Can statistics lie

Answer 1

No. However they can be used to mislead people or twisted to indicate things that seem good but aren't (this is called spindoctoring) in various ways. Examples include driving accident statistics - 45% of crashes are caused by drunk drivers, but that means 55% were caused by sober drivers - and crime statistics - 5000 murders were committed in one city but only 500 were committed in another; a person might assume the second city is safer, but if the first city had 100 times the population then it is actually only one-tenth, not ten times as dangerous as the second.