This is really cool and well done. I'm unhappy with the stats they chose though. I thought that 20x less robberies seemed unlikely so I clicked their source. They were using the absolute number of robberies not per capita. The way they used it is sneakiness bordering on outright lies. This kind of bad-faith abuse of stats always irritates me.
And not all robberies and thefts are reported to /recorded by the police. There are vast differences between countries in those numbers.
On the other hand, the car theft numbers is probably accurate, because people need to report to get insurance.