The general election of June 2017 marked an important watershed in the UK's trajectory towards Brexit. Until the June 2017 electoral contest, the path towards Brexit -of the so-called "Hard" variant- seemed inexorable. The unexpected outcome of the June 2017 election altered the prevailing calculus and re-invigorated alternative discourses regarding Brexit (and non-Brexit) pathways. In this paper we draw on VAA-generated data from the 2017 UK general election to explore citizens' preferences regarding different types of Brexit. To what extent can we detect shifts in Brexit preferences between the 2016 referendum event and the general election a year later? More substantively, how do Brexit preferences (and attitudes towards the EU more generally) fit within the overall structure of political contestation in the UK? And, what role do regional, national and European identities play in the articulation of Brexit preferences? To address the substantive questions we first apply dimension reduction techniques to the identity variables in order to explore how identity and Brexit preferences cluster in the UK context. In a second stage, we explore political dimensionality by applying scaling analysis to the lengthy battery of political attitude items typically included in a VAA. Apart from assessing the effect of latent identity and ideological constructs in explaining Brexit preferences, we are also concerned with demonstrating the added leverage of using Big Data analytics -such as cross-validation and machine learning approaches- that are typically not applied in election studies.