Simulators and Obbies Need Different Retention Clocks
The Benchmark That's Breaking Good Games
Here's a mistake I see Roblox developers make constantly: they build an obby, check their Day 1 retention against the benchmarks they've absorbed from simulator dev culture — somewhere around 30–40% — watch it fall short, and conclude the game isn't working. Then they rebuild, retune, or abandon it. What's actually happening here is they're applying the wrong clock entirely. Simulator games and obby games operate on fundamentally different retention timescales, and conflating the two causes developers to kill games that are, by the correct measure, doing exactly what they should.
What the Retention Curve Actually Measures
Retention is a proxy for something deeper: whether the core loop is delivering repeated value. The trouble is that "repeated value" looks completely different depending on the game structure. In a simulator — think Pet Simulator or Blade Ball (Roblox) — the loop is continuous accumulation. Players collect, upgrade, prestige, collect again. The psychological hook is variable-ratio reinforcement, the same engine that makes slot machines and Candy Crush stickier than they have any right to be. If a simulator doesn't lock that loop in during the first session, the player has no reason to come back. There's no external pull. Day 1 retention in simulators is existential because the entire value proposition depends on continuity of progress.
Obbies don't work that way. An obby like Tower of Hell offers something closer to what Jesper Juul calls the art of failure — the pleasure of a clearly defined challenge with a legible skill gap. Players who bounce off an obby in the first session frequently return days or weeks later, because the barrier they hit is memorable. The game planted a flag in their brain: I didn't finish that. That's not churn. That's a pending return visit.
Two Different Psychological Contracts
What's actually happening here is that simulators and obbies make different promises to the player, and those promises determine which retention window matters.
A simulator's promise is progress. The player's investment is time and consistency. Break that chain (Roblox) early — confuse the new player, gate progress too hard, fail to show a reward within the first few minutes — and the promise is broken before it's made. This is why simulator developers obsess over the first three minutes of the experience. It's not arbitrary. It's structurally correct.
An obby's promise is mastery. The player's investment is skill, not time. And here's the design insight that most obby metrics conversations miss: mastery-based games have a documented pattern of delayed engagement. Players sample, fail, leave, stew, return. Among Us saw the same dynamic at scale — millions of players who tried it in 2020, moved on, and came back when a friend invited them again six months later. The game hadn't changed. The player had ripened. Obbies live in that same rhythm.
If you're measuring an obby against Day 1 D1 retention targets built for simulators, you are measuring the wrong thing. You want to look at reacquisition rate — how often players who left in the first session come back within 30 days — and at session depth on return visits. Those numbers will tell you whether the difficulty curve is doing its job.
The Numbers Developers Are Actually Seeing
To be precise about this: Roblox hasn't published genre-segmented retention benchmarks publicly, so we're working from community-aggregated data and developer reports on the Roblox DevForum. The rough consensus among simulator developers is that a D1 retention rate below 25–30% is a red flag. For obbies, experienced developers in the same threads regularly report healthy games running D1 between 15–22% — numbers that would cause a simulator team to panic — with strong 30-day return rates compensating. The aggregate engagement looks fine. The snapshot looks alarming. Developers who know only the snapshot make bad calls.
This matters more now that Roblox's discovery and monetization systems increasingly weight session metrics. A game that looks underperforming on a 7-day dashboard might be accumulating a loyal returning base that the dashboard isn't surfacing clearly. Genre context isn't just a design question — it's an analytics question.
What to Do With This
If you're building or maintaining a simulator, the conventional wisdom holds: the first session is the product. Optimize that window obsessively. Show progress. Reward early. Gate nothing that matters until the hook is set.
If you're building an obby, stop benchmarking against simulator D1 targets immediately. Instead:
- Track reacquisition rate at Day 7 and Day 30. If players who left session one are coming back, your difficulty hook is working. If they're not, the problem is memorability, not initial retention.
- Look at session depth on return visits. An obby player who returns and plays 40% further than their first attempt is a player who is being served by your design. That's the metric that tells you the loop is sound.
- Don't smooth out the difficulty spike that's causing first-session drop. That spike might be exactly what's creating the planted-flag memory that drives return. Test carefully before you flatten it.
- Compare your numbers to other obbies, not the platform average. Genre-matched benchmarks are the only benchmarks that mean anything.
The broader principle is one that applies across game design: a metric is only meaningful relative to the psychological contract your game has made with the player. Raph Koster's argument in A Theory of Fun is that players stay as long as there's learning left to do. Obbies front-load the learning signal and let it pull players back. Simulators have to deliver that signal in the first sitting. Same theory, different clocks.
If you're not sure which clock your game is actually running on, that's worth figuring out before you read another retention graph. Use RoWatcher to track how your metrics shift after design changes — the before-and-after comparison is where the real signal lives, and you want to be measuring the right things before you start moving them.