Yes, I found this guy's explanation very accurate. Check it out.
'At the beginning of the first transition series (scandium), the 4s orbital is lower in energy than the 3d orbitals. Thus Sc has the configuration [Ar] 4s2 3d1. This continues as far as V which has the configuration [Ar] 4s2 3d3. We might expect Cr to have the configuration [Ar] 4s2 3d4 but, due to the shape of the 3d orbitals and the increasing nuclear charge as we go across the series, the 3d orbitals are now closer in energy to 4s than they were at Sc. The energy difference is so small, in fact, that it is smaller than the 'pairing' energy to put two electrons in the 4s orbital. Consequently the configuration [Ar] 4s1 3d5 is lower in energy than [Ar] 4s2 3d4. For Mn, the next electron goes into the 4s orbital as a paired electron in 4s is still lower in energy than a paired electron in 3d. As we go further to the right, the 3d orbitals continue to fill up but, by the time we get to Cu, the energy of 3d is now lower than 4s, so [Ar] 4s1 3d10 is preferred to [Ar] 4s2 3d9.
So you see that the unusual electronic configurations of Cr and Cu are a consequence of the increasing stability of the 3d subshell with respect to the 4s subshall as you go across the series due to the increasing nuclear charge. It is interesting to note that in ALL transition metal cations, because of the increased effective nuclear charge in the ion, the 3d subshell is always lower in energy than 4s. This is why 4s electrons are always lost first on ionisation of transition metals'