Readers at Climate Audit have been discussing the suggestion that a graph by Neukom et al produced a misleading rhetorical effect by centring the series - one paleoclimate reconstruction and a series of climate models - on the second half of the twentieth century. As a commenter observed, since the models run forward in time, the series should actually have been centred on a single point at the beginning. Then their divergence over time could be observed.
In response, Climate Audit regular UC has put together an animation showing how these kinds of spaghetti graphs might look centred on different periods. He has used paleoclimate rather than model data for convenience. It's rather beautiful I think.