In an article about bike theft for The Times, one computer scientist outlines how he attempted to convince the police to search for his stolen bike using math.
Bike theft is one of those crimes that doesn't get given a lot of police attention. In Cambridge, UK, where the computer scientist lives, the police will not investigate CCTV footage if you can't pin down a precise time-frame for the theft.
“The current force policy is that we will not trawl through CCTV if the time-frame exceeds more than four hours," a spokesperson for the police said after one theft in 2017.
“The time-frame of the theft that occurred is 10 hours, and we just simply do not have the resource to trawl through 10 hours’ worth of footage.”
However, as the computer scientist points out, you do not need to search hard, you need to search smart. Say you had a rough idea that your bike was stolen sometime over the last two days. You do not have to sit there, fast-forwarding through the footage until you find someone dressed up like a stock photo bike thief walking up to your bike.
If you employ a binary search (or a binary chop, if you'd like a cooler name for it) you can review the footage and find the bike thief in a fraction of that time.
Say you want to find an item in a sorted list of items. In order to find it using a binary search, you go to the midway point and see where in the list of items you are. You then go to the mid-point in the half of the list which you now know contains your item, and repeat until you find it.
In the case of the cyclist, what he attempted to convince the police to do was go to the mid-point of the long CCTV footage and see if the bike was there. If it was, they should skip to three quarters of the way through the footage and check again, repeating the process until they can quite quickly find the bicycle.
According to the Times, he pointed out that "if the CCTV footage stretched back to the dawn of humanity it would probably have only taken an hour to find the moment of theft".
Apparently, this did not go down well with the police.
[H/T: The Times]