defining event schma is really really annoying and wasting time, easy to make mistake, any way to work around this?
Defining an event schema for your custom events in advance of sending them does take some time but that time invested initially is well worth it and will help to prevent you from getting into a real mess further down the line.
-
The event schema you create in the event manager is used to validate the events we receive from your game. Any events that don’t match the schema you define will be rejected. This is an important step, it prevents your data set from getting corrupted with unexpected or malicious data.
-
You should always consider the design of your analytics events carefully to ensure that you are recording events and parameters that will be able to answer the questions you will want to ask in the most efficient way. An ill considered event schema can result in data that is inadequeate or difficult to work with and could impact the performance of your game in extreme cases. We have some blog posts that may help you with event schema design and we can provide pointers on best practice here on the forum.
An Introduction to Analytics for Games
Understanding Event Tracking in Games -
We provide tools to help you with your events. We support Environments, so you can define and test events in a development or staging environment before adding them to a production build. We are intending to extend the event management tools, promoting events between environments is still manual and as you say, error prone.
The Event Browser tool lets you view the most recent events and spot any that are failing validation to that you can rectify any problems and keep your dataset clean.
The time investment up-front realy is worth it, I have seen too many games where the lack of event validation against a predefined schema or poor event design has led to the analytics data becoming very difficult and in some cases almost impossible to use.
Please share any use cases or analytics scenarios that you need help with on the forums and we will be able to give you some pointers on best practice.
I hope that helps.
Hey @lauhonyeung ,
We’re currently looking into ways to make the event schema definition quicker and less error prone starting with providing an event management API.
I’d be interested in hearing your use cases and understanding how we can help make the event schema definition less time consuming for you while still providing the value that @Laurie-Unity mentioned above.
Thanks,
Clare
I know how to use this tool.
it’s just should not be designed in this way…
It’s basically anti-humanity.
Correct, we didn’t have this definition requirement with legacy Analytics. Since we had to process the event schema on the fly, the events took a minimum of 8-16 hours to appear on the dashboard. Sometimes under load, it would take up to 2 days. With UGS Analytics and defining the schema ahead of time, we can process events in minutes and they are displayed on the dashboard. Last time I checked, I saw events in 2 minutes, but never should take more than an hour. So not really “anti-humanity”, but as Clare mentioned, we are looking to improve this definition process.