According to me, $\ce{Fe^{2+}}$ should be a better reducing agent because $\ce{Fe^2+}$ after getting oxidized will attain a stable $\ce{d^5}$ configuration, whereas $\ce{Cr^2+}$ will get a $\ce{d^3}$ configuration. I think the half filled $\ce{d^5}$ configuration is more stable than the $\ce{d^3}$ configuration. Why is this not so?