Streams

Episode #601

Stochasticity

« previous episode | next episode »

Monday, June 15, 2009

Blobs of light Blobs of light (M I T C H Ǝ L L/flickr/CC-BY-2.0)

Stochasticity (a wonderfully slippery and smarty-pants word for randomness), may be at the very foundation of our lives. To understand how big a role it plays, we look at chance and patterns in sports, lottery tickets, and even the cells in our own body.

Along the way, we talk to a woman suddenly consumed by a frenzied gambling addiction, meet two friends whose meeting seems to defy pure chance, and take a close look at some very noisy bacteria.

Guests:

Laura Buxton, Ann Klinestiver, Jay Koehler, Little Wing Lee, Jonah Lehrer, Deborah Nolan and Carl Zimmer

A Very Lucky Wind

Laura Buxton, an English girl just shy of ten years old, didn't realize the strange course her life would take after her red balloon was swept away into the sky. It drifted south over England, bearing a small label that said, "Please send back to Laura Buxton." What happened next ...

Comments [29]

Seeking Patterns

Fine. Randomness may govern the world around us, but does it guide US?? Jonah Lehrer joins us to examine one of the most skilled basketball teams ever, the '82 - '83 '76ers, and wonders whether or not the mythical "hot hand" actually exists.

Then we ...

Comments [7]

Random Rules

The business of life is the business of taking the disorder of the world, little bits of matter scattered here and there, and giving them shape, regularity, order. Or so we all thought. Science journalist Carl Zimmer claims that at a microscopic level, in the inner workings of cells, ...

Comments [13]

News, weather, Radiolab, Brian Lehrer and more.
Get the best of WNYC in your inbox, every morning.

Comments [150]

DC from Baltimore

Love the show, and love this episode, but Radiolab should really divest any Jonah Lehrer stock they have left. I understand the desire to not reshuffle old episodes, and that you guys have fact checked and revised accordingly, but Lehrer is a toxic asset that throws the veracity of anything he touches into question. It's a distraction that taints this episode and really any feature he's had anything to do with

Nov. 08 2013 03:28 PM
_#I am Super Patatoe_ from Oklahoma

How do the two Lauras interpret the coincidences in their lives?

Nov. 08 2013 11:11 AM
ankit singh

really amazing must listen story :)

Oct. 21 2013 09:31 AM
Jennifer

Curtos, the music around 27:00 is "Marilyn Set Me Free" by Casino Versus Japan. Great pondering-and-revelation music. :D

Jul. 17 2013 05:13 PM
Terrible English person

I loved this episode....but I'm going to say something wanky on the basis that no one else has commented on it. I think you may have got English girl 1's voice and English girl 2's voice mixed up at some point. There's a point where you talk about Northern Laura Buxton and the Southern Laura Buxton speaks.

Kind of interesting that you found it difficult to tell the difference between their accents, though, because one has a very upper class southern accent and the other a regional northern accent. I wonder how many American accents I can't hear...

May. 14 2013 06:42 AM
Evan G.

A great book on this topic: The Taming of Chance by Ian Hacking

May. 04 2013 10:16 AM
Yanni

As soon as I heard the math professor say the probability of getting a streak of 7 in the 100-coin toss is 1/6, I thought it must be higher than that, because he was overlooking something in his analysis.

I admit I wasn't able to derive the probability mathematically (I'd love to see a correct analysis), but writing a simulation is easy. Here is what I get with a simulation:

Probability of getting at least 7 Tails in a row, out of 100 tosses:
0.318

Probability of getting any 7 in a row (heads or tails), out of 100 tosses:
0.542

May. 01 2013 12:02 AM
Ted Fridirici from Harrisburg, Pa

Either I dont get it, which is quite possible, or their logic regarding a "hot" basketball player is flawed. Their flaw seems to me to be that they only consider what a players field goal percentage (FGP) was or has been in the past....through his career so far ....and not what his final FGP is when his career is over. While its quite possible it would stay about the same throughout his career he could also get a new shooting coach or trainer or a new guru and make a much larger percentage of shots in the latter part of his career. This would result in a much higher FGP than what they are assuming or basing their discussion on. They are looking to the past to predict what he might do in the future. They cant know if he will be a better shooter from the point forward from which a discussion of his "hotness" ensues ...or worse ....or about the same. Taken to the extreme consider this, a lifetime FGP of 50% could be made up of 10,000 missed shots in a row followed by 10,000 made shots in a row. If you pick up the discussion after he's missed his 9999th shot what are you likely to predict? You cant know because you cant predict the future. Is he or will he be hot or not? If you knew his final FGP at the end of his career would be 50% you would say he is or will soon be VERY "hot". If you instead somehow knew his FGP when he ended his career would be 0% well then you would know he was not. If you knew it would be 75% than you know he is going to make quite a few more than he misses FROM THIS POINT IN TIME FORWARD and doing that over a given period is called being "hot".

Ok so what am I missing?

Apr. 29 2013 07:20 AM
Lee from Greensboro

If you want to see a great example of order from chaos, take a look at fractals. This also works in reverse as well. Have an ordered set of rules and see the chaos that arises from it. Look up "The Chaos Game" under fractals and you will see how "Random" acts can create the same ordered figure every time.

Apr. 28 2013 03:13 PM
Susan from Oregon

I enjoyed this show and loved the format: catchy enough for driving, interesting enough to engage. I found myself thinking about how a poor understanding of stochasticity seems related to the logical fallacy of attribution error. If attribution error causes me to think that something about "me" caused a success of failure, or caused me personally to succeed or fail, I'll never see that many things and people contributed to the outcome. "The race is not the swift, nor the battle to the strong, but time and chance happen to them all."

I wonder if the need to control people and things drives the creation of organizational systems--pedagogy, project management, or any system of practice--that set people up to fail by setting them up to believe. But I'm not sure what would happen to a life if the person lacked any ability to believe in something. The irrational conviction of personal value can do good things.

Apr. 27 2013 05:14 PM

:D

Apr. 26 2013 11:21 PM

This was beautifully crafted, RL. Keep it up! (:

Apr. 26 2013 11:12 PM
Wil Davis from Nashua, NH

I just had to switch this garbage crap programme off! So much rubbish plastered together with intrusive audio junk, absolutely awful production values - nothing has changed! I listened to one of these dreadful productions a while ago, and though much the same thing about the dreadful production; the ideas are quite good and the subjects are interesting, but the method used to get the point over is quite hideously overdone! - Wil Davis

Apr. 26 2013 12:18 PM
ajumajuj from New York

This was another case in which all the pizzaz and clever editing and sound effects got in the way of the science - the analogies led away from the subject and not towards it. though you have a fantastic choice of subjects, the general way you cover scientific topics is scattered and hyper, which may be fun to listen to but fails to actually convey what the ideas are.

Apr. 25 2013 10:30 PM

That's just a fantastic hour.. Has to be my favourite so far..
Stochasticity has serendipitously become my word of the day and this podcast is undoubtedly the highlight of it..

You guys are geniuses... RadioLab is just awesome..

Mar. 22 2013 04:35 PM
Ryan from NYC

Does anybody know the song that plays from 5:40 to 6:50? Sounds like it could just be something Jad threw together, but I'm hoping it's a piece from an available song!

Aug. 15 2012 02:19 PM
Mary Jo Marchnight from Savannah, GA

I suggest you and your audience would benefit from a study of Dr. Paul Kammerer (the most famous biologist in the world in 1920), particularly in this case his scientific studies of 'seriality'. He wrote a book to prove that what we call coincidence is in reality a manifestation of a universal principle in nature which operates independently from the known laws of physical causation. And because we are trained to ignore 'coincidence' in general, we miss manifestations which would stare us in the face if we were conscious coincidence-collectors. Arthur Koestler's "The Case of the Midwife Toad" is about Kammerer; and I have a wonderful book of Kammerer's on the subject of 'The Inheritance of Acquired Characteristics". He is a case of someone whose reputation was destroyed and needs to be reinstated for the benefit of science.

Jul. 22 2012 02:55 PM
Bob

When I heard the piece on the lady with Parkinson's Disease taking Requip and gambling her life away I saw myself 15 years ago, when I was taking Requip and happily gambling my families future away. Had it not been for my strong willed wife the nearly $20,000 I burned through very easily could have gone into the hundreds of thousands. I was diagnosed with young onset pd at 28, I am 52 now. I remember doing research at the time and coming across studies linking Mirapex with gambling (as well as other compulsive behaviors). There was talk of a lawsuit against the maker of Mirapex. Parkinson's Disease is an endless battle for the right balance in meds. It aint fun, but life is still sweet.

