The booster's flight uses the rocket equation and a 2-DOF ascent / coast / landing model (gravity, thrust, mass depletion). Each successful landing raises the vehicle's reuse count N, which feeds a real amortized cost-per-seat formula — the same logic that lets reusable boosters cut real launch prices from tens of millions of dollars toward far lower marginal costs per flight:
Cost/flight = Vehicle_cost / (N+1) + Refurb_cost + Fixed_ops
Price/seat = Cost/flight / Seats
Vehicle_cost is fixed at $62,000,000 (a realistic medium-lift reusable booster build cost). As N grows, the amortized share of that build cost falls hyperbolically, so price/seat asymptotically approaches Refurb_cost/Seats + Fixed_ops/Seats — the floor real reusable-launch economics is chasing. The bottom chart plots this exact curve against N, live.
- Launch — runs one ascent burn → unpowered coast → a single suicide-burn landing ignition. A soft touchdown (below 4 m/s) counts as a successful reflight and increments N; a hard landing destroys the booster and resets N to 0.
- Landing guidance gain — sets how much stopping-distance margin the autopilot reserves before it ignites the landing burn. Below about 1.0 it waits too long: even full throttle can no longer supply the deceleration needed in the remaining altitude, and the booster slams down fast. At or above ~1.0 it ignites with a safe margin and the closed-form burn (computed once at ignition, then held) brings it to a gentle touchdown — try dragging the slider down mid-descent to watch a landing fail in real time.
- Seats / Refurbishment / Fixed ops sliders — explore how flight cadence and per-flight cost assumptions change the ticket price a real space-tourism operator would need to charge to break even.
- Flight view camera — drag to pan, scroll/pinch to zoom; the camera auto-follows the booster otherwise. "Recenter view" clears any manual pan/zoom.
This is the mechanism the article "Space Tourism: An Overview" points to when it credits reusable rocket technology with dramatically reducing launch costs — reuse count, not raw thrust, is what ultimately makes a lower-priced seat possible.
Fixed vs. the original 3D build: the source engine's ascent thrust gave a thrust-to-weight ratio below 1 (it could never leave the pad), and its landing-burn guidance never added gravity into the commanded acceleration (so a "needed deceleration" of zero — near the top of the arc — silently authorized a hover, and away from that point the commanded burn always fell one full g short). Both were confirmed numerically (see repo scratch scripts) and corrected here: ascent thrust was retuned to a liftoff T/W of ≈1.3, and the landing guidance is a proper single-ignition suicide burn sized as g + requiredDeceleration.