We examine the kinematics of the Galactic halo based on SDSS/SEGUE data by Carollo et al. (2007, 2010). We find that their claims of a counter-rotating halo are the result of substantial biases in distance estimates (of order 50%): the claimed retrograde component, which makes up only a tiny fraction of the entire sample, prone to contaminations, is identified as the tail of distance overestimates. The strong overestimates also result in a lift in the vertical velocity component, which explains the large altitudes those objects were claimed to reach. Errors are worst for the lowest metallicity stars, which explains the metal-poor nature of the artificial component. We also argue that measurement errors were not properly accounted for and that the use of Gaussian fitting on intrinsically non-Gaussian Galactic components invokes the identification of components that are distorted or even artificial. Our evaluation of the data leads to a revision of the estimated velocity ellipsoids and does not yield any reliable evidence for a counterrotating halo component. If a distinct counterrotating halo component exists it must be far weaker than claimed by Carollo et al. Finally we note that their revised analysis presented in Beers et al. (2011) does not alleviate our main concerns.

17 pages, 12 figures, submitted to MNRAS