The science we can perform with an instrument is determined by the information contained within the data and the ability of our methods to extract this information. Next-generation low frequency wide-field radio interferometers, such as the Murchison Widefield Array (MWA) in Western Australia, are capable of delivering exciting new science results, but present new challenges to understanding our instrument and designing our analysis methods. In this talk I will present some of the information theory techniques being used to understand the limitations of MWA datasets, as a crucial testbed for SKA-low design. I will then present one example of where an understanding of the instrument informs the strategies used to detect the Epoch of Reionisation signal from high redshift neutral hydrogen.