[stale2002]

The way science works is by doing experiments.

The people have been doing experiments on the EM drive and have been getting unexpected results.

Nobody really knows why yet, but that's kinda the point.

Is thermodynamics flawed? Is it pushing off of "dark matter"? Is it a completely new and novel scientific effect?

Any of these circumstances would be interesting.

[DanielBMarkham]

Yes. Even if it's the most likely thing, experimenter error, for a half-dozen teams to all have errors and not spot them? That's quite interesting! It's the same situation we had with the FTL neutrinos a while back.

[euyyn]

Except the neutrino paper then was a big lot of physicists asking colleagues to help them review, so as to find where the problem was. What I see here is just hype. The former, as opposed to the latter, is what gives confidence that the people involved are doing their darnest not to fool themselves. That what they're doing is actually science.

[DanielBMarkham]

I don't know. I see a few universities looking into building their own setup. I see the Chinese filing patents, testing them in orbit. I see a 2017 microsat launch. I see a healthy debate continuing over on the reddit Emdrive board.

That doesn't look like people trying to sell things to me. That looks like a lot of different people from different backgrounds playing around with an experimental setup that does something unexpected.

The only difference I see is that with the FTL neutrinos, folks were saying this breaks a mathematical model that we've proven over and over again. With the emdrive, folks are saying If this works, it's by some model we do not yet understand. It doesn't attempt to disqualify an existing model; it suggests there's some other model applicable that we don't quite understand yet.

For most folks, this is quite more upsetting than the FTL guys. I get that. Still, observe first, theorize second. I'm sure the theory folks will catch up, and it'll probably be some sort of EM leakage we're talking about. But maybe not. Maybe we've accidentally come up with a new kind of photon rocket. That's what makes the story fun.

[euyyn]

> It doesn't attempt to disqualify an existing model

It would very much disqualify conservation of momentum, which is another way of saying that the laws of motion are the same here and one meter down the road.

Conservation of momentum has been observed in countless experiments for centuries, in a broad range of conditions. It was indeed observed first; there's no catching up needed for the theory. It's always valuable, and in a sense exciting too, to test it again beyond the range where it's already been tested, even if you expect a negative result (the null hypothesis holding). But that's not what's happening:

If your goal is to learn, and you see momentum not being conserved in your prototype space engine, you start dismantling it, simplifying it, until you stop seeing it. And then go back one step to see it again. Your device might not be a space engine anymore! But now it's simplified; easier to reproduce, and easier to model and analyze. The laws of nature don't care about the purpose and coolness of your device, they apply the same. You also take your measurement apparatus and calibrate it against something unrelated, in many conditions, to make sure it's not lying.

You might have a different goal than learning: to achieve cheap space travel, which is a great thing. If then you build a prototype space engine with the hope that it won't conserve momentum, and you see a positive result, there's already the first red flag. If when you see that, what you do is trying to explain it with new (or with wrong) theory instead of questioning it, there's the second red flag. Healthy debates, and playing with technology in a lab, are good things. But not all good things are science. Our minds are too easy to fool, and people are doing this without condoms.

[mhneu]

> With the emdrive, folks are saying If this works, it's by some model we do not yet understand.

A summary of the parent article is "The experimenters did not do the obvious control experiments [e.g. testing under a null load], thus there are currently no results at all."