL
JEXL (Java EXpression Language) conditions are evaluated against a variety of factors in each request. This method uses contextual data attributes to define the audience for which you want a Rule to apply.
Minimum detectable effect
Within a mobile game experimentation platform, the Minimum Detectable Effect is the smallest difference or change (in metrics such as player engagement, time spent in the game, etc.) that the experiment is designed to spot between the control and the variant groups. For instance, if you test a new level difficulty, the MDE would be the smallest increase in player completion rates significant enough to implement the change for all users.
Override
Game Overrides give Unity developers the ability to create personalized in-game player experiences and understand their impact.
P-value
A p-value determines if a change in the game (like a new character design) did improve player engagement or if the observed improvement was due to chance. If you see more players using the new character and the p-value is low, it suggests that players likely prefer the new design.
Peeking
Peeking checks the results of a game test before it's fully completed, risking hasty decisions.
Power analysis
Power analysis determines how many players you need to test a new feature on to decide if it's beneficial or not. Before introducing a new in-game purchase, for instance, power analysis can help decide how many players should be exposed to this feature to gauge its success.
Real Money Purchase
A purchase made by the player with real currency.
Remote Config
Remote Config is a cloud service that you can use to tune your game design without deploying new versions of your application. It consists of a set of namespaced Key-Value parameters, and you can optionally define a set of values that override or add to these parameters.
Reporting
If you’re using Unity Analytics, you can see the impact your changes have on your KPIs. On the Details page, select Reporting. You can select a KPI and the dates (today, last seven days, last 14 days, last 30 days, last quarter) to view your data.
Return rate
Rate of players that come back to your game.
Sample ratio mismatch
When different versions of the game are exposed to unintended unequal numbers of players. For example, if you want 500 players each to try out two different game backgrounds, but due to a glitch, 700 players see one background and only 300 see the other.
Samples
Samples are a subset of players selected to test new game features or changes. Instead of rolling out a new enemy character to all players, the game company might first introduce it to a sample of 1,000 players to gather feedback. The more samples in an experiment, the less likely it is for results to be due to chance.
Scheduling
Choose when you want to run your Override. You can start immediately and run indefinitely.
Seasonal event
A seasonal event provides themed content for your players. For example, you may run a winter campaign with items related to snow, Christmas, and presents.
Simulation Experience
Use Simulation Experience to understand what a given Audience or player’s experience will be with your configured Game Overrides. In addition, you can add user IDs and their attributes, view the keys that would be overridden through Game Overrides, and look into any relevant conflicts.
Statistical significance
Statistical significance tells us if observed changes, like increased game time or in-app purchases, are because of the new feature or by chance. If a new game tutorial leads to statistically significant increases in level completions, it means the tutorial likely had a positive effect.
Target
The group of players you want to reach.
Type
The type of key used in key-value pairing: int, string, float, JSON, or long.