"The problem is it's a computer simulation and as we used to say about computers back in the 70s - 'junk in, junk out.' In this case, Ralph even admitted in one of the interviews, 'However, I emphasize it is a simulation – we “cheat” when we prepare the 2nd photon that represents the time traveler in the past' (see http://thespeaker.co/
Marshall continued, referring to information he has in his new book based on his time travel report to the U.S. Congress, Paradox Lost:The True Geometries of Time Travel.
"So they have a situation where a photon goes through a wormhole to the past but they ignore the past has to be in a parallel universe. In the other example, with the closed time-like curve, each iteration of going back will be another parallel universe. If it doesn't, then you have the geometry that Stephen Hawking talked about in his Chronology Protection Conjecture which I proved wrong with a physical classroom experiment in 2012 (see http://network.nature.com/
"What they've done," Marshall continued, "is create a scenario that fails my MCEBPS formula which tests for paradoxes and if they're there, it kicks them out (MCEBPS stands for Marshall's Copenhagen Everett Barnes Paradox Solution). In this case they send a photon to the past in simulation where it meets itself. That's a violation of the Copenhagen Interpretation of quantum physics because we have a photon in the past that's now meeting itself from the future which didn't happen before and Copenhagen says you only get one outcome per measurement which means this new outcome has to exist in a parallel universe. Notice, it has nothing to do with the quantum state of the photon, but where it is in the space-time geometry, in this case, the past. So, the parallel universe part of this is provided for by the Everett/Wheeler hypothesis which is the 'E" part of the formula, thus providing a foundational alternative to Copenhagen. Next is the B for Barnes part of the formula where I weigh in with my contribution which is to insist on either having created a record of the past event when it was the present, or using evidence from the past to show we now have a new past which exists in a parallel universe. In this case, since it's a computer simulation, we know the photon is in the past and alone. By introducing its future self through the wormhole, it becomes obvious it didn't happen before, so this new encounter requires a parallel universe solution".
Marshall says the overall results of the Ringbauer/Ralph experiment do mirror those of thought experiments of Igor Novikov and his Principle of Self-Consistency that Marshall referenced in his paper, Experimental and Theoretical Analysis of Chronology Protection Conjecture Failing On The Discovery Channel. However, though Novikov's calculation shows how paradox can be avoided, it ignores the space-time geometries that result from true time travel and his experiment violates what Marshall calls, a FAOC or first action of cause. In this case, although Novikov claims his principle proves any event that gives rise to a change in the past would have a probability of zero, he ignores that the act of time travel to the past causes a change, just by taking place because it didn't happen before. The arrival in the past, from the future, would be the FAOC which then renders Novikov's idea moot, despite its self-consistency.
"The bottom line is there's so much misunderstanding and misinformation about time travel, because no one in physics is actually getting paid to figure this out and there's so much reliance, conceptually, on ideas from science fiction which are not the science part but the plot twists for the fiction part."
Marshall will be doing a special presentation on time travel at Harvard in July. For more info on him, you can see his profile page at http://about.me/