There's a real debate in game design about whether reward-driven retention creates genuine engagement or just defers churn.
The skeptic's version: you've trained your players to expect rewards. When the rewards stop or decrease, they leave faster than players who were never in a reward program, because their baseline expectation shifted upward. The reward didn't create loyalty, it created dependency.
The optimist's version: retention is retention. A player who stayed because of $PIXEL rewards had 30 more days in your game, made more purchases, invited more friends, contributed more to the community. The chain of events that started with a reward created real value regardless of what motivated the initial stay.
Stacked's AI game economist is designed to resolve this empirically: measure actual LTV trajectories of rewarded vs. non-rewarded cohorts over long windows. If rewarded players show flat or declining LTV over 6-12 months, you have a dependency problem. If they show increasing LTV, the rewards are doing genuine ecosystem-building work.
The 200M rewards and $25M revenue figure suggests the optimist's version has at least some support in the Pixels data. But this is an 18-24 month question to answer properly, and the data has presumably been accumulating for less than two years.
I find Stacked genuinely compelling for studios that want to test this empirically rather than assume either answer. Whether the long-term cohort data supports the optimist's version is still a question worth watching.