Jul. 21 2012 06:57 PM
Boris Pro from Sacramaneto

Enjoyed Stochasticity but riddle me this Radiolab. In the case of the NBA streak shooter, what of "serial correlation", when draws across time are correlated not independent? Did your statician test for this? Interestingly, your gambler with the dopamine problem provides a physical process that might explain serial correlation in NBA data. A shooter that is hot may have a chemiacal activity that enhances brain functioning for a period of time till the reservoir of that chemical is depleted or neourons require higher levels to be stimulated. That was Kofi from NY's point. Run a statitical test to prove independence/post & will give you $50. I am too lazy to doit.

Jul. 18 2012 04:00 PM
Manhattan Blues from NYC

Thanks for the heads up on this old episode. I couldn't have found it without you. Okay. Okay. I guess I'm spoiled by This American Life's putting up new episodes almost once a week. Perhaps finding interesting science stories to write about is more difficult than my small, and cynical mind, can fathom.

Jul. 18 2012 06:39 AM

Today's LA Times has an interesting article about brain biology and behavior which I think could be the subject of a future program--http://www.latimes.com/news/opinion/commentary/la-oe-sapolsky-brain-and-behavior-20120715,0,5186378.story

Jul. 15 2012 05:43 PM
Darlene from COLORADO

THE SETH MATERIAL BY JANE ROBERTS, PUBLISHED IN 1970, TALKS ABOUT ELECTROMAGNETIC ENERGY THAT FORMS BASIC EXPERIENCE, IS ORGANIZED, MAINTAINED BY ALL THE EFFORTS OF OUR INNER SELVES. IN FACT THIS
CHAOTIC ENERGY MAKES OUR VERY AIR. I'D BE INTERESTED TO KNOW IF ANYONE OUT THERE HAS READ SETH OR WILL READ THIS COMMENT.

Jul. 15 2012 05:13 PM
chadrah from Ojai, CA.

Sorry, "Can someone"

Jul. 15 2012 02:21 PM
chadrah from Ojai, CA.

Does someone tell me the name of the music during the interlude from the first story, 'a very lucky wind' ?

Thanks

Jul. 15 2012 02:14 PM
Miriam Brooks from Boston

I was very disturbed by your story about Ann, the Parkinson's patient. My mother has been on Requip for many years with no side effects. I think it's irresponsible of you not to address the fact that 1) no one should stop taking their meds without consulting their doctor and 2) there are other medications very effective for Parkinson's. If her neurologist took her off the Requip without putting her on another drug or combination of drugs, then I suggest she might want to look for another neurologist. Your story may have caused many people who are currently taking Requip to panic unnecessarily. You left out a large portion of the story, and could have caused many people serious harm.

Jul. 14 2012 05:29 PM
julianadoremi from Boston area

I agree with Sara. Was this a breech of medical ethics, or an oversight? Now that we know this, can her neurologist consider another option, or get her some help for her gambling? That's a pretty big lifestyle changer to simply ignore.

Jul. 14 2012 03:54 PM
Sara Oaklander from Boston, MA

I wanted to scream when I heard the story about the Parkinson's drug, suggesting that the gambling addiction was a known side effect. How could her physician not have monitored her for side effects and detected that her life was in shambles - possible as a result of this medication? I would go so far as to suggest that her physician - and the drug company - as well should be held responsible - at least in part - for her addiction. I was taken aback that you did not include some discussion of this in the story, as well. I kept waiting - at the end - for this to be discussed. But it was not.

Jul. 14 2012 03:47 PM
Rob Shepard from Roanoke, VA

Stochastocity, a lucky wind, for years I have called these events examples of cosmic synchronistic ridiculosity, CSRs! Have had my fair share. Thanks for your wonderful show. Rob

Jul. 14 2012 02:18 PM
Drsnap

what a great episode! I loved that one!

May. 21 2012 10:09 PM
erica from Milwaukee, Wisconsin

How is it that Joe Damaggio is the only outlier mentioned when Willie Keeler had 45 consecutive hits, Pete Rose had 44, and Bill Dahlen had 42? Also Nolan Ryan had 383 strikeous in a single season and Sand Doufax had 382. Are they also outliers? Eric Gagne had 84 consecutive save chances converted...wouldn't he also be an outlier? Though Joe was a great ballplayer, there are many more athletes that did amazing feats that could be outliers. Joe played for roughly 13 short seasons...would that factor into this equation?

Apr. 30 2012 03:30 PM
Patrick

This was one of my first episodes I heard and is still one of my favorites. I found the golf ball analogy and its implications fascinating. At one point in the cast, a guest mentions that the only sport streak that defied statistical expectations was Joe Dimaggio's hitting streak. Does anyone know if there are actually others or if anyone's done a thorough analysis of sport records/streaks looking at significant stats?

Apr. 25 2012 12:10 AM
Ned from Warwick ny

Please have pity and identify that big band clarinet at 7:24!
Tremendous show, team.

Apr. 17 2012 09:59 PM
C. from Toronto

True story, when you guys mentioned his phone ringing, my roommate's cell started ringing.

Dec. 04 2011 11:28 PM
Shaz from Canada

Whoa whoa whoa....did I just hear you say hockey is the sport that no ones cares about?! I'm not a huge fan but I am enough of one to know that you are grossly incorrect. Come down to the Red Mile and say that during playoffs and see if you make it out alive.

Oct. 13 2011 09:36 PM
Emily from Hawaii

The song is Simon and Garfunkel's "Mrs. Robinson,' the famous tune used in "The Graduate."

Sep. 26 2011 02:40 AM
Curtos

I was just wondering if anybody new the name of the song or sample used toward the end of the seeking patterns section, when Jad is lamenting over his realisation that being hot is often not quite as improbable or amazing as it seems. The music that plays before the Joe Dimaggio song? It occurs around 27:00 into the episode.

Jun. 13 2011 11:33 PM
stig from norway

A good example of order coming out of chaos is slot machines in a casino. I remember once being in a casino in Atlantic city and being really facinated by the noise collectively being made by all the slot machines. Now all the slot machines have sounds that play when certain buttons are pushed or levers are pulled and the timing of those sounds are pretty random, yet when you put all of these slot machines together in one big room (as was the case in this casino) all the random noise combine into a super tone at a spesific pitch. It was a pretty surreal thing to hear.

Jun. 03 2011 04:45 AM
Stephanie

@William L: I agree. My favorite story: I used to be a proofreader and at one point I was correcting typos in "random number tables." How weird was that?!

Mar. 22 2011 10:27 PM
Jol from Toronto

When I was ten I watched a program about probability on TV and one story really struck me, and I am kind of disappointed d you didn't feature it, because it is a good example of chance. I don't remember all the facts, but here is how the story goes.

One night, in an office in London, a woman was working late on a project, and she needed to print the papers. However, during somewhere during the process the printer stopped working. She tried to fix it, but it was hopeless. What she did then, is she looked up the information of a technician who usually fixed such things. I don't remember his name, but let's call him Jay. So, she looked in the book where all of the information of the employees is held, and besides Jay's name she saw a number. Again, this was at night, but she really needed him, because the project was very important. So, then she dials the number she saw besides Jay's name, and few seconds later, he picks up the phone. She says "Hi Jay, the printer isn't working I wonder if you could fix it?". And Jay replied "How did you find me?". "What do you mean? I dialed your phone number?" "I am not at home, I am in East London, talking on at a telephone booth."

To clarify, what the woman saw wasn't Jay's phone number it was his office ID, you know the one that he uses for identification. However, when she dialed it, totally by chance, from the first try, she called a random telephone booth in East London. That night Jay was walking with his girlfriend beside that telephone booth, and just for curiosity picked up the telephone. Imagine how astonished he was when she talked to him?. This was in the 80s, when there were no cell phones, in a city the size of London.

What are the odds of that? I hope you will find that story, because I was looking for it everywhere and wasn't able to find it. I hope you understood what I had to say, because English isn't my native language.

Good show!

Mar. 22 2011 12:59 PM
LAH from Chicago

So my friend and I spent a long time arguing about this, and here's our final conclusion on the coin toss probability to add to the rest of the discussion here. My tool: MATLAB, since I'm not a probability genius.

- A run of 7 heads or tails has a 1.56% chance of happening.
- In 100 consecutive coin tosses, you will get runs of 7 or longer about 72% of the time.
- But if you count the runs longer than 7 as multiple runs of 7 (a run of 8 is two runs of 7, a run of 9 is three runs of 7), then you'll get "runs of 7" about 144% of the time. (i.e. In 200 consecutive coin tosses, you'll get roughly three "runs of 7.")

