Mann did finally admit to being wrong about the hockey stick.
>We conclude that model overestimation of tropospheric warming in the early twenty-first century is partly due to systematic deficiencies in some of the post-2000 external forcings used in the model simulations.