In Frauchiger-Renner's paper, the authors propose a thought experiment which suggests that taking QM together with certain natural assumptions, one arrives at a contradiction. They go on to say that Many Worlds Interpretation throws out the assumption:
"I am certain $x =\xi$ at time $t$" implies "I deny $x \neq \xi$ at time $t$"
It is true that getting rid of the above assumption does permit the results of the experiment, but that assumption seems true by definition of equality. Further, it is my understanding of many worlds that in a given "branch", things are consistent. That is in any given "world", the assumption above holds, while in general, or universally, every possible outcome happens. Thus I don't think MWI discards the above assumption in a way that can explain the experiment. Where I am confused?