So it seems a pretty good estimate that in 100 consecutive coin tosses you'll get:
- 100*(0.5^n) runs of n or longer
- 100*(0.5^(n-1)) "runs of n"
And it's an overestimate because, as mentioned before, for a run of n, you only have 100-(n-1) slots instead of 100 slots.

Love the show!

Mar. 07 2011 11:53 PM
Livingston from New York

http://www.youtube.com/watch?v=VEUVL0U7QyY

this episode reminded me of this scene from American Splendor

Jan. 19 2011 11:54 AM
Kofi Asante from New York

All I got to say is you guys couldn't be furthest from the truth about the whole basketball thing. It is purely skill when it comes to shooting a basketball. I think your problem was that you fell victim to hubris. You did admit that you are a basketball fan, so I can understand. A basketball player knows what it entails to make a shot. I've had several experiences when I was locked in and couldn't miss a shot, and I've realized that the reason for that was because I was totally focused at every single shot I attempted. I've also realized when I miss shots that it's because I wasn't totally focused, mainly because of tiredness; not following through, the ball not coming off my fingers correctly because of lapse of focus, not fully or properly jumping, etc. My point being that reasons why real basketball players miss or make shots is not because of chance, but rather tiredness and skill. Now I understand some people take wild shots and sometimes it goes in, but if you're skilled enough, those shots are not necessary.

Jan. 04 2011 02:01 AM
Lowell

It strikes me very strongly that this question of how order emerges from chaos is key in a number of ways. Learning how noisy we are at the molecular level, and the predictable regularness that comes from that, I immediately began thinking about quantum and classical physics. We still don't know why/how particles that behave insanely, randomly appearing, disappearing, etc. can beget a larger universe as ordered as a clock. It sounds like the same issue-- any thoughts?

Dec. 29 2010 04:06 PM

Great Episode as usual. I tried the coin-flip experiment with my 11th grade math class and it worked. Every case we were able to determine if the student had actually flipped the coin or pretended to flip it.

I did modify the activity. I started by giving every student a note card. I told them to imagine that you are flipping a coin 100 times, and record the results on the card. I told them to make it look as realistic as possible. On the back of the first card I told them to write "pretend". Then, when everyone was done, I passed out a coin to each student and I told them to repeat it, but this time, actually use a coin. One the back of the second card, I told them to write "actual". When they were done, I displayed the cards on the screen using a document camera, and once they got the hang of it, they were able to guess with 100% accuracy which cards were the pretend cards and which ones were the actual cards.

Dec. 17 2010 03:44 PM
Jordan from Brookline, MA

I love when Robert Krulwich doesn't believe something and Jad is telling him how true it is, or vice versa, and they both start chuckling.

Oct. 26 2010 10:20 AM
stay

i had to read this for a science class and i thought the things were interesting

Oct. 14 2010 09:16 PM
prasun from Montreal

Shouldn't the Stochasticity song at the end of the episode give some credit to the brillaint Bob Dorough and his song Electricity?
http://www.schoolhouserock.tv/Electricity.html
Excellent show by the way.

Sep. 08 2010 06:50 PM
Ant

Surely stochastic implies probability of next thing depends on present state, whereas random implies probability of next thing is independent of present state? This show seemed to equate the two. Otherwise fun as usual.

Jul. 17 2010 02:49 PM
Tim M

I wrote a script to test the odds 7 coin flips in a row. After millions of these 100 flip trials, my script estimates this has a chance of 56.90%.

Out of curiosity I tested the odds of other streaks of happening. here is the odds from my program:

5 in a row: 97.58%
6 in a row: 81.64%
7 in a row: 56.90%
8 in a row: 32.66%
9 in a row: 13.71%
10 in a row: 5.35%
11 in a row: 3.03%
12 in a row: 0.61%

Mar. 05 2010 01:53 AM
ASA kaplan

Running the numbers and the chances for the lottery with one in ten million and six billion people are 54% for once and 1in a hundred thousand twice for anyone

Jan. 03 2010 05:29 PM
Jesse

Izzy, I suspect that few of the scientists and statisticians interviewed for the story began their careers because they "enjoy proving others wrong." I would guess that most of them did so because understanding the world better is a rewarding experience. Also, enjoying "the thought of complete randomness in our live" does not make that thought real or accurate. Finally, I fail to see how any result involving coin flips or basketball shot percentages or any of the other examples in the story affects your ability to come up with non-sequiturs for your various classes or conversations.

Nov. 11 2009 10:03 PM
izzy carter

not only was i forced to listen to this podcast.....i had to do a project on it... yes i got an A but while i was listening i realized....some of us enjoy the thought of complete randomness in our lives...and some people enjoy proving the others wrong. scientists and anyone else who likes proving that complete randomness isn't really random. i personally enjoy randomness in my life....i do random things all the time. i say random things in the middle of class or in a conversation. we can be talking about maybe...stars in the sky...i'll come out and say something about food or cows anything!

i guess what i'm trying to say is enjoy and enhance the randomness in your life anyway you can cause at any point in time...you could be gone...

Nov. 11 2009 08:14 PM
George

Fantastic show as usual, but there was an error at the beginning. You defined Stochasticity as a word that comes from Latin. That is incorrect. the word has a Greek derivation.(see link)

Oct. 27 2009 11:51 PM
Tammy

Found it interesting that this is my first visit to radiolab (thanks to link from Seth Godin's blog), and after scrolling through the podcasts...the one I 'chose' to listen to first was Stochasticity....hmmm, random?
Great show...time to download a few of these.

Oct. 24 2009 01:19 PM
Kimberly Zepeda

I thought the whole laura thing was so wild. This podcast on randomness is really interesting and fun to listen to. I didnt really ever think about how many random accurences happen to us. The classroom coin toss was pretty remarkable and made alot of sense.

Oct. 24 2009 01:20 AM
Lynden Price

Hot hand I truly believe in this , I believe that a person tends to have an amazing streek in sports. Many atheletes tend to have their day of unexpected moments and hot hand is one of those times .

Oct. 22 2009 03:02 AM
Lynden Price

This was very intense, must check out !!!

Oct. 20 2009 02:48 AM
Irene Cardenas

Unlike coins, people gain awareness of their interconnections when survival is on the line. Research into psychic connections proves this. In Entangled Minds, Dean Radin describes Thomas Young's double-slit experiment. In it, a photon is shot through a screen with one slit open. Behind the screen, a photographic plate records a particle. With two slits open, the plate records a wave. Each photon acts entangled with itself, going through both slits.

If the experimenter decides to close the second slit after the photon has already passed through, it still records a particle. If he decides to open the second slit after the photon has passed through, it still records a wave. The photon responds to the experimenter's later decision (a mental phenomenon), or becomes entangled with time. Dynamic things act like static objects if people choose to view them as such, i.e. through fixed definitions, prejudices or stereotypes. Consciousness affects them.

Oct. 13 2009 10:09 PM
Gal Haspel

ION CHANNELS!!!!
You missed a great opportunity here to sum it all by speaking about how science did solve this problem in another system. By "this problem" I mean Carl talking about cells being stochastic all the way down there but working all the way up here. Ion channels are proteins that let ions go into cells (most interestingly for me: neurons). They are individually stochastic but because the actual probability for a channel to be open can change and that there are MANY channels in every cell, the common activity is predictable!
One cannot predict a coin (even a weighted coin will fall on the "wrong" side sometimes) but can easily predict 10000 coins tossed at once.

Sep. 18 2009 01:51 PM
Johan von Boisman

The great British mentalist Derren Brown makes uses of the coin toss sequence in one of his shows.

http://www.youtube.com/watch?v=X1uJD1O3L08

Note that you have to watch the entire program to hear him explain it.

Aug. 21 2009 07:13 PM
Aaron

I was recently browsing through some Yahoo! news articles, and one jumped out at me. It was an article about flipping coins... 51/49 odds... blah, blah. What caught my attention was a paragraph in the article explaining that a deck of cards must be shuffled seven times to be "truly random."

Link to the cards article:
http://www.dartmouth.edu/~chance/course/topics/winning_number.html

What struck me as odd about this proposition is that there is no point when a deck all of a sudden becomes random. Now, if they're assuming a new deck, then sure: seven times would be great. They're not, though. They say that every time a deck is shuffled, it must be done seven times to assure randomness. Here's a quote from the article: "...but seven shuffles is a transition point, the first time
that randomness is close."

