Join 3,432 readers in helping fund MetaFilter (Hide)


Probability senses tingling batman!
December 11, 2005 7:38 AM   Subscribe

Experts can suck at predicting the future. Their intuitive sense of probability is no more developed than lay-people's. A classic experiment is to present two indistinguishable choices are presented, but with unequal probability of reward. Humans look for complex patterns, which don't exist, and preform quite poorly. Rats quickly recognize the choice with higher probability, and preform optimally.
posted by jeffburdges (34 comments total)

 
Poor human "probability sense" could have played a major role in the development of religion, as well as more desirable human traits.
posted by jeffburdges at 7:38 AM on December 11, 2005


This was just posted a few days ago...
posted by sfts2 at 7:46 AM on December 11, 2005


List of Cognitive Biases.

Also, I could swear I've seen this somewhere before.
posted by Eideteker at 7:48 AM on December 11, 2005


back in 1981 there was a book published called "the people's almanac presents the book of predictions" ... i have a copy ... in it psychics, experts, science fiction authors and others make predictions of what the future will be like ... nearly 25 years later the vast majority of them are amusingly wrong

if you can find a copy of the paperback it makes for intriguing reading now ...
posted by pyramid termite at 7:57 AM on December 11, 2005


It's an interesting article, and since I didn't catch the previous post, I would have been sorry to have missed it.

(Particularly enjoyed all the reference to Tolstoy's hedgehogs and foxes; the realm of pundits, pontificators, predicters and the 'plain-spoken folks' outperforming them all seems ripe for Aesopean analogies.)
posted by Haruspex at 8:02 AM on December 11, 2005


...predictors

I mourn the loss of spell check.
posted by Haruspex at 8:04 AM on December 11, 2005


that's perdicters, this post is abowt the preformance of perdicters. and how they preform. i perdict jeffburdges will preform badlee in a spelung be.
posted by quonsar at 8:23 AM on December 11, 2005


procisely.
posted by Skygazer at 8:30 AM on December 11, 2005


oh poo.
posted by Haruspex at 8:58 AM on December 11, 2005


So what do the rats say about Peak Oil?
posted by jalexei at 8:59 AM on December 11, 2005


......which is that most of them are dealing in “solidarity” goods, not “credence” goods. Their analyses and predictions are tailored to make their ideological brethren feel good—more white swans for the white-swan camp. A prediction, in this context, is just an exclamation point added to an analysis. Liberals want to hear that whatever conservatives are up to is bound to go badly; when the argument gets more nuanced, they change the channel.

A great analysis of the state of punditry and political debate.

Also:

Metafilter: More white swans for the white swan camp.
posted by lalochezia at 9:01 AM on December 11, 2005


quonsar: Oh, it's even worse than you think: I use a Mac. :)
posted by jeffburdges at 9:06 AM on December 11, 2005


It is possible to predict the past with 100% accuracy.
posted by Rothko at 9:09 AM on December 11, 2005


Pyramid Termite: I have the People's Almanac book, as well as a lot of other old books that try to say what'll happen in the next year, decade or century. Whether they're relying on fortune telling (Old Moore's Almanac), statistics (the Economist Guide to the Year N), or science/science fiction insights (too many to name), or talking about the short, medium or long term... nobody gets it even close to right. Or, as far as I can tell, ever has.
posted by Hogshead at 9:20 AM on December 11, 2005


Had I been a hedgehog I would have dismissed the article due to its confusing FPP. Instead I am a fox and read it and found it to be quite interesting.
posted by Slothrop at 9:38 AM on December 11, 2005


it certainly worked well for george orwell.
posted by rodney stewart at 9:39 AM on December 11, 2005


that's perdicters, this post is abowt the preformance of perdicters. and how they preform. i perdict jeffburdges will preform badlee in a spelung be.

He probably read the article posted just before this. It's about Merrian-Webster's open dictionary. Merrian hated Merriam, and made a dictionary of misspelled words to spite him.

spellchecks
posted by gunthersghost at 10:13 AM on December 11, 2005


If anyone can predict the future with any reliability, they're not writing books about it for pennies a word.
posted by Richard Daly at 10:15 AM on December 11, 2005


"Can't see the forest for the trees" said that quite simply.
posted by HTuttle at 10:20 AM on December 11, 2005


"Sometimes a cigar is just a cigar." - Freud
posted by clevershark at 10:46 AM on December 11, 2005


Double. I still haven't gotten around to reading it yet...
posted by Chuckles at 11:31 AM on December 11, 2005


Rothko already spotted it, sorry... Vague link text is evil!
posted by Chuckles at 11:32 AM on December 11, 2005


I am the black swan. Therefore your view is wrong, you just haven't realized it yet.
posted by mystyk at 11:53 AM on December 11, 2005


Indeed. The tags on this one are singular, the tags on the old one plural. There's only one overlap.
posted by Eideteker at 11:56 AM on December 11, 2005


I predict I'll have another cup of coffee.



Hey! I'm right!
Me: 1
Experts: 0
posted by The Great Big Mulp at 12:11 PM on December 11, 2005


This deserves to be a double, because most people need to see the black swan at least twice before they'll believe it.

I hope this article changes the way I think, but being an expert on myself, I predict it won't.
posted by Citizen Premier at 12:37 PM on December 11, 2005


By the way, I'm emailing this to my dad, because I predict that he'll feel bad for sending me a bunch of low-grade jokes...
posted by Citizen Premier at 12:38 PM on December 11, 2005


Citizen Premier: "This deserves to be a double, because most people need to see the black swan at least twice before they'll believe it."

Damn, and I'm still recovering from the last time I saw the 'white rabbit.' Now that was some crazy shit.
posted by mystyk at 12:57 PM on December 11, 2005


Also, Amazon link, just in case you're thinking of buying it.
posted by mystyk at 1:01 PM on December 11, 2005


Too bad we arn't wikifilter, we could just merge the threads. I just realized the author passed up a chance to compare us to feces-throwing monkeys, hate to see an oppertunity like that go by unexploited.
posted by jeffburdges at 1:24 PM on December 11, 2005


This was just on NPR's "On The Media" this afternoon (mp3 link)...at least this afternoon on WBUR.
posted by tpl1212 at 2:31 PM on December 11, 2005


I think the wrong criteria are being used, that is, "experts" vs. "lay people". Properly, shouldn't it be "people who predict the future" vs. "people who don't"?

Whose predictions of future events would be more reliable? An "expert" historian; an "expert" in current events, such as a news reporter; or someone whose business it is to forecast the future, such as an "expert" stock-market analyst?

The historian relies on past performance to make his prediction. The news reporter relies on "linkages" between news items. However, the stock-market analyst not only relies on past performance and current indicators, but also has a hundred different trend indicators; and his livelihood depends on the accuracy of his predictions.
posted by kablam at 2:58 PM on December 11, 2005


It is possible to predict the past with 100% accuracy.

So then, whatever happened to Jimmy Hoffa?
posted by JHarris at 5:22 PM on December 11, 2005


Basically, if you only gleem one lesson from this post, it's not to spend hundreds to get thoroughly researched sports predictions.
posted by Citizen Premier at 7:01 PM on December 11, 2005


« Older Merrian-Webster open dictionary...  |  He is “fine” by a certain very... Newer »


This thread has been archived and is closed to new comments