A single-string mission needs every one of its subsystems to work — power, thermal, comms, propulsion, avionics, attitude control and payload are all on the same critical path. Overall mission reliability is the product of every subsystem's individual reliability, so even subsystems that each look "pretty good" multiply down to a surprisingly low mission-level number.
Giving a subsystem a redundant backup means the mission survives if either unit works, so its effective reliability becomes 1 − (1 − r)² instead of just r. A subsystem at 85% jumps to about 97.75% with one backup — a huge gain. A subsystem already at 97% only gains about a point.
With a limited budget of redundant units, the efficient strategy is to back up the lowest-reliability, most fragile single points of failure first — that's where 1 − (1 − r)² buys the most extra reliability per unit spent. Backing up an already-reliable subsystem instead wastes the budget.
- Click a node (in the 3D view or the list) to toggle a backup unit, budget permitting.
- Overall mission reliability updates instantly as the product of every subsystem's effective reliability.
- Reset all backups returns every unit to the budget and the tree to single-string.