The ability to filter results according to user preferences, to bring meaning to complex information, is an indispensable functionality of any great data analysis and visualisation tool. The San Francisco Crimespotting map provides a good example of how well designed controls can enhance user ability to visualise only a subset of available information, specific for the problem under investigation.
In particular, the map uses coloured map markers to show the location of different types of crime. Users can filter the types of crime displayed by simply selecting the items from a list. A click on individual marker brings up information about a particular event. The 'date' and 'time of day' controls allows to filer the results shown on the map by any date range or for any period, or combination of periods during the day.
The 24 hour clock (literary!) takes a few moments to get used to but is a very nifty idea. I have seen various approaches so far, for example using sliders or day-time matrix, but this one takes the prize.
For other examples of visualisation of crime statistics please refer to my previous posts:
Mapping crime in Queensland
Crime maps - complex stats visualisation
Victorian version of crime stats
UK crime statistics revisited
Mapping Crime in Canberra
First spotted on Google Maps Mania