Ponder that, and see if you agree with me that this article might be a little off in its definition of "random."

Aug. 03 2009 07:32 PM
aram

The "14 chances" argument in the original podcast is wrong, but so is the "94 changes" argument in the corrections. There are indeed 94 chances, but they're not independent: if flips 1-7 are all the same, then flips 2-8 have probability 1/2 of being all the same. The exact calculation is not so easy.

Simulating with matlab gives roughly 32%. Here's the code:

clear a;
for i=1:10000;
a(i) = max(conv(floor(rand(1,100)*2),ones(1,7))); end;

sum(a==7)/10000

Jul. 25 2009 02:06 PM
Spike Dunn

Ye-arg! One of my favoritest podcast doing one of my favoritest subjects and you didn’t talk to my favoritest author on the subject: Nassim Nicholas Taleb. “Fooled By Randomness” and “The Black Swan” were my introduction to the questions of stochasticity. And he’s a very energetic interviewee. Fun missed, is all I’m saying. Still a really good bit though.

Jul. 21 2009 11:44 AM
Amy McCracken

One month later and the Stochasticity song is a persistent earworm for both me and my husband! Interesting and wonderful episode. Keep up the good work -- wouldn't mind more songs, too.

Jul. 19 2009 09:18 PM
Aaron

That's an awfully odd pingback....

Jul. 15 2009 10:25 PM
William L

I would say congratulations, but I wouldn’t want to appear too sycophantic or hackneyed, despite the fact that I do immensely enjoy Radiolab, and I do wish you well (as long as you continue to make shows that is ; ). I have enjoyed every show that you have aired, despite whether I have agreed with all of it or not; however, this last show, which I also enjoyed, I cannot restrain myself from commenting on. Randomness is a subject I know a little more than something about. I have taught a course or two on complexity and chaos theory. The show was fascinating as always, but I have to say that your last show talked about lions, but managed somehow to miss the zoo.

THE JANUS FACETS OF RANDOMNESS

The problem is there are two competing definitions of randomness. One common dictionary definition goes as follows. Randomness: proceeding, made or occurring without definite aim, reason, or pattern: the random selection of numbers. The other typical definition goes like this. Randomness: of or characterizing a process of selection in which each item of a set has an equal probability of being chosen. The last definition is from statistics and is, at least, well defined. The first definition is a mixture of the last and something else. Something ambiguous that seems to rely on the intentions of a human or the qualities of a human observer. The first definition is conflated with the second and refers to a frequently misunderstood phenomenon that is related to but not the same as the second. I will try not to digress too much as I attempt to disambiguate and clarify the phenomenon that is commonly understood as “randomness” extrinsic to the field of statistics. Statistics does not seek to understand randomness but merely to use it. Statistics is like engineering, which uses physics to build technology, but does not itself inform us as to how or why it works.

RANDOMNESS AND THE PERPLEXITY OF THE PERCEPTUAL MACHINE

Randomness is not a phenomenon independent of the observer. It is relative. Randomness is a property of a perceptual machine, a machine capable of recognizing patterns. Simply put, randomness is the category that all patterns not recognized by a perceptual machine fall into. If people cannot identify the ultimate cause of a pattern or predict the pattern, it is referred to as random. But one man’s, or computer’s, random is not equivalent to another man’s. In my past research I have found patterns in data that others have not. In that case, randomness depends on who you talk to. With the assistance of computers, many patterns that previously have been assigned to the random category have been re-categorized as deterministic. What changed in these instances was not the phenomenon observed, but the perceptual programming of the perceptual machine itself.

ILLUSIONS AND FLAWS IN THE PERCEPTUAL MACHINE

There is a reluctance, I believe, for some people and some scientists to accept the above definition based on the mistaken belief that what people perceive with their senses is truth, not perception, not relative, and certainly not the result of a process that could be broken down and eventually described like a machine’s. But this is exactly what neuroscientists, evolutionary biologists, psychologists, and people working in artificial intelligence have been doing for the past few decades. Donald D. Hoffman’s book “Visual Intelligence” chronicles many of the rules that our visual cortex uses to construct the reality that we perceive. It is by no means a passive process. We interact, interpret, and interpolate the data our senses accumulate. Hoffman shows the readers numerous examples of the failures of the rules we use to construct our visual perception, each of which has a corresponding illusion associated with them. He also uses compelling examples of certain brain damaged individuals to show what happens when the perceptual machines that we take for granted are damaged.
We need not limit our analysis to the errors of the visual perceptual machine. Our proprioception can be tricked into thinking we have a “phantom” limb when people have lost one of their limbs. Our auditory perception is regularly tricked into believing we hear “phantom” rings from our cell phones when enough background noise is present. People working in the video game industry and artificial intelligence research are busy emulating how our visual and auditory perceptual machines create the reality we perceive so as to create more realistic video games and more intelligent machines that can independently navigate the real world, respectively. People in these fields have no doubts that our brains contain perceptual machines that have a profound yet limited capacity to identify patterns.

THE HISTORY OF RANDOMNESS AND ITS COMPLEMENT: THE HISTORY OF SCIENCE

The history of randomness has developed simultaneously and in parallel to the history of science. At the boundary of our knowledge our explanations and predictions begin to fail and so bifurcate into two categories: stochastic or deterministic. The complementary nature of science and randomness is inevitable because scientists seek to identify, explain, and predict patterns in nature. When scientists have failed they have unwillingly contributed to the unsung history of randomness. We Scientists don’t like to admit that we don’t understand something so we find euphemisms instead. Randomness is just one of those euphemisms.
I will give two examples, out of a plethora of examples that I could give, of the hubris of some scientists regarding the denial of our collective failures and limits. In 1980 Stephen Hawking, famous in scientific circles for not being able to admit when he’s wrong, predicted that by the year 2000 we would have a “theory of everything.” Here we are nine years past his deadline with no theory in sight. His confusion has still not been rectified in that he just keeps postponing the date without recognizing that a universal theory can never be a theory of everything because we will never identify all the patterns in nature. The uncertainty principle and the existence of chaos cement randomness as a permanent relationship and physical effect of every measurement and therefore every perceptual machine. The term “dark matter” and “dark energy” are further examples of hubris. They are euphemistic postulations for what amounts to simple ignorance. We cannot see 96% of the universe and so instead of assuming that 96% of the universe is beyond our present ability to recognize, or understand completely, or that our theories about nature only account for 4% of the behavior of the known universe scientists actually incarnate and encapsulate our failure into what they imagine as two real physical phenomena. When the simplest explanation by far is that our theories simply do not possess the power to explain or predict all of what we observe. In other words our current theories like our perceptions are not always right. They are merely the best we’ve got. Most scientists suffer from an all consuming fear of having to say, “I don’t know" or worse yet, "I'm wrong."
Random number generators have come a long way. Their history is a microcosm of the history of randomness as a whole. That history also helps to clarify the distinction between “statistical randomness” and the randomness that is more commonly referred to in the vernacular, and that I am trying to disambiguate. Pseudorandom is the description given to a deterministic (predictable) process that produces statistical randomness. If that statistical randomness refers to an artificial process that produces a number, we call that process random number generation. Many of the random number generators of the past have had to be discarded because somebody found a means by which to predict their behavior. Linear congruential number generators used to be used until it was found that they deviated from perfect randomness in embarrassing ways. Other random number generators have had to eventually be discarded because somebody eventually divined some pattern in their results. Stephen Wolfram’s Mathematica uses cellular automata to generate its random numbers, which as yet no one has been able to identify a pattern in. So, random number generators produce randomness that is only considered random until someone can take its results, with the help of a computer to enhance one's perceptions, and identify a pattern. If somebody plots the results of the Lorenz equations, used to model convection in the atmosphere, in one dimension the results appear totally random to our eye, which is ultimately the perceptual machine that we usually use to identify order. If we plot the results in three dimensions, however, a chaotic attractor that appears something like a butterfly with circular wings appears as a beautiful demonstration of order emerging from what was apparently chaos.

