The Falcon 9 has had 126 missions so far and only 2 of those were failures. It's not a perfect record, but it's pretty decent. And presumably SpaceX has learned from those failures. Compare that to the Ariane 5 rockets (which will actually be launching the James Web Space Telescope) which has had 109 launches, and 5 of them ended in failure.
I'm just going to assume you're referring to Starship's various explosions. It's far too early in the Starship development cycle to draw any kind of conclusions about it's reliability.
Starship is a prototype. It's a completely new vehicle with completely new engines, and they're building it with the expectation that the early versions are going to blow up. It's like saying Falcon 9 is unreliable because their early "grasshopper" prototype (for testing landing) exploded and at times. Also probably a good idea to note that these explosions all happened during their landing attempt, so in theory any payload onboard would have already been deployed. It's just the vehicle that would be lost. Of course they're still so early in the development cycle that Starship hasn't even attempted an orbital flight yet.
I get that it's weird watching these very early prototypes blow up so spectacularly and publicly, but that's the development model SpaceX has chosen. And we're not used to watching rockets being built and tested so out in the open. Personally I think it's exciting watching the progress they're making.
That's how space travel development works. It doesn't take off; it crashes; it explodes; it fails to land; it works. From then on pretty much, it works.
What is that based on? How do you distinguish launch systems that just don't work well? Everything works in the end? That would make engineering much easier and less stressful!
I’m unaware of any launch system that gets many tries at failures - it either gets what needs to be done relatively quickly or it gets cancelled. The number of tries you get is usually single digits.
Falcon 9 (and Heavy) has not experienced any kind of mission failure since 2016, the latter period accounting for (if I'm counting right) 69% of its total launches including the Amos-6 pad failure (which did not actually launch). Its actual failure rate is 1.5%.
It's not a track record that deserves any more derision than its contemporaries. Doing so in such vague terms just makes you look like you don't know what you're talking about, which goes double if you're thinking not of F9, but rather of those big shiny rockets they've been blowing up in Boca Chica recently. That (Starship) is a development program.
