I have known about MW for 25 years or more, thanks to popular science writers. I don't see it getting more traction than it already has absent some way to falsify it. It's stylish from a distance, but like imaginary numbers, you should tidy up your multiverse when you are done playing with it.

That's a common criticism, but many-worlds is a consequence of a theory, not a theory itself. Plus all the other candidates are worse. (And Copenhagen isn't really a theory at all since it never gets around to defining what a 'measurement' is.)

Many worlds requires postulates (Everett's word!) that are not a consequence of theory, in particular that the wavefunction is an objective property of a particle. This is very much in dispute, and is rejected by the ensemble interpretation, consistent histories, etc.

I don't agree that those interpretations are inferior to MWI. In particular those theories postulate that the world is essentially probabilistic, and so do not have MWI's trouble with predicting probabilities.

You might find this interesting http://lesswrong.com/lw/q4/decoherence_is_falsifiable_and_te...

>Many worlds is by far the best hypothesis to date.

well, to me it actually works in the opposite direction, like prove by contradiction - if the Many Worlds is the best/logical/natural consequence of the indeterministic view of QM than the indeterministic view is definitely have to be overturned.

All the Bell experiments i read about so far have obvious gaps allowing for obvious explanations. And if you take the latest one, the heralded loop-free, then there is obvious question - why didn't they published the statistics of the unconnected, non-entangled runs? I mean it is obvious from current QM orthodoxy that non-entangled runs wouldn't produce S>2, yet from alternative, locally-realistic, experiment result explanations the S>2 would be produced in that case too (the S>2 isn't possible in the locally realistic world, yet it is possible to locally realistically explain S>2 in all the Bell experiments that so far i've seen)

Many worlds is not a logical consequence of the non-deterministic view of quantum mechanics. One of the reasons cited in favour of the MWI is that it is wholly deterministic. That is probability doesn't appear at all in it's framework. This is much more in keeping with previous physical theories (Newtonian mechanics and Maxwell's laws of electrodynamics for example) which were also wholly deterministic.

If you're interested other reasons the MWI is often considered "nice" is that it is: linear, local, causal, realist, entirely unitary and in some sense (this is often disputed) minimal. Other interpretations have to add mechanisms (for example wavefunction collapse) in a rather ad-hoc way in order to stop many worlds happening.

That is also one of the most common critiques: because MWI is wholly deterministic, there is no way for probability to arise in it, and therefore no way to derive probabilistic phenomena like the Born Rule.

Many worlds is a consequence of the idea that the laws of physics apply equally to physicists and experimental apparati as they do to the particles under study.