THE CHALLENGE OF CHAOS AND COMPLEXITY

If our sensory perceptions are for the most part fixed then what, outside of our perceptual faculties, leads to what we call randomness? Why do some things appear random and others do not? As Stephen Wolfram points out randomness in an observed system can either come from the environment or it can be generated by the system itself. When Wolfram says this he is accidentally coupling complexity and our response to it, the perception of randomness. When uncertainty is continuously injected into a system from the environment the system will exhibit randomness, since the environment is typically far too complex for us to understand. Also when uncertainty about the initial conditions of certain known systems is extremely sensitive to initial conditions the system can exhibit random behavior, which is the scientific definition of chaos. The development of fingerprints on our hands and the coat patterns of animals like dogs and zebras are everyday examples of chaos. Twins and clones will never have the same fingerprints or coat patterns, respectively. Tiny variations in the initial conditions of development guarantee that no two fingerprints or zebras will ever look alike. Other systems, like computer generated cellular automata, which operate according to simple rules can themselves, from perfectly known initial conditions, generate enough complexity that we perceive their results as random. If one takes the last example to be what happens in nature without our knowledge then we can understand our relationship to the world as so. The universe operating according to simple rules, some of which it seems we know, can generate, nevertheless, complex patterns that confound both our enhanced (computer aided) perceptual faculties and the innate perceptual faculties that we were born with. When a pattern becomes too complex for us to understand we call it random.

UNIVERSAL-PSEUDO-RANDOMNESS

Pseudo-randomness is probably universal. There is no evidence to suggest that beneath all the things we don’t understand there isn’t a deterministic and orderly process that we cannot yet perceive an order to. Because of chaos theory and the uncertainty principle, we will never completely eliminate randomness. The common understanding of randomness is an ancient, obdurate, and anthropocentric theory, more pernicious than the idea that the Earth was the center of the universe, which posits that all the things we perceive with our senses are true and real. “Something appears random; therefore, it is random.” This naïve point of view imbues objects with a quality they do not in fact possess, which requires us to continuously adjust what we consider random and what we consider deterministic as our knowledge progresses. The old interpretation of randomness cannot admit the possibility of our minds own imperfections and weaknesses despite the fact that it would takes less than five dollars of materials and probably thirty seconds for me to perform an experiment that would prove to any rational person that something as sacrosanct as the very colors he or she has perceived his or her entire life do not actually exist. Our visual cortex constructs them as it does everything we see. Colors are merely a tool, in a repertoire of tools, which our visual perceptual machines use to understand our natural world.

Jul. 13 2009 03:27 AM
Steve

To toasterhead:

To answer your question: Yes, coin flips are totally random. To qualify myself, I study theoretical mechanics at Northwestern.

The thought is that, if we know all the forces on a body, we can exactly describe its motion (this would be nice for roulette!). So, Newtonian physics equations of motion should be able to form some 'equation' of the coin flip.

However, the exact state of a coin will never be known. If you remember from high school physics, the Heisenberg uncertainty principle states that one can never simultaneously know the momentum (velocity) and position of an electron. This means that many of your variables describing the state of the coin, e.g. surface texture, can never be completely accurately characterized down to its atomistic structure.

To look at it another way, one could never put the exact same initial conditions on a coin (height of drop, orientation, normal force between finger and coin, wind intensity, etc.) due to the same principle.

One other way to answer your question: There IS some equation to describe the motion of a coin IF we know all the variables you described. However, we can never know with 100% certainty those variables. Further, slight variations in each variable superpose and can make large changes in the result of the flip. Sadly the flip is random...don't go trying to hustle casinos (at least not that way).

Jul. 13 2009 12:15 AM
Chris

Awesome show.
I, too, have the same ring tone when I get new text messages. I think it's from a DJ Shadow song?
Thanks for the good work, fellas.

Jul. 12 2009 06:05 AM
Aaron

Way up above, someone mentioned the deaths of Farah Fawcett, Michael Jackson, and Ed McMahon. It's worth mentioning (albeit belatedly) that Billy Mays also died near the time of the other three deaths. Groups of four celebrity deaths can seem almost apocolyptic, yet statistics says that it's almost assured to happen. Wierd.

Jul. 10 2009 02:15 AM
Patrick

This morning I was wondering what word we used to use before we started using "random." Later today I got around to listening to this podcast that had been on my iPod for a while. It was an unexpected event.

Jul. 09 2009 02:58 PM
Aaron

While sitting around the other day, a stochastic thought popped into my head: In computing, there is no "grandma," just as in our bodies. There are only 1's and 0's in computing. The CPU makes sense of these digits, without any specific final goal. If what I gathered from the show is true, then humans are more like computers that I would care to think. Is this an effective analogy?

Jul. 08 2009 12:55 PM
Carie Kottman

Once again, loved the podcast. I always love the podcast, every single time.
Radiolab always gets me thinking about something (that's the best part--it stirs my neurons). This particular 'cast got me thinking enough that I made a little blog post about it. If you want to take a look, it's at http://knitcrush.typepad.com/home/2009/06/stocha-what.html
Great job, once again, guys. You never disappoint.

Jul. 08 2009 12:09 AM
Charles Platter

Minor correction: stochastic is not Latinate. It is Greek < stochazein, "guess."

Jul. 07 2009 12:25 AM
seabourneclouds

woo hoo! another great episode! congratulations, guys. This one is fantastic. My friends and I have been looking forward to that classic Radiolab goodness we know and love. This is it. Smart, scientific, sexy, and despite the errata, which you have tactfully broadcasted, it is, in a word, bueno. With you from the start.

Jul. 03 2009 04:00 PM
Paul May

Correction: the Yemenia flight was not inbound to France; some of the passengers were from France.

Jun. 30 2009 10:44 AM
Paul May

Having read Jonah Lehrer's book, and listened to the stochasticity episode - today's Yemenia plane crash gave me the shivers. There's still no doubt that air travel is incredibly safe, but two major disasters (both apparently involving flights inbound to France, with similar aircraft) lead the brain to see troublesome, awful patterns.

Jun. 30 2009 10:22 AM
Ben Gawle

Day 15: "Stochasticity" song still firmly stuck in my head on repeating loop

Jun. 30 2009 09:15 AM
OneEarJack

One complaint with this episode: Near the end, at the start of the noisy bacteria segment, the noises are so jarring that I had to turn it off. Really like nails on a chalkboard, it hurt my ears!

Jun. 27 2009 03:22 PM
Jeremy

Isn't radio lab supposed to come out on a regular schedule?
Interesting episode. Making you look at the big picture instead of just you yourself.

Jun. 27 2009 01:07 PM
Dr. Drang

Michael Weiss & Sam:

At the risk of repeating myself (and being a recidivist blogwhore), there is a fairly detailed discussion of the math behind runs of coin flips in this post

http://www.leancrew.com/all-this/2009/06/stochasticity/

on my blog. It contains several links to pages on Mathworld (http://mathworld.wolfram.com) for more information. Interestingly, the calculations involve Fibonacci numbers and generalizations of Fibonacci sequences.

Jun. 27 2009 01:06 AM
John T

So left unspoken in this podcast is the distinction between individual outcome sequences and outcome patterns overall.

The coin flip discussion and others in the podcast reminded me of an unsourced quotation by the famous physicist Richard Feynman on the issue of the seeming randomness of patterns: “You know, the most amazing thing happened to me tonight. I was coming here, on the way to the lecture, and I came in through the parking lot. And you won't believe what happened. I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!”

In statistics (and statistical physics), this is related to the fact that any single combination of outcomes (known as a microstate, like ARW357 or a specific sequence of H's and T's in a series of coin flips) is equally probable, assuming fair and identical systems, but it's the larger-scale pattern (known as a macrostate, like three identical letters and three consecutive numbers on a license plate, or so many heads and so many tails in a set of coin flips regardless of the order) that we tend to focus on. The point is that there can be several microstates that make up a macrostate - there are many ways to get three identical letters and three consecutive numbers on a license plate: AAA123, AAA234, AAA345,... BBB123, ...

One other question I wondered was, how many heads did each group end up with out of the 100 flips? How close to 50? The probability of getting *exactly* 50 heads out of 100 coin flips is something like 8% (I think). The more coins you flip, the *lower* the probability of exactly 50% heads(tails) is. Bizarre and counterintuitive.

Jun. 26 2009 06:13 PM
toasterhead

Perhaps it's too soon for this, but the deaths of Ed McMahon, Farah Fawcett, and Michael Jackson in the space of two days has me thinking about the urban legend of celebrities dying in groups of three. Looking at it through the lenses of some of the statisticians featured on this show, it would seem to me that this phenomenon has more to do with the human tendency to compartmentalize information than the actual spacing of such events. In other words, we perceive them in groups of three, and thus assume that they happen in groups of three.

