A handy clip 'n' save guide:
1. Vice presidential selections rarely affect election outcomes.*
2. The selection is therefore only likely to be important insofar as the VP choice (a) helps or hurts the president they serve during his time in office and (b) becomes more likely to be a future president.
3. The selection should therefore be assessed primarily in light of #2, not #1. (It will not be.)
* You could tell a story where Obama's VP could help prevent defections from white working-class voters who would otherwise have voted Democratic (a possibility that was obviously not relevant in past elections). However, this idea is purely speculative and would be difficult to test even after the fact.