back to article Oi, London thief. We KNOW what you're doing - our PRECRIME system warned us

Italian boffins claim to have devised a method of predicting future crime trends in London using creepy footfall data collected by a mobile phone provider. A team at the University of Trento said they are able to work out whether an area will have a low or high crime rate with an accuracy of 70 per cent. The researchers' …

  1. Anonymous Coward
    Anonymous Coward

    70% accuracy?

    I daresay the local coppers can do better than that at predicting where and when crime occurs.

    1. Vociferous

      Re: 70% accuracy?

      I'd expect simply looking at last months statistics would give you a better than 70% estimate of which areas will have high crime this month.

      1. frank ly

        Re: 70% accuracy?

        The local coppers could even predict the type of crime and its daily, weekly and seasonal variation in different areas. They don't need mobile phone data to determine footfall either.

        1. Dan 55 Silver badge

          Re: 70% accuracy?

          Bet it won't stop 'em asking for it though.

          1. Trigonoceps occipitalis

            Re: 70% accuracy?

            What is this "local copper" of which you speak?

            1. Fungus Bob Silver badge

              Re: 70% accuracy?

              "What is this "local copper" of which you speak?"

              That would be the wires installed by the local telephone company...

              1. Eltonga

                Re: 70% accuracy?

                "What is this "local copper" of which you speak?"

                That would be the wires installed by the local telephone company...

                Or the 230V electrical ones. Some are 3-phase, some are one-phase. Your choice.

  2. Anonymous Coward
    Anonymous Coward

    So next month Elephant&Castle will still be a "Clockwork Orange" shithole?

    Precrime should tell me that I'll get mugged if I walk down that particular street right now, not just be a slightly higher resolution crime incidence map.

    Maybe they've got some interesting observations about where people go when & how this correlates to types of crime, but it's drowned out by this disappointing precrime angle?

  3. gazzton

    I read the first paragraph as "creepy football" data

    Now I've read it properly, creepy footfall conjures up an entirely different but no less odd mental image.

    1. Kubla Cant Silver badge

      Re: I read the first paragraph as "creepy football" data

      "Creepy footfall" - it sounds like they are identifying criminals by the way they walk.

      1. breakfast
        Paris Hilton

        Re: I read the first paragraph as "creepy football" data

        It is a matter of record that the way someone walks can indicate that they are a ladies man and that they have no time to talk. This is basically a straight extrapolation on that information.

        1. Anonymous Coward
          Thumb Up

          Re: I read the first paragraph as "creepy football" data

          It is a matter of record that the way someone walks can indicate that they are a ladies man and that they have no time to talk.

          Well played, sir.

  4. lglethal Silver badge


    Am I seeing an increased likelihood of crime around the El Reg office in the next month? Well I guess it is Friday, and some things have to head out the back door whilst the bosses are enjoying their Friday liquid lunches, no? wink, wink, nudge, nudge...

  5. russell 6

    Statement of the obvious

    Looking at the map, most of the red zones are around central London tube stations most used by tourists and targeted by pickpockets. Was it really that much effort to work out crime will happen in such obvious areas?

    They should have used a normal town as an example, it would be much more representative. It is pretty much guaranteed that crime is going to happen around Oxford Circus tomorrow afternoon, it is a pickpocket's paradise

    1. Anonymous Coward

      Re: Statement of the obvious

      Drug dealing / Shooting - sink-hole estates

      Pick pocketing / bag snatching - tourist areas and tube stations.

      Drunken violence - city centres and nr pubs, peaking on Friday and Saturday nights,

      Under-age drinking, nr parks, co-ops & bargain booze

      Football violence, pubs near football grounds, mainly Saturday and Sunday afternnons.

      I make that a lot more accurate than 70%

      1. Marvin O'Gravel Balloon Face

        Re: Statement of the obvious

        ... Ursine defecation - woods.

    2. Elmer Phud

      Re: Statement of the obvious

      "Looking at the map, most of the red zones are around central London tube stations most used by tourists and targeted by pickpockets."

      The pick pockets often work in conjunction with the 'find the lady' scammers.

      Are they listed as well?

  6. Anonymous Coward
    Anonymous Coward


    Sounds more like the stat of Asimov's Psychohistory to me!

  7. Obitim

    Bit Obvious

    Did the Ministry of the Bleedin' Obvious sponsor this study?

    Go to any town and ask a local where you're likely to get mugged, get your car broken into or get pick pocketed and I bet they'll be able to tell you with greater than 70% accuracy (they may even be the one to do one of the above!)

  8. Anonymous Coward
    Anonymous Coward

    Ah, the mythical footie fan

    that lives within walking distance of the team he/she supports. Well, at least for most of the Permier League this is true. I walk to my teams (conference) ground in around 10 mins.

