So a $\psi$ is a meson with a charm and an anticharm quark. The particle data group lists the following branching fractions for the following decay modes of the $\psi$:
hadrons: 87.7%
- virtual $\gamma\rightarrow$ hadrons: 13.5%
- g g g: 64.1%
- $\gamma$ g g: 8.8%
Now if I understood it correctly, the virtual photon should just contribute $\alpha$ to the probability amplitude (because it introduces two new vertices in the feynman diagram with each having $\sqrt{\alpha}$). So I would have assumed that the branching fraction for the virtual photon decay mode, which is proportional to the square of the probability amplitude, would have been
$BF(\gamma\rightarrow h) = (BF(ggg)+BF(\gamma gg))\cdot\alpha^2$
Now, obviously, this is way less than 13.5%, so where have I gone wrong?
EDIT:
Here the feynman diagrams I thought of for the "g g g" decay mode and the equivalent with a virtual photon:
From this, I thought that the virtual photon just added two vertices, and that I have to multiply $\sqrt{\alpha}$ to the probability amplitude for each of them. Are these even correct?