There were 135 space shuttle missions over 30 years, with 2 failures resulting in 14 lives lost, giving a success rate of 98.5%. The second failure happened when the program was mature, which means that either NASA didn't analyze certain failure modes or they didn't take steps to address them. The Space Shuttle's design is inherently less safe than a normal capsule on top. With the orbiter on the side of the stack, any debris from other components is more likely to damage it. The orbiter also had no launch escape system or ability for crew to eject. Also, the solid boosters could not be throttled or shut down early if they malfunctioned. In contrast, capsules like Dragon and Soyuz are above the booster stages, reducing the chance of damage from any malfunctions in the stages, and allowing a launch escape system to get the crew away in the event of an emergency.

Falcon 9 has had 531 launches over 15 years (394 of them have happened since January of 2022), with 3 failures (one on the pad before launch, two during launch), for a success rate of 99.4%. Had these failures occurred during manned missions, the Dragon capsule's launch escape system would have likely saved the crew.

The mature version of Falcon 9 (block 5) has had 466 successful landings out of 472 attempts, giving it a success rate of 98.7%. This likely means that riding on a Falcon 9 first stage with no additional safety devices (such as a parachute or a launch escape system) is safer than riding in the Shuttle.

▲

kemotep 4 days ago | parent [-]

Thanks for providing this to other users but my point was about other rocket programs beyond Falcon 9 and Space Shuttle such as Starship. Starship is behind where Falcon 9 was at this point. By the same timelines Apollo was sending people to the moon too.

▲

ggreer 4 days ago | parent [-]

It's not useful to compare timelines. Of course the Apollo program went fast. Adjusted for inflation, NASA's lunar program cost over $300 billion. It also killed three astronauts. And it didn't have the regulatory hurdles that exist today when trying to launch rockets.

Starship's budget is 2-3% of the Apollo program, and its goal is to become profitable long term. I would assume that given a sliver of the same budget, and a much harder problem (fully reusable super heavy lift vehicle), and more regulations than the 1960s, it would take significantly longer.

It's also not useful to compare failure rates yet, because Starship is currently a test program. SpaceX believes that it's cheaper to build, test, and revise rather than to try getting it right the first time. They know Starship is not reliable, which is why they don't have real payloads in their test flights. Contrast this to the Space Shuttle, which NASA thought was so safe that they put a schoolteacher on it and broadcast the launch to children across the country.

	▲	kemotep 4 days ago \| parent [-]
		This is the 4th launch where they are effectively attempting the same thing they were going to do 8 months ago. Musk himself has a deadline of December 2026 for Mars, ignoring Artemis. How many more launches do they need to work out orbital refueling to make that deadline if they don’t test actually sending a real payload into space?