Game Developer

September 16, 2016 - Game
Game Developer

We must pick specific streams of information to evaluate! A variety of metrics are in our disposal, but we can’t track everything. Each stream of information we monitor includes a processing cost. Both around the server and also the human side. Although we are able to collect and treat lots of data today, we can’t get everything. We must determine what metrics are highly relevant to us, and which of them aren’t.

What metrics don’t let track?

The character and quantity of metrics you need to track completely changes from project to project. You have to plan your utilization of game developer analytics throughout the game’s pre-production phase. Based on your game’s genre and scope, you’ll find pretty much relevant metrics. For example, community and monetization related metrics don’t matter in single player games.

Game Developer

Game Developer

Beyond a couple of generic metrics, each project has different needs.

Small teams do not need to trace considerable amounts of information though. It normally won’t have to monitor all user actions and input like big companies do. A couple of simple metrics will help you solve critical game play issues. Here are a few helpful general picks to balance your game developer progression. You’d like to learn:

When gamers leave the sport

Their average session duration

What lengths gamers go hanging around

At what time gamers uninstall your application

The couple of streams of information uncovered above are extremely simple to track. By mixing individuals streams of information, you can aquire a feeling of what content frustrates your customers. Mixing this data with user census, you may also refine your analysis according to your audience. E.g. you are able to remove ten years old customers uninstalling your hardcore game for grown-ups.Game play related analytics go very deep. The simplest game uses a lot of variables. Objects possess a position, states, parameters, etc. We can’t track all of them individually: examining the outcomes would cost considerable time. Thus, we want clever methods to group teams of data together.