Correlation Versus Causality

I saw this neat graph by way of Instapundit. It shows “correlation versus causation” in a way that words might fail, proving with statistics that Microsoft's Internet Explorer was the cause of so many murders.

After all, the number of murders fall in proportion to IE's market share, implying that IE drives people to murder. Therefore, in order to reduce murders even more everyone should abandon using IE.

Makes perfect sense, doesn't it?