Does anyone know of any studies of this phenomenon? It would seem like a pretty simple thing to test - you plot obituary dates for a few years on a graph and see if any clumping emerges. And, on the other hand, if there is some clumping, that would seem to prove that it is indeed random - it'd be suspicious if these events were all evenly spaced, right?

Rest In Peace, Michael, Farah, and Ed.

Jun. 26 2009 10:14 AM
Luis Giron

I love the stochastic change...This episode is one of my favorites now...keep up the amazing work!

Jun. 25 2009 10:18 PM
dani

I realized after rereading my comment that it sounded as though I was advocating intelligent design! "...most gamblers don’t consider the man behind the curtain in their calculations." This was completely unintentional!

I was simply trying to say that "there are more things in heaven and earth, Horatio, than are dreamt of in your philosophy" (as I'm sure science will reveal to us in the next 20, 50, 100 years).

And for all of science's proof that the universe is random, the very existence of the scientific disciplines is itself a human imposition of order.

Could it not be that the universe is both ordered and chaotic?

Jun. 23 2009 09:13 PM
Ryan Feeley

The slot machine segment reminded me of your earlier episode about the connection between delaying pleasure and success (the kids and the marshmallows, March 9/09). Really seems to provide evidence that this inclination, and success in life, is hard-wired.

Jun. 23 2009 07:14 PM
Michael Weiss

Sam,

You're absolutely right. After I posted my comment I realized I had made a basic freshman statistics error: I was treating the 94 "positions" where a run of 7 could occur as if they were all independent, which they are not. Obviously if the first "position" has 4 heads and 3 tails, then the second "position" cannot possibly be a run of 7. So my analysis is off. I haven't sat down yet to think through the calculation more carefully. But I have learned (once again) to think before posting.

Jun. 23 2009 05:55 PM
Sam

Re Michael Weiss: isn't there even more to it than that? A run can start on any of the first 94 flips only if the previous flip was different than the type of run in question. That is, if we're looking for runs of tails, then it can only start on the 2nd flip only if the first flip was heads. Otherwise, we'd say that the run started on flip #1. I haven't run the numbers, but I think failing to account for this would lead to a run of 8 being counted as 2 runs of 7, which doesn't really seem fair to the point being made.

Jun. 23 2009 04:56 PM
RadioLab team

Thanks to all of you folks for sharing your thoughts about our latest episode. Since the podcast went live, a Georgia man won the lottery for the second time in a week. He got $1000 the first time and $777,777 the second time. Random or not?? you decide :)

http://news.yahoo.com/s/ap/20090622/ap_on_fe_st/us_odd_twice_lucky

Thanks to Patrick Roberts for pointing this out to us!

Jun. 23 2009 03:52 PM
Berto

Here is evidence that the numbers from the Iranian election are most likely falsified, and they figured it out because they are forced to look like they are random: http://www.washingtonpost.com/wp-dyn/content/article/2009/06/20/AR2009062000004.html

Kind of goes right to the heart of this podcast, huh?

Jun. 23 2009 02:14 PM
brian t

RE: 30 people in a room - at least 2 matching birthdays: this is an example that comes up in undergraduate statistics lectures, as an example of "conditional probability", I think. Basically, as the number of people in a room increases, the odds of two people in the room having the same birthday also increases, as you would expect. On Wikipedia this is under "Birthday problem".

The surprise comes in with how few people it takes: when there are 23 people in the room, the odds are just over 50%. When there are 30 people, the odds are just over 70%. Assuming a 365-day year, it's possible that any 365 people would all have different birthdays, so the only way to be 100% certain two people share a birthday is to bring in 366 people!

Jun. 23 2009 01:37 PM
Gianie

Yay! Thanks so much for this episode, I was thrilled to find a new hour-long one out, and am excited for all the ones to come.

Jun. 23 2009 11:25 AM
Michael Weiss

Guys, if you're going to do a show about statistics and probability, you really should ask someone to check your math.

In the show, a 100-flip experiment was described as roughly 14 "opportunities" to get 7 straight. (Because 100/7 is a little bit more than 14). But actually it's many more opportunities than that, and the probability of getting 7 straight is way, way higher than 1/6.

The probability of getting 7 straight somethings in a row is, as others have pointed out above, (1/2)^6, which is 1/64, or just over 1%. Or, to put it another way, the probability of NOT getting 7 straight somethings in a row when you flip 7 coins is 63/64, very close to 99%.

But when you flip 100 coins, you have to think carefully about how many "slots of 7" are included in 100 flips -- that is, where a run of 7 could be located in the map of 100 flips. Such a run could, of course, start out with the very first flip, and end with the 7th flip. But it also could start with the second flip, and end with the eighth. Or start with the third flip, and end with the ninth. Or start with the... well, you get the idea. The very last possibility is that it could start with the 94th flip and end with the 100th.

So, in all, there are *94* opportunities in a 100-flip experiment to have a 7-straight-run. The probability of every single one of those opportunities failing is (63/64)^94. Which means that the probability of getting at least one run of 7 is 1 - (63/64)^94... which crunches out to about 77.24%. Way, way more than one in six.

Jun. 23 2009 10:42 AM
Jordan

The lesson I learned from Douglas Adams: coincidence happens.

Jun. 23 2009 12:58 AM
dani

I love love love this show! And oh my, how I've missed you. I am thrilled to have the hour-long format return. Thank you!

As for this particular show, if the universe is completely random, noisy, chaotic, unpredictable, what about the golden section? Is that just humans imposing an order on top of a chaotic natural world--seeing patterns where there are none? What about waves, artichokes, sunflowers? (It seems pretty obvious that there is, indeed, a pattern there.)

We have the ability to consciously find and recognize patterns (cycles of the cosmos, tides, etc.). In the segment on the woman with Parkinson's, you explained that our minds *seek* order. Why is that? Simply to get food and procreate? And further, why have our minds evolved to do complex mathematical equations (like the ones in this show that debunk a theory of an ordered universe)? Isn't this just imposing order on our chaotic natural world? Or science for that matter! All of our classifications, hypotheses, and theories--are these not artificial constructs that seek order (where science then concludes there is none)!?

Are we delusional to think there is a greater order, when really it's all random? Or might it be more likely that our science is too young to explain all the mysteries of the universe? Not too long ago scientists told us the earth was flat, the sun revolved around the earth, and the atom was the smallest particle.

Lastly, didn't you do a show about casinos? I remember the casinos secretly monitored the players and then just when they were about to lose interest, someone instantly appeared to offer them some sort of prize in order to keep them playing. In their desire to figure out slot machine patterns, most gamblers don't consider the man behind the curtain in their calculations. http://www.wnyc.org/shows/radiolab/episodes/2008/11/14/segments/113278

Jun. 22 2009 08:57 PM
ron

hey guys. great show (as always).

you kind of touched on something here that I'm endlessly fascinated with -- dark side of the rainbow.

for those unfamiliar, this is the strange marriage between pink floyd's dark side of the moon and the movie "the wizard of oz." the movie and cd appear to sync up in many ways - certain musical changes match up with visual changes, lyrical references in the music sync up to moments in the film.

granted, most people are probably stoned when viewing this, but even when straight and sober, the seeming connections are hard to deny.

but, but, but.... there are no connections. it is simply a trick of the brain (unless you believe in some kind of syncronicity of the universe - or as one person I know described it - the movie & album are the sending the same messages from parallel universes) otherwise, it's just a completely random phenomenom. the brain is working completely on overtime creating order where there is only chaos. it's quite amazing.

from wiki: "Fans have compiled more than one hundred moments[5] of perceived interplay between the film and album, including further links that occur if the album is repeated through the entire film. This synergy effect has been described as an example of synchronicity, defined by the psychologist Carl Jung as a phenomenon in which coincidental events "seem related but are not explained by conventional mechanisms of causality."[6], although most accounts assume that the effect was deliberate on Pink Floyd's part. Detractors[7] argue that the phenomenon is the result of the mind's tendency to think it recognizes patterns amid disorder by discarding data that does not fit. Psychologists refer to this tendency as apophenia. Under this theory, a Dark Side of the Rainbow enthusiast will focus on matching moments while ignoring the greater number of instances where the film and the album do not correspond."

here's one sequence (you can watch the entire thing on youtube if you search about a bit):

http://www.youtube.com/watch?v=2OxFtS9RYpw

keep 'em coming, guys! you make me smarter every show!

"The brain is the most complex thing in the universe, and it's right behind the nose!" - michel gondry (in "the science of sleep")

Jun. 22 2009 12:57 AM
Rosalie

stochasticity and the Iranian election:

http://www.washingtonpost.com/wp-dyn/content/article/2009/06/20/AR2009062000004.html?hpid=opinionsbox1

just like your coin-tosses.

