Actually, that’s completely different. If there is going exploration soon (and until recently that was the plan), then it will need massive amounts of propellant, not foodstuffs. It will cost a lot of money to launch that propellant. That money can be used to develop cheap lift that is available to everybody or heavy (and expensive) lift that is available only to NASA.

What I’d also like to do is re=purpose our spending on space on missions that aim to explode such markets as quickly as possible.

That’s precisely what an exploration program using freely competing commercial propellant launchers would do.

There’s nothing artificial about launching propellant in support of exploration, in fact for the foreseeable future there is no way around that.

Then there’s nothing artificial about farm subsidies.

If you don’t like NEO missions, or unmanned missions, then there are alternatives.

It has nothing to do with whether or not I like a particular mission, but whether or not taxpayer money should be spent on missions that do away with even a pretense of contributing to the economic bottom line. Once again, we probably agree that government has a role to play until such time as markets emerge that can sustain the launch services industry without taxpayer intervention. What I’d also like to do is re=purpose our spending on space on missions that aim to explode such markets as quickly as possible.

Driving up launch rates does not require artificially propping it up on billions in taxpayer dollars sunk into projects with no evident prospect of return near, mid, or far just for the sake of lowering transportation costs. This approach wasn’t “market-driven” when we called the Tennessee Valley Authority and it certainly isn’t dressed up as you’ve proposed. That’s not to say that government has no place intervening in an emerging market. She certainly does; for half a century starting 150 years ago, Americans cottoned to railroad and homesteading legislation, fort building, and eventually a powerful navy to expand agriculture and industry into the old frontier and foreign markets.

Beancounter did it quite well. To simply tot up the failures and successes as though they are equivalent is meaningless and misleading, because it implies a Bayesian probability of a 3/7 chance of failure on the next flight, which is clearly nonsense.

It was certainly not my intent to extract a measure of reliability from 7 launches, which is why I said: “[i]t’s misleading to focus on launch rates for any young family of launchers in the first place. That’s the point.”

Nonsense. SpaceX has flown nobody and it is a bogus argument to propose future performance from them will meet expectations when they have not orbited a soul while government space agencies have been orbited crewed spacecraft for half a century. It’s just a bogus. But if you have such high confidence in SpaceX one would expect the firm to be flooded with investments from Aussie sources.

Not at all, if we did that we might have to wait a very long time.

Awesome. So what should NASA being doing with the $20 billion a year we’re willing to shell out?

Explore. Create as much demand as possible for commercial launch services. This will start the probably lengthy process of driving down launch prices by enough so that we’ll see commercial development of space before everybody here is dead and buried. All else being equal, spend as much as possible on launching propellant and as little as possible on space hardware and directly funded R&D. Markets, not bureaucracies, will determine how much of that should be channelled into R&D. My current near term favourite would be a set of NEO tagging / sample return missions using an unmanned reusable spacecraft based at L1/L2, not LEO. Relatively cheap and low-tech and hard to screw up. This would create a large demand for propellant at L1/L2 and in LEO (good for RLVs) as well transport services beween LEO and L1/L2 (good for SEP tugs and aerobraking R&D). A market driven solution.

Beancounter did it quite well. To simply tot up the failures and successes as though they are equivalent is meaningless and misleading, because it implies a Bayesian probability of a 3/7 chance of failure on the next flight, which is clearly nonsense. The order in which things fail and succeed is very important. At this point, Falcon 9 has a perfect flight record, and it’s reasonable to estimate that its reliability probably exceeds ninety percent. If they have another ten consecutive successes, we could assess it in the high nineties.

“How would you characterize it [SpaceX launch record]?”

I would say their launch record is 1-0 from an operational standpoint. They have had only one paid payload mission since Falcon 1 became operational (Malaysia’s RazakSAT), so their customer record is perfect. Falcon 9 is still in test, so I don’t count it yet.

If you want to count their test flights, then that’s really open to interpretation, especially since this is a new company. I think what is the most relevant thing to look at is how they have progressed through their test programs, and that they did not repeat problems – each test accomplished more and more of the mission profiles, until they finally solved the problems and orbit was achieved.

With all things, I gauge companies based on what they do when things go wrong, not only on when they go right. And so far SpaceX seems to be doing well.