The thing I keep coming back to about Stacked is how specific the problem it solves actually is.
Stacked is a LiveOps engine built by the Pixels team — it tracks player behavior and intervenes before disengagement becomes a decision. The signal it reads is RORS: reward output relative to activity. When a player's farming output starts dropping relative to time invested, Stacked catches that window before the player consciously registers it. That's not a feature you design from theory. That's a feature you design after watching the window close too many times.
Which is why 2023 matters more than the official story suggests.
Late 2023, Pixels migrated from Polygon to Ronin — a blockchain network built for gaming. Better wallet infrastructure, smoother onboarding. All reasonable, all true. But I kept coming back to the timing. The migration landed right after Axie Infinity collapsed and Ronin went quiet. Almost no active games left on the chain. I used to read this as an infrastructure call. It took me a while to see it as a market position. Finite attention divided by near-zero competition means each game captures nearly all of it.
What I missed for a long time: retention without competition doesn't generate learning pressure. Pixels couldn't learn why players leave when players weren't leaving. The signal looked like product-market fit. It was a monopoly artifact.
Then Pixels helped build the Ronin ecosystem — and created the competition that made retention hard again. What replaced the default was four years of granular data: exactly when in the crop-and-harvest cycle players stopped refilling energy, what their reward output looked like the week before they never came back. That's the pattern Stacked was built to recognize before it completes.
What Pixels chose in 2023 was a market with almost no competition. Stacked looks like proof that market no longer exists — and that they knew it wouldn't.