Bucketing New Users

When using a percentage rollout, the variation will remain the same for a given user key as long as the rollout percentages are not changed. To learn more about how this works on a technical level, you can check out the How Percentage Rollouts Work (optional) lesson at the end of this unit. For now, just know that the user key needs to change in order to receive a different variation when using a percentage rollout.

We can simulate a new user by changing our user key, using the handy user key changing function that we implemented earlier in the lesson. All we need to do is click the "Toggle Runner" header, input our new user key, and click "ok". Let's try changing the user key to dev1234 now:

If your new user key changes the variation, you will see a log in the console area informing you of the new speed value:

If you did not see a console log, it means that the new user key received the same variation. Try a different user key to see if your variation changes. Remember, we have a 1 in 3 chance of receiving the same variation.

Randomizing the initial game speed will likely have an interesting impact on our users. This means that some users playing our game might find it more difficult, while others may find it easier. If we were to measure the impact of randomized game speed on our users, it could provide us with valuable data about how to improve our game so users spend more time playing. This is where experimentation comes into play. In the next lesson, we'll turn our flag into an experiment.