Jun. 21 2009 11:14 PM
Tatter Demalion

really enjoyed the video. think this is a great idea... the podcast you did last year featuring zoe keating was a great and unique bonus to all the fascinating science programs.

Jun. 20 2009 11:15 PM
Jesse

You missed a very important and relevant aspect of this question.

Professor James Conway, of the famous Conway's Game of Life, has recently proven that if you grant the assumption that humans have free will then electrons must have a kind of 'pure' free will of their own. The will of the electron is, in fact, so free that it cannot consult any part of the past history of the universe! In his lectures, he calls it more akin to 'Free Whim'.

What's more, and where it gets interesting, is that the result of their purely free decision CANNOT BE RANDOM, and simultaneously cannot be a function of any past or present bit in the universe.

I really hope you read this comment, and take a little time to watch at least some of his very interesting lectures on his proof:

http://www.princeton.edu/WebMedia/flash/lectures/2009_03_04_conway_free_will.shtml

Jun. 19 2009 11:29 PM
Robert

Thanks so much for being back. You've been missed, so keep them coming!

Jun. 19 2009 01:51 PM
Dr. Drang

I loved this episode and couldn't get the coin flipping demonstration out of my mind. Ended up writing a longish post on the math behind it:

http://www.leancrew.com/all-this/2009/06/stochasticity/

toasterhead: What makes coin flips and dice rolls random is our inability to control all the inputs that determine the outcome. There's a Fritz Leiber story called "Gonna Roll the Bones," the setting of which is a craps game between the hero and the Devil, both of whom have such great control over the Newtonian mechanics of their throws that it's a contest of skill rather than chance.

Jun. 19 2009 12:43 PM
Yisroel

Fascinating! I love the show. One of my favorites.

Consider this…as you said in the show regarding coin flips, 7 in a row out of 100 is a normal characteristic of randomness but 7 out of 7 would be wondrous (my word, not yours).

Isn’t the fact that the sun rises every single morning without fail wondrous? And isn’t "brushing off" the wonder of the sun with facts and figures about gravitational pull and such, kind of like saying "it is because it is"? Defining a word by using it in a sentence (a big “no-no” according to my 6th grade teacher, may she rest in peace)?

Aren’t the rules and laws that science has deduced and distilled so well over the years, no more than a description of what is? They don’t seem to go very far in explaining how such elegant order came to be and continues to be…

Perhaps order is the default state and chaos is introduced to make discovering the creator more challenging…

Jun. 19 2009 12:21 PM
Applepaul Wilson

I absolutley LOVE RADIOLAB. I was so excited to see some new episodes, i've listned to some of the episodes over and over. I have suggestion. There's alot of stories that i'd like my nephew to hear, he's 10. He's very much into science and ants in particular. I made him a copy of the emergence episode as well as the E. O. Wilson "Chasing Bugs" podcast. But, there'a lot of material that's very adult oriented, so i'm not sure if my sister will let him hear it. What about presenting some podcasts for kids that have just the material from radiolab episodes that is appropriate for children? Thanks and keep up the beautiful work you guys do! - Applepaul

Jun. 19 2009 12:13 PM
Connor Walsh

The talk of two people of 30 sharing a birthday reminds me of a joke we used to have when I worked on a radio show in China. We used to say that if we ran out of material before the end, we'd just ask the presenter to sing "Happy Birthday to you, Happy birthday to you, Happy Birthday Mr Li…" because it was bound to apply to a ton of listeners!

Jun. 19 2009 10:14 AM
Adam

Robert, I have the same ringtone. I thought my phone was ringing. Random. : )

Jun. 19 2009 10:06 AM
Scott

RE: 30 people in a room - at least 2 matching birthdays.

(First off - I am excited to hear we're getting a regular longer show!! Thanks Radio Lab!)

Odd little Stochasticity with the 30 people and matching birthdays you mentioned, Rachel. I lean heavy on the I can't believe that side of any issue that I can't get my head around easily -- Like this 30 people and 2 matching birthdays scenario.

It just so happens that I first heard of this idea when my 7th grade math teacher threw it out to our classroom during a probablity lesson. She said: "In this very classroom, of just a little over 30 students, there is a high probability that 2 or more of you were born on the same day of the year."

She didnt explain the math behind her theory, and the rest of the class seemed to take the statement at face value. I had to stop and say, I don't understand. She said should we go through the class to see if the idea is right? Everyone else seemed to want to go on with the lesson. I really pressed the issue, so we got a demonstration. She started at the back of the room, furthest row from her left.

She got maybe 4 students into the demonstration, when I heard my own birthday . We stopped the demonstration. But I still didnt have the math, but I definitely saw she was right, of course.

Jun. 19 2009 02:21 AM
Rachel

In response to Jaime: When I was a kid, I read that if you take any 30 people in the world (for example everyone in your class), the odds are better that two of them will share birthdays than that all 30 will have different birthdays. This is true because of the "zoom out" phenomenon--there are billions of people in the world, but only 366 possible birthdays, so there are millions of people who share each one. My guess is that, like most people, you spend a lot of time with people about your age. So, finding someone with the same date and year as yourself--out of all the people you have met--is not a super amazing coincidence, but if it happens to you, just like with the gambling example, you want to find a pattern in it. The cool thing is that there is a pattern--it's just really big. I think that in some ways, this whole show just backed up the idea that we're all part of something bigger than ourselves--and there is order in the universe, after all. It's just an enormously huge and complicated order. It wouldn't be nearly as fascinating if it were simple, would it?

Jun. 18 2009 11:08 PM
Paul

Soren!

The first segment of this show may be my favorite Radiolab segment EVER. The show lost steam near the end (which is too bad, because Carl is great), but that opening story was brilliant.

I'm a teacher, and I do the exact same coin flip exercise with my students. Love it.

Now, as much as I like the Stochasticity video/song posted separately, I was hoping for the mp3 of the stochasticity song that was featured at the end of the show. How about it, Radiolab gang?

Jun. 18 2009 09:57 PM
Ryan

YES - you made something clear to me here - and probably made my marriage a lot happier - I'll check it out anyway - here's the problem, my wife says I never surprise her anymore - like I used to when we were dating, and that if I just bought her chocolates or flowers every once in a while, then she would love it... but I DO buy her chocolates and flowers, just not often enough, it seems... maybe if I treat it like the basketball hot streak that you talked about here... I'll give it a try anyway...

Jun. 18 2009 07:48 PM
RG

Hate to burst your bubble even further, but I think your math prof was confused about the coin scenario. If you did 14 sets of 7-flips, the probability that one (at least one) of those 14 sets was all (heads or tails) is 19% (1-63/64^14). However, that's not what you did. What you did was look for the longest consecutive string. The way I look at it, what's the probability that you flip 100 coins and none of the strings is longer than 6? Well, it's looking at a string of numbers, 1-6, with probability 1/2, 1/4, etc. that add up to 100. How many numbers are there? It's an expected value of n*1/2^n, sum from 1-6, which is 1.875. There are an average of 53 numbers. Okay, picking a number from 1-6 has a probability of 63/64. Doing that 53 times, gives about 40%. So, the probability of having a string of 7 OR MORE (heads or tails, again) somewhere in that string of 100, is about 60%.

Jun. 18 2009 05:35 PM
Andreas

Hy there,
can someone name the piece that plays at 4:05, my girlfriend threatens to kick me out if I´m going to repeat this 10 seconds of great horns on and on... Instead of being kicked out I would prefer to switch to my MP3 player ;)

Cheers,
Andreas

Jun. 18 2009 04:15 PM
Matt

In biological systems, introducing stochastic 'noise' has been found to help improve the signal strength of the internal feedback loops for balance and other vestibular communication. It has been found to help diabetic and stroke patients with balance control. see http://www.bu.edu/abl/files/fulltext.pdf

The financial markets use stochastic models to represent the seemingly random behaviour of assets such as stocks, commodities and interest rates. These models are then used by quantitative analysts to value options on stock prices, bond prices, and on interest rates, see Markov models. Moreover, it is at the heart of the insurance industry.

Jun. 18 2009 02:56 PM
Kylie

Great show. But HEY lots of us love hockey!! Boo to 'the sport no one cares about'! Thanks guys, I love to laugh, smile, tear up, the whole gamut while listening during my work day...I get some funny looks from co-workers, all because of you!

Jun. 18 2009 02:33 PM
Michelle

Thank you for the hour long shows again. I really appreciate all the hard work that goes into this.

