Why "connecting the dots" is the wrong way to think about stopping terrorism

Bruce Schneier has a great op-ed on CNN on why it's stupid to talk about whether the FBI should have "connected the dots" on the Boston bomber. As Bruce points out, it's only in hindsight that there's a neat trail of dots to connect, a narrative we can make sense of. Before the fact, it's a hairy, swirling hotchpotch of mostly irrelevancies, and it's only the "narrative fallacy" that makes it seem like a neat story in retrospect. The risk here is that intelligence agencies and the press will push this fallacy as grounds for taking away more rights and more privacy in order to "connect the dots" next time.

Rather than thinking of intelligence as a simple connect-the-dots picture, think of it as a million unnumbered pictures superimposed on top of each other. Or a random-dot stereogram. Is it a sailboat, a puppy, two guys with pressure-cooker bombs or just an unintelligible mess of dots? You try to figure it out.

It's not a matter of not enough data, either.

Piling more data onto the mix makes it harder, not easier. The best way to think of it is a needle-in-a-haystack problem; the last thing you want to do is increase the amount of hay you have to search through.

The television show "Person of Interest" is fiction, not fact.

There's a name for this sort of logical fallacy: hindsight bias.

Why FBI and CIA didn't connect the dots (Thanks, Bruce!)

(Image: connect-the-dots, a Creative Commons Attribution Share-Alike (2.0) image from whitneywaller's photostream)


  1. From TFA:
    “In hindsight, we know who the bad guys are. Before the fact, there are an enormous number of potential bad guys.”

    The key word there is *potential*. Before the fact, there aren’t *any* bad guys.

    1. I think that there may have been some serious unintended consequences as a result of the FBI investigation.

      The older brother had started to become radicalized, and was maybe posting inflammatory stuff online, enough to attract attention. Then he gets interviewed by the FBI, who conclude at that point that he is not a threat. After this, he travels to Dagestan, and returns ‘radicalized’.

      If you are feeling out of touch, paranoid and angry with the culture around you, having law enforcement show their hand in this fashion is going to validate those feelings. Without the interview, he would have walked into that mosque in Dagestan an angry fantasist. Instead, he walked in with bona fide credentials, as a radical that was already warned by the FBI.

  2. “Or a random-dot stereogram. Is it a sailboat, a puppy, two guys with pressure-cooker bombs or just an unintelligible mess of dots? You try to figure it out.”

    It’s a headache. It’s always a headache.

    ‘Connecting the dots’ and other ‘pre-crime’ models are a recipe for harassment, and entrapment, but they are a recipe for profits for surveillance companies, and new powers for police/military/intelligence agencies and their extralegal “private” partners, and that ensures they continue to expand..

  3. There’s also a lot of money in pretending that the dots could have been connected with some company’s patented e-woo. As far as I can tell no government has been able to grow a resistance to e-woo.

    1. Also, the very phrase “connect the dots” is insulting to law enforcement, implying that any small child could do their job. This allows us to wag our heads and wonder what’s wrong with those morons. But I’m willing to bet that actual detective work is harder than “connect the dots.”

      1. But I’m willing to bet that actual detective work is harder than “connect the dots.”

        From what I read in the news every day, it also seems to involve making up new dots where the correct picture can’t be drawn with the existing ones.

        1. An awful lot of that.  Hopefully it’s the exception, and that’s why it makes headlines, but I have no statistics.

    2. Irresistible e-woo may be our saving grace.

      If we encourage the powers that be to double the data they hold on citizens every 18 months, the processing power needed to effectively mine that data will never be attainable.

      1.  Unfortunately not.  The processing power needed to ineffectively mine the data, coming to stupid and prejudiced conclusions, will always be available.  If the job can’t be done right, trust power to do it anyway.

  4. One of the first thoughts I had after it came out that the bombers were Caucasian but not crazy, right-wing white guys, is that no two terrorist attacks in this country are alike. At least in my lifetime and in this country. Maybe I’m stating common knowledge here, but it’s something easily forgotten.

  5. But the CIA needs taxpayer money to overthrow other countries’ democratically elected governments so that the resulting terrorists give the CIA a reason to get more taxpayer money.

    1. It sounds like you’re suggesting shutting down the CIA. That’s crazy. The CIA is more important to our national security than ever. Who else is going to identify overseas threats? Should we rely on our allies Pakistan and Afghanistan to keep us apprised?

      The CIA has been involved in some notorious incidents (usually with the approval of the US President), but they do critical intelligence work for us. Suggesting shutting them down is like suggesting shutting down the local police department because of some isolated corruption incidents.

      1. Given the insane culture apparently native to the CIA, shutting them down doesn’t sound so silly.

        No reason they can’t be replaced by a fresh organisation, starting from a clean sheet of paper.

          1.  Well, they could just replace them with a TSA that has heightened security clearance.

            /sarcasm off

          2. ‘No sense’?

            What is your standard of sense?

            If an organization is involved in corruption or abuse of power, breaking it up and starting over is one of the few effective ways of limiting the corruption.

        1. Would you really be willing to roll the dice and hope they don’t come up with something a hundred times worse?

      2. Every time someone says “Now more than ever” history kills a kitten.
        If you are seriously saying that the CIA is more important now than it was during the Cold War you might want to get your perspective checked.

        1. Suppose I’m wrong in that particular statement. The fact that the CIA was extremely important during the Cold War only serves to emphasize how important it is now, when the worldwide political landscape is much more complex.

