In this paper a new method of distinguishing shower-like from track-like neutrino interaction events using event reconstruction data is introduced. This method is then used to minimise the error...Show moreIn this paper a new method of distinguishing shower-like from track-like neutrino interaction events using event reconstruction data is introduced. This method is then used to minimise the error that a shower-like event is misidenti ed as a track-like event and vice versa, which is necessary for measuring the neutrino mass hierarchy in ORCA. In its current format, the method leads to errors of around 0.40 for noisy simulation data and approximately 0.15 for noiseless data. However, the method can still be extended to take into account more information of the reconstructions, thereby possibly improving the results.Show less