If the data being collected never needs to be updated (e.g logs) than saving the information to
BigQuery datasets/tables is ideal (it now actually allows for
updating data). BigQuery is optimize for performing fast queries on massive amounts of data for reporting purposes. You can then use
Google Data Studio with your BigQuery datasets/tables to create your graphs.
2. Second step is to
choose your backend which will be the proxy between your clients and your cloud database.
You can of course have your PC game directly interact with your cloud database (e.g
BigQuery) via the
Client Libraries (or via the
HTTP REST APIs which you can play with in the
API Explorer), but giving your clients access to directly manipulate your data is not secure. It is therefore recommended to restrict this access to your own 'backend' (e.g
App Engine). Your PC game would then make HTTP requests to your own custom backend APIs, and your backend will act like a proxy to save and return data to/from your cloud database via the Client Libraries or HTTP REST calls.
3. Once you have a backend deployed running the code to read/write to your cloud database, the third step is to expose your backend's abilities to your clients via your own API using
Cloud Endpoints.