How we use machine learning to constantly improve Neyon Clash

Updated: Jun 26, 2020

Neyon Clash is improving itself by constantly running machine learning algorithms.

Neyon Clash Machine Learning

Constantly testing.

As a location based game Neyon Clash is very dependant on item locations. To avoid spawning items at inaccessible area the app is persistently testing item locations.

Self learning.

Whenever an item has not been picked up the quality of the POI (point of interest) drops. The POI is therefore being tested less regularly but still tested until it is very unlikely to use it for an item spawn again.

The positive tested POIs are preferred for item spawns. With that algorithm Neyon Clash is steadily improving the quality of games played.

Reaktor Berlin's focus is on Real-World-Multiplayer-Action-Games. Games that encourage people to get active and meet each other in real life.


Bautzener Straße 35
10829 Berlin

© 2019 by Reaktor-Berlin GmbH