Claims of detected anomalies have been made as soon as the WMAP first year data was released. Is the WMAP power spectrum anomalous? What about the hot and cold spots in the WMAP maps? How can we best constrain the properties of the primordial potential fluctuations using current and future CMB data? I will present Bayesian and frequentist methods to address these questions and show results from our analysis of the WMAP data.