I review the observational prospects to constrain the equation of state parameter of dark energy and I discuss the potential of future imaging and redshift surveys. Bayesian model selection is used to address the question of the level of accuracy on the equation of state parameter that is required before explanations alternative to a cosmological constant become very implausible. I discuss results in the prediction space of dark energy models. If no significant departure from w=-1 is detected, a precision on w of order 1% will translate into strong evidence against fluid-like dark energy, while decisive evidence will require a precision of order 10^-3.