Skip to content
Casheeno
All posts

Tuning a reward economy with real player data

Launch numbers are a guess. What matters is being able to change them without a release, and knowing which number to move.

By

Tuning a reward economy with real player data

You cannot design a reward economy correctly on paper. Player behaviour is the only real input, which means the important decision is not the launch values - it is how fast you can change them.

Everything is server configuration

Drop rates, prize tables, streak bonuses and payout thresholds live in config the server reads. Nothing in the binary. A tuning change is a config push, not a store review.

Watch the distribution, not the average

An average reward tells you nothing. What matters is the shape: the share of players who never win, and the tail who win far more than intended.

Cost per engaged player is the number

  • What a session costs you in rewards
  • What it returns in ads or purchases
  • How both change as a player gets deeper

Change one thing at a time

Move two levers at once and you learn nothing. We change one value, wait for a full weekly cycle, and only then decide.