We’ve spent some more time looking at the functionality of the Toronto Crime Map. As mentioned in an earlier post, we find it astonishing that the icons disappear at a normal zoom level.
One continuing thought we have about these types of maps is to compare the Toronto Crime map data to press releases of the Toronto PD that provide more specificity to the incidents address.
If privacy is such a concern, then shouldn’t the press releases have off center information. We recognize the absurdity of this comment because then no one would trust or use the press releases. We, at SpotCrime, feel that this lack of consistency will result in the same treatment for the Toronto PD crime maps. No one will trust or use them.
In picking out the press release data to compare to the map, we noticed a string of assaults in the Thorncliffe Park area in October – mostly against children. We’ve mapped these events on SpotCrime as best we can based on the address information provided.
Here’s the link:
To get to this map on SpotCrime. Sign in. Go to our Toronto Map. Change the date range to October 1, 2015 to Today. Deselect all crimes and then just select the assault icon.
Finally, add the word “Thorncliffe” in the keyword search.
(In this particular case, we’ve moved the icons a few feet from each other because the address information is not specific or provides the same intersection. In a regular case, the icons would be on top of each other and details would just be on the crime list below the map.)
The end result of the search is three assaults against children and one where a man with a knife entered a woman’s apartment in the early morning. The assaults against children for the month of October in the Thorncliffe area range from a sexual assault of a eight year old girl, attempted abduction of a 9 year old boy, and another 9 year old boy sexually assaulted during a tutoring session. The Toronto Police have indicated there may be more victims of the tutor.
It is unclear to SpotCrime if there is a pattern in Thorncliffe or if October is any different from any other month in terms of sexual assaults. Toronto PD took great care to bring this information to light in their press releases, and provided some minimal address information. But their crime map has little information and or search capabilities to find this information.
Let’s say you heard that there were a few sexual assaults against children in October in the Thorncliffe area and you wanted to check the Toronto Police Crime Map. Got to the “public” crime map. Click past the disclaimer immediately limiting your rights in what you can do with the information. Then over to the right select assault and sexual assault. Then you need to zoom in to the area of interest. Typically to navigate a modern interactive map, one would use the address search bar. Notice the great detailed Esri map below – complete with topography details.
But where are the crimes we saw in the first map before we did an address search. If you remember from our last post, you now need to zoom out a few times to see crime icons.
The point here is that it is not intuitive to not see icons when you search and address, and unclear without instructions that you need to zoom out. Anyone using a map for personal use, is likely to search on their personal address, and then see no data being presented. In our view, most members of the public would determine this to be a glitch and good portion of the users give up on the search. At Spotcrime, we understand that you have to honestly engage the view with relevant information or you’ve lost their attention. If you really want to drive tips, you’ve got to capture attention and some semblance of usability.
We know very little about the crimes highlighted above except date and time. There’s no approximate address listed. By zooming out and in a few times, I can see that the incident on the 19th probably occurred at or near the Thorncliffe Park Public School. The incident on the 20th looks like it is near the Marc Garneau College Institue.
Let’s say you saw an assault near the public school on the 19th, and you considered providing more information. Would anyone take the time to zoom in and out to verify that was the right assault, or the tipster conclude there wasn’t a report for the event. There does not seem to be enough information to drive any tip information from the public.
Why do the assaults mapped by SpotCrime from the Toronto press releases not show up on the Toronto PD Crime Map?
The assaults SpotCrime has mapped do not show up on this map because they are out of the date range. The current date range on the Toronto map is October 19th to October 25th. When Toronto Police map gets updated, the public does not have the capability to search older crimes from even a week prior. Again, without any historical search, how is this product going to create use and drive crime tips?
If you add the assaults on the Toronto map and the SpotCrime map, you get six assaults in the area. What is missing is the assaults not reported through press release and not accessible on the Toronto map. There’s no immediate way to see any assaults that happened from October 1 to October 18th. How is this limited access useful for the community?
In an interview with CP24, Ian Williams indicated that this map was created because the Toronto police “recognized a need from the community”. However, I think you would be hard pressed to find any member of the public to determine this map as satisfactory. The effort to control the information under the auspices of privacy has rendered the map useless.
When you have a map with so little identifiable information, it is very difficult to check the accuracy and hold the data output accountable. Even if any of the dates matched up, it would be a guessing game if the incidents on the press release were the same on the map.
Since none of the Thorncliffe assaults we had mapped showed up on the Toronto PD map. We picked another press release a sexual assault of a 15 year old girl on October 23rd in the Empringham Drive and Sewells Road area.
SpotCrime has it mapped here:
Here it is mapped on the Toronto Police map:
How did we get the Toronto map to zoom in? After playing with the function, we realized that if you select a crime and then zoom in – the description bubble stays. Likely not intended, but we think we are able to pinpoint the random assignment of this particular crime two blocks away at another location.
Think about the level of difficulty to audit and correlate this one crime. We’ve taken the time because public crime mapping is our business, but the general public will have no patience for this goose chace. This map will not be used by anyone.
The problem with technology implementations like this is often the utility metrics never get measured. Toronto police have had the opportunity to mention they have entered the era of open data and no one has questioned whether this website meets any definition of open data.
Disappearing historical data and disappearing icons is not open data. Open data has no restrictions and allows the public to self tabulate the data and question the results if needed. The Toronto Police map brings no risk of accountability and provides no way to measure effectiveness. A most importantly has not utility to the public, provides no measure of public trust and does nothing to help reduce crime.
It’s a gossamer press release with an ephemeral crime map.
Solutions for Toronto:
1. Understand the true definition of Open Data. Know that the current crime map website for the Toronto PD meets no minimal definition of open data.
Produce an open data portal for the public that includes historical data. If there is a concern for privacy, use typical convention of block numbers for crime incidents.
2. Scrap the existing map. Throw it in the dustbin. If an open data portal is not within the comfort level of the police department, higher a vendor for free to map the data. Two vendors that will do this for free are Crimereports.com and Raidsonline.com. Both restrict the press from accessing the information. SpotCrime is not a fan of this solution because the data will not be open. But the public will get a much better implementation of a crime map with historical searches and a better visual representation of the data.