Comment by adventured | Hacker News Reader

Comment by adventured | Hacker News Reader

Parent Full thread

NASA, for example, spends immense amounts of money on trying to ensure that their projects have a high success rate.

They'd rather spend $2b to get a 97% likelihood of a positive outcome, than $1b with an 80% likelihood.

Time is a big deal when you're going to spend a decade on a project. NASA will have a ~$250b budget for the next decade. The people that they have access to to work on projects, are finite however. Their lifespans are finite, how much time they're willing to spend on one thing is often very restricted, and just nailing them down for a given project can be difficult. There are a lot of other things they can be doing with their time. You want to get things right the first time and not spend five years doing it all over again (and more likely, the project is simply cancelled). NASA's budget being sizable, gives them some flexibility on spending more to increase positive outcome chances.

It's better to spend $10b on James Webb (if necessary), than $3b with a meaningfully higher chance of failure. If it fails at the lower budget, they still won't do it all over again, it'll just be dead. It'd be a massive loss of time and potential.

I for one like the fact that different national space agencies are going with completely different approaches. I doubt that the skycrane used for the NASA's Martian Science Lab mission could be accomplished on a shoestring JAXA budget. On the other hand, an asteroid sample return mission would've cost billions and taken decades if NASA had done it.

At the end of the day, we all take different paths but ultimately get to the same place. Humanity wins.

> They'd rather spend $2b to get a 97% likelihood of a positive outcome, than $1b with an 80% likelihood.

I don't know if you picked your numbers out of a hat, but napkin statistical analysis (if I'm not mistaken) is that two attempts with an 80% likelihood would have a 96% likelihood of at least one success. That's not quite exactly 97%, but if your numbers are not their actual assessment, it's worth wondering if their bets are the good ones.

The thing that skews their monetary needs towards the expensive option with a better likelihood, as I understand it, is that they can't afford the bad PR that comes from failure.

But in pure, statistical terms, they might, actually be better off betting less money on projects with less likelihood of success. They would fail more often, but overall, would spend less money for the same amount of success.

"as I understand it, is that they can't afford the bad PR that comes from failure."

I think that's the real problem.

NASA tried "faster, better, cheaper". It didn't work out so well.

Mars Climate Orbiter, Mars Polar Lander were two missions executed under the FBC regime. They failed, arguably due to underfunding that elevated risks. It's hard to say, because the statics are sparse.

I'm personally ok with paying taxes to fund missions at the level of, say, New Horizons (~$700 million over 15 years) or Cassini (>$3B).

But the point of a cheaper project that has more likelihood of failure is that you could try again. You can even learn from the previous failure to reduce the likelihood of failure on the second attempt. Plus, since you've done a lot of the development for the first mission, you probably can save a lot of time and money. But that's armchair space exploration engineering. I could be completely wrong in my assumptions.

That being said, what is non compressible, though, is the time to get to where we need to go (Mars, Ryugu, or whatever else in space), and having to retry is a set back that can be counted in large number of years.

Don't forget that every "try again" adds 80-150M to the cost for new launch vehicle.

SpaceX is starting to change this, but most of the NASA history / planning was done in context of high launch costs.

The James Webb scares me. They now have put so much money and time into it that a failure would just be unimaginable. I am sure they have run the numbers but personally I am not sure if they wouldn't be better off with iterating faster and cheaper while accepting some level of risk.

Maybe a failure of the James Webb would be a good motivator for a manned mission to the Lagrange point :)

Reply on news.ycombinator.com