Hacker News Reader

>SpaceX's 93% launch success rate is still below average for ALL launches (94-95%), and has a higher incident rate than its "competitors" in any form. That's what was asked. You don't need to apologize for SpaceX.

How do you get to this number? Even in the least charitable way to calculate it, excluding Falcon Heavy, using Amos in the numerator but not the denominator, and counting the secondary payload loss on CRS-1, Falcon 9's failure rate is 3/57. That's 5.2%, or a success rate of 94.74%.

>Some of the rockets in your "thorough" have radically different missions, specifications, and costs than the type of rockets we're discussing. Let's talk about what we're actually comparing. Orbital launch rockets actually beat "industry" averages by a signification margin. And let's talk about failures, unrecoverable ones.

Zero of those rockets are non-orbital. The only one not in Falcon 9's class is Pegasus (though the TESS mission was intended for Minotaur). Ariane and Proton are Falcon's main competitors in the commercial market, where ULA is uncompetitive.

>See a pattern?

Yes, designs tend to get safer as they mature. Falcon is at 58 launches and is likely to exceed 80 before any crewed launch. Its last failure was mission #29.

>All of the above families have mission histories of 100+ launches.

Counting 'family' legacy back to ICBMs is ludicrous. Atlas V has flown 78 times, Delta IV 36 times (Medium and Heavy), H-IIA 39 times.

And can you cite those family success rates? The Thor/Delta family, by a quick calculation per Wikipedia's launch lists[0], has an 87.42% success rate.[1]

>No they didn't. Are you just purposefully misreading the article?

No, but the article's author is purposefully misleading readers by citing that number alone:

>Against that backdrop, two recent government reports raise questions bearing upon the reliability of SpaceX products and processes. The first is an evaluation of quality controls among launch-vehicle suppliers to the military space program. That report, prepared by the defense department's Inspector General and dated December 20, found 181 deviations from quality standards at contractor sites.

Moreover, ULA rockets consist of both ULA and Aerojet Rocketdyne parts (and indeed RL10 caused Atlas' partial failure). If we're correlating nonconformities with risk, their numbers should be added together. And, of course, the inspector general didn't evaluate RD-180 production facilities.

[0] https://en.wikipedia.org/wiki/List_of_Thor_and_Delta_launche...

[1] https://i.imgur.com/izkMH14.png

>Counting 'family' legacy back to ICBMs is ludicrous.

I didn't do that and you know it. Those are all rockets used in orbital payload missions. All of them have been used within the last 5 years.

I used some inline JS for this table here:

https://en.wikipedia.org/wiki/Comparison_of_orbital_launch_s...

That table doesn't show success/failures...

You can click on each rocket and look at the launch history. I am again, done with this conversation because of how disingenuous you are willing to be.

> let's talk about failures, unrecoverable ones.

Then why is spaceX at 93%? 2 unrecoverable failures in 59 campaigns equals 96.6% success.

All the competitors you listed have roots in government space programs going back many decades. Rockets tend to have a lot of "infant mortality" that those programs are way past.

Also, some of your claims are value/wrong: the Delta and Long March families are very large and very old and if you count the earliest attempts of governments to build ICBMs you get reliability figures worse than SpaceX.

Conversely the H-IIA, Delta 4, Atlas 5, Ariane 5 have fewer than 100 launches each.

Good data can be found here: http://www.spacelaunchreport.com/logsum.html

  (ATLAS ORBITAL     325(46)     .86    .86)

  (ALL ATLAS         582(116)    .80    .80)

Ouch.

>the Delta and Long March families are very large and very old and if you count the earliest attempts of governments to build ICBMs you get reliability figures worse than SpaceX.

"If you do this thing that you didn't do, you get worse figures."

Um, okay.

But you did write (entire family) for Delta and Long March. Your data is confusing and wrong.

...for orbital launches, which is quite literally what 99% of these rockets have been used for. Which also is exactly what we are talking about.

It's only confusing if you're wanting to be disingenuous.

>Altas V (actually built by Lockheed Martin, not ULA)

Er, what?

(Missed this before.)

You could maybe, I dunno, click on the wikipedia link and read it.

Oh?

>It was formerly operated by Lockheed Martin and is now operated by United Launch Alliance (ULA), a joint venture with Boeing.

https://twitter.com/torybruno/status/580076118339756034

https://twitter.com/torybruno/status/887340287550345216

Yes, it wasn't built by ULA.

>The Atlas V was developed by Lockheed Martin Commercial Launch Services

Honestly just stop.

Do you mean originally? ULA is a 50/50 joint venture that Boeing and Lockheed Martin were forced into after almost corporate-espionaging themselves into oblivion. It owns, operates, and manufactures the factories and rockets that each company used to provide independently — Delta from Boeing and Atlas from Lockheed.