There are many times as a skeptic that I wish I were wrong. I really want to detect an alien artifact, and would love free energy, cold fusion, and a cure for cancer. I completely understand why these ideas have endless allure and the temptation to engage in a small bit of motivated reasoning to see the science from a particular, if odd, angle. But science does not progress this way. It progresses through the cold and heartless removal of error, by brutally smashing the pillars of our own vanity, fear, and desires, and by controlling for our own biases and shortcomings. I often refer to the peer-review process as a meat-grinder – it chews up and spits out ideas, but there is a product at the end – and that goes right back into the meat-grinder for another round.

One more really tempting idea now bites the dust – the EM Drive. I first wrote about this almost seven years ago. The idea is to create propellantless propulsion. This would revolutionize space travel, and could potentially even create that flying car we always wanted. Now, in the world of physics, in order to accelerate something there needs to be a force acting on it. If you want a rocket to go up, then you need to throw some mass from the rocket down so that the mass and velocities match (equal and opposite). So rockets need propellant, something to throw out their back. Ideally this is something very light that gets accelerated to really high speeds to produce the maximal thrust to the rocket.

While this concept works just fine, it is also extremely limiting, by something known as the rocket equation. The rocket needs to carry enough fuel to accelerate the entire rocket, including all the fuel it is carrying. So it needs fuel to carry the fuel to carry the fuel… This means there is a geometric rather than linear relationship between speed and range and how big a rocket and its fuel has to be. For many chemical rockets the fuel is the propellant; when you ignite it the fuel rapidly heats and expands and gets pushed out the exhaust. Other rocket designs may have a separate energy source and propellant. Ion drives, for example, create energy to power magnets which accelerate charged particles to extreme velocities.

But what if you did not need propellant? What if all you needed was energy, and could somehow use that energy to create thrust without having to throw any matter out the back end? That would drastically alter the rocket equation. This would reduce the cost of space travel and open up the solar system. It might even make it practical to get to nearby stars – in a hundred years we might have a fusion powered ship that can zip around the galaxy at a constant 1G acceleration.

There is a small problem with this vision, however – it breaks the known laws of physics. It is the equivalent of a free energy device or perpetual motion machine – it’s easy for people to convince themselves that they have finally cracked this problem, with a clever-enough arrangement of magnets or whatever, but they never work because physics. In this case Newton and his equal and opposite thing gets in the way. But what if we cleverly design a chamber to bounce around microwaves so that they push a little bit more in one direction than the other? Just as with perpetual motion, it just can’t work. The laws of physics will not be denied.

Now of course, any good scientist or philosopher would admit that our current laws of physics are just approximations of reality. They are models we use to predict how the universe works, but are not the ultimate actual rules of existence. So it is possible that the laws of conservation, thermodynamics, and motion may be correct as far as they go, but there is a deeper level to reality that allows for exceptions in special cases. This would be like special relativity and general relativity tweaking Newton’s laws of motion. Maybe there is something even deeper.

What proponents often say, then, is that their machine (which they can demonstrate to work – but more on that in a bit) represents an anomaly, which points in the direction to a deeper understanding of the laws of physics. This is exactly what lead to relativity – anomalies that could not be accounted for in classical physics, like the orbit of Mercury, because relativistic effects were measurable.

Skeptics acknowledge all this. Sure – maybe you have discovered new physics. But the burden is on you to prove it. It’s not as if we don’t have hundreds of years and countless examples to inform our thinking about how likely it is. On the one hand, you made some minor error in calculations or measurements. On the other hand, you have shattered the modern world with new physics, and you are basically the next Einstein. I know what I want to be true – but I also know what is overwhelmingly likely to be true. So far, every single free energy/perpetual motion claim has flopped, but sure, that next one just might be the one. I’m just going to bet on the laws of physics, and will likely always be right even though I wish I would be wrong.

Let’s now deal with the – evidence for an anomaly – aspect of these claims. That is always at the core of such claims, an experimental setup that demonstrates anomalous energy or, in this case, anomalous propulsion. But there is always a red flag in these claims (I am not even including the deliberately fraudulent claims, just the true-believers). The measured anomalies are always tiny. This absolutely matters, because it means the signal to noise ratio is also tiny, and as this ratio decreases the effect of small experimental errors become geometrically greater. This doesn’t mean we cannot measure tiny signals against a loud background of noise. Some experimental setups are mind-blowingly able to do so, such as LIGO which measures gravitational waves. But in these cases we have massive installations or extremely high-tech setups painstakingly designed to eliminate noise. They are not running on someone’s tabletop.

In any case – when on the one hand we have the known laws of physics, and on the other hand we have a tiny anomaly barely above the level of noise, I will bet on experimental error every time. That’s just playing the odds. At the very least, the massive burden of proof is on the person claiming new laws of physics to prove that their tiny anomaly is real. Of course, they always promise that it will “scale up”. Great – talk to me when you make that happen. That is one of the main ways such claims fail – because the small experimental errors don’t scale up like the promised real effect would. The other way they fail is to conduct tighter experiments where small errors are better controlled for.

That is now what has happened to the EM drive. The Institute of Aerospace Engineering in Dresden, Germany conducted a “High-Accuracy Thrust” measurement to remove “false positives” and that is exactly what they did. They tweaked the experimental design to eliminate possible false positive thrust measurement. They powered the device with a battery to avoid effects from external energy. They further report:

After many iterations, we developed an inverted counterbalanced double pendulum thrust balance, where the thruster can be mounted on a bearing below its suspension point to eliminate most thermal drift effects.

With this new experiment designed to eliminate tiny false signals, they detected no anomalous thrust. The EM drive is effectively dead, gone the way of thousands of perpetual motion machines before it. But of course that does not mean it is truly dead. It will likely continue to exist as a zombie theory, thriving on the fringe along with cold fusion and zero-point energy. The idea is simply too good to give up simply because it’s wrong.