Originally Posted by MG23
"It was basically designed to minimize the risk of failure, rather than to have backups if it did fail. The ascent stage engine, for example, was about as simple as you can make a rocket engine; if I remember correctly, it was pressure-fed and hypergolic, so just a couple of tanks, a couple of valves, and a rocket nozzle."
That is correct, and I would not describe the LM as lacking contingencies. The design philosophy was extreme reliability through simplicity -- even at a significant cost of performance.
The LM used batteries instead of fuel cells to improve reliability, the engines had no turbopumps, no ignitors, no engine gimbal in the ascent stage, no regenerative cooling -- it was as simple as could be made. The propellant valves were double redundant and the plumbing had multiple contingency paths.
The LM was the first digital fly-by-wire vehicle, and like modern aircraft, flight control was critical. The LM had dual-redundant fly-by-wire guidance & navigation computers hand-built and tested with extreme rigor, and the backup computer had independent software developed by a separate company to avoid a generic failure.
If both LM flight control computers failed, there was a manual reversion mode including a non-computerized analog pathway from the hand controller to separate redundant solenoid propellent actuators on each RCS thruster. Activating this required only a full deflection "hard over" on the hand controller to trigger the alternate path.
There was a practiced contingency procedure to ascend from the lunar surface and achieve orbit using no computers and no inertial platform whatsoever. It involved using charts, a stopwatch and making stepped pitch changes to align window etchings with the lunar horizon. Astronaut Gene Cernan said he achieved this in the simulator and felt it was possible.