It is easy to account for ion formation for the alkali metals. These are supposed to have only the 1 valence electron, which is lost after it oxidizes a non-metal: cf #NaCl, KCl, CsI# etc. These metals, upon oxidation, have lost all their valence electrons (the which means the outermost electrons). Transition metals have valence electrons amongst the #s# shell and the #d# shell. Electrons from the #d# shell are less penetrating (i.e. and as a consequence less attracted by the nuclear charge), and multiple oxidation states are possible: cf. #Fe(I)#, #Fe(II)#, #Fe(III)#, and even further oxidation states. We remember that oxidation numbers are a formalism, but they are still useful to track metal reactivity.
Most transition metals, #Fe#, #Cr#, #Ni#, for example, have multiple states (including, curiously, negative ones and ones with zero oxidation states). Oxidation states are our attempts to categorize and classify this chemistry.