          1. Do you think that terrorists are an existential threat, but the USSR with it’s thousands of nuclear weapons, and its large and highly capable conventional military forces wasn’t … ?

          2. Do you think they’re not? Who is the only foreign organization to kill more than 2,000 US citizens in one shot, on US soil, since 1941? Don’t they deserve careful surveillance? Who should handle it?

          3. Who is the only foreign organization to kill more than 2,000 US citizens in one shot, on US soil, since 1941?

            In other words, who is the only true, true, true, true, true Scotsman.

          4. “Do you think they’re not?”

            Rofl. Are you serious? Of course terrorists aren’t an existential threat.

            Most of the time they can’t find their arse with both hands. *Occasionally* one or two of them get extraordinarily lucky, but luck isn’t a viable course of action. And even when they DO get lucky, the amount of damage they cause is trivial (yes,in the grand scheme of things even 9/11 caused trivial amounts of damage).

            By the by, you’re in the midst of weaving yourself a mighty strawman. The question is not really “who” should handle it, but “how.” It’s the same three letters, but they lead to VERY different conversations.

          5. Who is the only foreign organization to kill more than 2,000 US citizens in one shot, on US soil, since 1941? 

            How did the CIA do at predicting and preventing that?  The fall of the Berlin Wall?

            How much blowback (including your example above) can be attributed at least partially to CIA actions?

            It’s arguable whether the CIA has caused more damage to the US than it’s prevented.

          6. Actually, the CIA did an unbelievably good job of predicting an attack by Al-qaeda, including a prediction that groups within the US would be trying to hijack aircraft from US airports. Their reports stated that the threat was not in the distant future, but imminent. The agency presented multiple reports expressing these concerns to President George W. Bush in August of 2001. On August 6, 2001, Bush received a briefing from the CIA titled, “Bin Laden Determined to Strike in US.” One month later, they were proven correct.


          7.  The world political landscape is always been very complex. To wish for otherwise is not a good idea.

      3. Who else is going to identify overseas threats?

        Overseas threats? Seriously? I’m not clear that such things really exist. Or if they do, they’re sufficiently rare to be discounted.

          1.  There is a fairly compelling theory that the greatest threat to peace currently is the US. So, yeah. The major threats to the US do come from inside North America.

          2. ahh yes, “peace”…it was all over the world before the US was an entity.  Nice flag.

          3.  “The major threats to the US do come from inside North America.”
            Mexico and Canada (in unison): Keep us out of this!

          4. So you’re going to (not) answer my questions with a question? What threats?

          5. I’ll see your snark and raise it: Among the literate, the device I used is called a, “rhetorical question.”

          6. > the device I used is called a, “rhetorical question.”

            In other words, you’re answering a question with another question that no one is expected to answer. In your head, what does the next reply you expect to receive look like?

          7. I thought your questions (“Overseas threats? Seriously?”) were rhetorical, because my comments above make my opinions clear.

            Your new question is, “What threats?” Beyond the obvious in-the-news things, I can’t name any. I’m not in a position to know. But that emphasizes my point – the primary reason to have international intelligence/surveillance is to discover and keep track of them for us. For discarding our agencies to make sense we need to know that: (a) there are no threats, and (b) there never will be any in the future.

          8. I guess I could make the argument a little differently

            You haven’t made any argument at all.

          9. No, quite seriously. You’ve made a bunch of assertions, but absolutely no argument.



      4. I’m suggesting the CIA causes more problems than it solves. They need to be rebooted.

          1. Seriously?  You’re being sarcastic, aren’t you?  I’m not trying to gang up on you, but you basically claimed you’re pro-CIA, and now you’re saying you’re for a reboot of a clearly-corrupted department? I think you’re just trolling this thread, man.

          2. As I said above in one of my first comments, “Shutting it down and starting ab initio makes no sense. Fix what’s there.” I think that’s pretty much what Mark says. And yes, I was being sarcastic. I obviously didn’t change Mark’s mind.

    2.  The CIA is better in concept than in practice.  There is the usual whackdoodles saving the world for Jesus, there are bribe taking prostitute loving appendages of the defense industry, there’s the bureaucrats having turf battles, there’s people doing stuff that’s very risky for the US and of questionable strategic value…

  6. Unfortunately, most random acts of terrorism can not be predicted or prevented. Any time security measures are put in place, they only serve to annoy the public. Someone determined to create mayhem will find a way. 

  7. I see what the post is saying.  Totally get all that.

    But, fwiw, in general, I think, if someone has a license plate that reads, like, “TERRORISTA#1,” you know, that might be sort of a tip-off.  Just a hunch.

      1.  I know some of my local roller-derby’s girls. They’re pretty fucking scary.

  8. The argument that is always trotted out to support grabbing more data is that by grabbing more data, they can winnow down to a manageable pool of suspicious activity, and rule out the innocent.

    The problem with this argument lies with the fact that every novel thing a terrorist / criminal does increases the “pool of suspicious activity”, and nothing reverses that.

    1.  so your argument is basically the same as the NRA’s against background checks. Why bother?

      1. Because gun ownership by criminals is a common and easily detected occurrence, whereas terrorist plans are an extremely rare and expensive to spot occurrence. 

        Threat assessment needs only basic probability skills – but most people who do it don’t have them.

Comments are closed.