Another media rant
I was listening to NPR yesterday on the way to work, so it would have been Morning Edition. Specifically, this story.
Listen to it, or read the transcript (the transcript, not the writeup at the link.) then come back.
So, this Marc Salem guy, the "pick a number" guy. I assumed he's a professional mentalist. (And, in fact, he is.)
Anyway, the routine he does where he asks Tovia to pick a number between 50 and 100. Wow! I thought. 50 possible choices! "Both digits even," he sort of mumbles quietly after the big flashy loud 50-100 thing. Oh. So really, there are only TEN actual choices: 60, 62, 64, 66, 68, 80, 82, 84, 86, and 88. Right off the bat, I think I can say that most people won't pick 60, 66, 80, or 88 - because they don't think of 0 as an even number, and 66 and 88 don't seem "random" enough. So already, no matter how he "reads" you, his chances are only a bit worse than one in six, and no worse than one in ten. Pretty good odds for "a number between 50 and 100."
Now read the writeup at the link. "[H]e begins to count as fast as he can from 50 - until he stops dead at 68." Wait, did they listen to the same interview I did? He stopped dead at 64, too. Maybe he has some additional reason to not consider 62 likely; I dunno. But I think there's a pretty obvious case of selective perception there on the part of the person who wrote the summary. And of course he can count as fast as he wants to from 50 - his own restrictions say he can spit out ten - maybe even twelve - numbers before there's even a chance he'll get a hit!
All in all, as a demonstration of how effective behavioral profiling is, this makes a good demonstration of why magic tricks aren't anything like as unconstrained as real life. (Which is to say, it's a pretty dismal demonstration of what it's supposed to be a demonstration of.)
So, the lesson I should take from this is that NPR sucks at finding people to make a reasonable case for behavioral profiling. The lesson I'm choosing to take from this instead is that we're putting our nation's security in the hands of stage magicians who don't even try very hard. (And note, it does say that he is "a consultant to law enforcement," so that's not an entirely unreasonable conclusion to jump to.)
Listen to it, or read the transcript (the transcript, not the writeup at the link.) then come back.
So, this Marc Salem guy, the "pick a number" guy. I assumed he's a professional mentalist. (And, in fact, he is.)
Anyway, the routine he does where he asks Tovia to pick a number between 50 and 100. Wow! I thought. 50 possible choices! "Both digits even," he sort of mumbles quietly after the big flashy loud 50-100 thing. Oh. So really, there are only TEN actual choices: 60, 62, 64, 66, 68, 80, 82, 84, 86, and 88. Right off the bat, I think I can say that most people won't pick 60, 66, 80, or 88 - because they don't think of 0 as an even number, and 66 and 88 don't seem "random" enough. So already, no matter how he "reads" you, his chances are only a bit worse than one in six, and no worse than one in ten. Pretty good odds for "a number between 50 and 100."
Now read the writeup at the link. "[H]e begins to count as fast as he can from 50 - until he stops dead at 68." Wait, did they listen to the same interview I did? He stopped dead at 64, too. Maybe he has some additional reason to not consider 62 likely; I dunno. But I think there's a pretty obvious case of selective perception there on the part of the person who wrote the summary. And of course he can count as fast as he wants to from 50 - his own restrictions say he can spit out ten - maybe even twelve - numbers before there's even a chance he'll get a hit!
All in all, as a demonstration of how effective behavioral profiling is, this makes a good demonstration of why magic tricks aren't anything like as unconstrained as real life. (Which is to say, it's a pretty dismal demonstration of what it's supposed to be a demonstration of.)
So, the lesson I should take from this is that NPR sucks at finding people to make a reasonable case for behavioral profiling. The lesson I'm choosing to take from this instead is that we're putting our nation's security in the hands of stage magicians who don't even try very hard. (And note, it does say that he is "a consultant to law enforcement," so that's not an entirely unreasonable conclusion to jump to.)