    Will the group of Ipswich fans who travel from Norway a few times a season spark a crimewave?

    Will all the Man U fans from Brighton lay waste to the Regeny Palace when the get beaten or non of their expensive signings fail to score?

    These are in my mind more important questions.

  9. Anonymous Coward
    Anonymous Coward

    The proposed approach could have clear practical implications

    As a hardened criminal, or at least one who likes to be on the ball, I would be delighted to subscribe to the service (ideally an app) that shows me where my competition (and plods) are most likely to appear in large numbers. Bring in real life, dynamic results, please!

    1. Joe Harrison

      Re: The proposed approach could have clear practical implications

      Crimbr 2.0 is an innovative cloud-based solution which blends leading-edge academic research with dynamic crowd-sourced information to deliver attractive realtime offence-commission opportunities that (contd. p.94)

  10. Stretch

    Interestingly if anyone was to make use of this information it would instantly become wrong.

    1. Wensleydale Cheese
      Thumb Up

      "nterestingly if anyone was to make use of this information it would instantly become wrong."

      Quantum computing then.

    2. Eltonga

      Except if it was a first time (and pretty clueless) criminal looking for target areas.

  11. Anonymous Coward
    Anonymous Coward

    If I was living in London -

    I sure wouldn't mind knowing that crime will happen in my street next month, but I'd like it a bit more specific: Pickpocketing and burglary would be high on my list of concern, while domestic violence would be somewhat lower (because I'd know in advance if I was in a risk group for that). Maybe those Italians could refine their research into varioius crime categories. Then again, I'd very much like to be assured they don't get funding from a certain well-known Italian organisation...

  12. Bernard M. Orwell

    More like...

    ...CtOS than Minority Report...?

    1. Aggrajag

      Re: More like...

      Connective Tissue Oncology Society?

      1. Irony Deficient

        Re: More like…

        Convergent Technologies Operating System? shudder

        1. Anonymous Coward
          Anonymous Coward

          Re: More like…

          Complete Testicle Opinionated System? double shudder

        2. Bernard M. Orwell

          Re: More like…

          It's a reference to "WatchDogs", but your definition is close enough, Irony.

  13. Elmer Phud

    Hmm . . .

    . . .If you are a crim (or potential crim) and are aware how the software works then there should be no problem in getting round it.

    But - these 'detection and analysis 'products are usually sold to people who find Powerpoint an important marketing tool.

  14. i like crisps
    Big Brother

    "Don't leave home without it"...

    ...i think that was for American Express, but as regards crime, just leave your mobile at home if you're going to commit some....simples....but obviously DON'T commit crime.

  15. Graham Marsden

    Isn't it convenient...

    ... that London is built around a regular grid pattern so they can output their map in nice square blocks that line up with the roads...

  16. John Brown (no body) Silver badge

    footfall trends by time, gender, and age

    Wait...what? How do they get that information? Tracking where a mobile phone is one thing, but that level of data scraping is making a lot of assumptions about who is actually carrying it. Even the telco doesn't know who is carrying the phone or even if the actual user is the same person who bought or registered it.

    1. This post has been deleted by its author

  17. roger stillick

    NHK World ran this last night, called it 'Public Data'

    Joke Alert= the city maps were online... the boxes are areas to have a police presense, suggests limited coverage in the non-boxed areas...the police are convinced they are 'supressing crime'...per the TV show, NOT EVERYONE is convinced this is a good idea.

    IMHO= this is really a map of where the Police Aren't... pricess to criminals w/ computer access...RS.

  18. Alister Silver badge

    Incident prediction

    Back in the early 90s I worked for an NHS Ambulance Service which was approached by researchers at one of the local Universities.

    They had this great idea for a piece of software they wanted to trial, which would predict the likelehood of an emergency call in any particular area the service covered, based on historical data from the last ten years or so.

    The idea was that the service could position it's vehicles close to the predicted scene of an incident, and therefore cut down on response times.

    Unsurprisingly, it didn't work, as real life doesn't work like that, but the statisticians were convinced their methodology was sound... In more cases than a little, the vehicles were at the wrong end of the county when the next emergency came in.

    Those of us working at the sharp end could have told them it wouldn't work, but the Trust spent thousands on the system before they eventually gave it up as a bad job.

  19. Anonymous Coward
    Anonymous Coward

    So the unshaded squares are areas with no crime predicted?

  20. cortland

    Late to the game but...

    I daresay that if footfall is indicative of criminal activity, one need look no further than for poor lighting, the presence of lurking places, and roads and walkway conditions that make it more difficult to flee from an assailant.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019