This show was wonderful. Can't wait to hear what else you guys have in the works.

Jun. 18 2009 08:18 AM
Bummus

This was a really great show. I feel like this show really hit the core of the "RadioLab" concept. Just wanted to say "good work."

Jun. 18 2009 12:11 AM
j.vanpelt

Did I hear music by Casino Vs. Japan at 27:03? I knew him (Eric Kowalski) in Milwaukee. His music is _very_ good, I highly recommend it.
http://blip.fm/~8enw0

Jun. 17 2009 05:39 PM
Matijs

Great show and awesome to hear there are going to be regular hour long episodes again! Keep up the awesomeness.

Jun. 17 2009 05:18 PM
radiolab

Roger,

You are the very man we want to talk to. If you don't mind, and we are sure you didn't expect this, could you please send us your contact info here: radiolab@wnyc.org.

Thanks,
Radiolab Team

Jun. 17 2009 05:10 PM
Ric

I'm captivated by this show...just started listening.
I am a clarinetist and hoped you could share the composer of the music at time 21:03 on the podcast.
Cheers

Jun. 17 2009 04:34 PM
Andrew

I love this program. I recently started listening online after my girlfriend told me about it, and I particularly enjoyed this show because I spend all day thinking about neuroscience and mathematics, which very quickly gets into matters of chaos, stochasticity, order from disorder, etc.

I do want to offer a clarification/correction : shortly after the 51 minute mark, the montage describing the messy reality of gene expression and protein synthesis intermixes the words "chaos", "random" and "messy", which reinforces the perception that chaos = random = stochastic. To a scientist, chaos and stochastisity are NOT the same, although the difference is subtle. A truly stochastic process is by definition unpredictable. A chaotic process is theoretically predictable. The details get complicated. For anyone who wants to hear it from the experts, Edward Lorenz discovered chaos, and Steven Strogatz wrote a great text book on the subject.

Thanks, and keep up the good work!

Jun. 17 2009 03:49 PM
eso_si_que

Another person curious about that clarinet piece here.

Looking forward to more Radiolab!

Jun. 17 2009 03:45 PM
Kurt

This was one of my favorite episodes. Thank you much!

The coin flip reminded me of how the iPod created an option to make their shuffle play less random to seem more random. (Which is exactly how the professor knew which coin flips were random and which were made up).

Sometimes I think about all the things happening in the world, where the odds are "one in a million". Well, it doesn't take long until a million of these things happen around the world. Thus, odds are, these freakish events will happen.

Jun. 17 2009 03:42 PM
Roger

At the end of the Hot Hand segment, it's mentioned that DiMaggio's 56-game hitting streak is the one true exception to the randomness of the Hot Hand.

But that isn't true at all! It's not as if DiMaggio got a hit in 56 straight at-bats, that would truly be astounding. Hitting in 56 straight games, in which a player gets at least 3 at-bats, is much easier. In fact, simulations have been run, and DiMaggio's streak appears to not be as much of an outlier as we thought. See http://www.nytimes.com/2008/03/30/opinion/30strogatz.html

Otherwise a great episode, keep up the good work!

Jun. 17 2009 03:35 PM
synchronicity happens

The reality will not be mathematicized! :)

Jun. 17 2009 02:37 PM
Tim

Jaime -
Your story about meeting someone with so much in common with you is definitely incredible...but to me it becomes less so when you think of how many times you flipped the coin, so to speak. You've met countless people in your life, and many who shared several of those characteristics with you. Eventually, it doesn't seem that unbelievable to me that you would meet one person with even more similarities. We seek non-randomness, and the time between non-random events is completely nondescript - as opposed to those rare salient moments that we tell stories about. That's the way I see the balloon story, too. There are thousands of other balloon stories that are almost as incredible but fell short on one detail - first name, last name, age, etc. We don't share those stories, so the Laura one seems even more unique than it is.

Jun. 17 2009 02:30 PM
Michael

Every time I listen to your show I think to myself that there is nothing better on the airwaves (or downloadwaves). Thought I should share that with you. Amazing work. Can't say enough about it. Thank you.

Jun. 17 2009 02:28 PM
Shana

I really enjoy your shows. I have to say though that the high pitched screeches and icky prolonged static noises bother me...somehow literally to the point of nausea. They prompt me to dive for the volume control, and eventually dive for the off switch.

I'd like your show even more if you found some other sound effects that got across the feel you're going for.

Thanks!

Jun. 17 2009 02:20 PM
Jaime Devereaux

I have to say...
To simply dismiss the idea that something greater is taking place rather than coincidence is utterly ridiculous. Let’s look at it in reverse. What if that balloon landed in any other person's yard? It would have been no big deal. Right? That’s coincidence. To land in the yard of someone that has the same exact name, looks the same, has similar likes and dislikes, and a matching rabbit... That’s a force of something. What, I don't know. God, energy, fate, destiny, or even necessity. It was not just coincidence. Everything happens for a reason and life develops and reacts the way it does for a reason. Simply putting some scientific notions to the event just gives us reason not to believe in something or someone greater than us. I once meet a woman (I am a man) with the same first name, exact birth date - including year, exact height, weight, and skin color as myself. Coincidence? I think not.

Jun. 17 2009 12:52 PM
Quincy Holmer

Confusing the debate between stochasticity and determinism with the debate between freedom and determinism is just a classic case of equivocation, i.e. calling stochasticity and freedom by the same name. The probabilities of quantum mechanics don't guarantee free will. To me, the senses of self and choice seem biologically adaptive but perhaps constitute some of the many illusions we experience as the human perception of reality.

Maybe you could ask Brian Greene or Oliver Sacks, or any of the other geniuses to whom Radiolab provides due attention.

Jun. 17 2009 12:25 PM
toasterhead

Very interesting show, as always!

Though I can't help thinking about coin flips - they're not truly random, are they?

The outcome of any given flip is really a function of the initial orientation of the coin, its weight and dimensions, the height and direction of the toss, its rate of spin, and how far it has to fall before landing, with perhaps some corrections for wind resistance and the amount of wear on the coin. There should be a relatively simple formula that could predict the outcome of every flip of any coin given the initial data.

And unless I'm missing something, you could do the same for dice rolls, roulette wheels, ping-pong balls in lottery machines, even the path of a water molecule in Brownian motion. It would seem that a lot of what we call "random" is really predetermined by Newtonian physics.

Jun. 17 2009 11:04 AM
zadkiel

This episode was so intense...

Jun. 17 2009 10:08 AM
Timothy

I knew I used the term correctly. It is most appropriate in the context of discrimination.

Jun. 17 2009 06:27 AM
Julie F.

Wonderful episode as usual.

I have one question.

What is the name of that lovely and haunting clarinet bit you had in there?
Thanks!

-Julie from
Long Island, New York

Jun. 16 2009 11:02 PM
Luis

Hey! First time here! Wanna know how I got here to listen to your show for the first time? I read a science blog from the country I live in that leads to a blog that led to this site, and BANG! I find that the podcast thats just before this one is about who? Juana Molina! That by chance is from the country and the city I live in! A MIRACLE!

Loved the show guys, now i'm going to download all the previous "episodes".

Greetings from Buenos Aires, Argentina

Jun. 16 2009 08:44 PM
Kevin H

I really liked the show, but I think it's missing one part at the end. In some ways, your first story explains your last. Think back to that coin. You might get 7 heads or tails in a row, and that shouldn't be very surprising, but I bet if you totaled out all of the heads and all of the tails in your 100 flips, you would probably be somewhere close to 50.

That is exactly how even stochastic gene expression can end up in a beautiful song. If you look at any one instant of gene expression, it's noisy, but if you look at a bigger picture. Say, the gene expression of one bacteria over 2 hours, or the gene expression of 1 million bacteria, well you would start to see a pattern emerge.

That's the key to randomness. It all depends on the scale you are looking at. Look at one streak, or one balloon, and it's unexplainable. Look at the bigger picture, and it becomes (more) predicable.

Jun. 16 2009 06:54 PM
Tony Bruguier

Regarding the slot machine, it is not "totally random", I mean not iid. They design slot machines so that they attract us and make us think we can predict them. Each trial is not independent.

Jun. 16 2009 05:20 PM
JL

I laughed, I cried, I made dopamine, and will come back for more.

btw The e-coli fluorescing chaotically sounded like BLACK DICE!

so are those wall street journal thumbnail portraits truly stochastic or purposefully stippled?

Does PI make the best random generator and is there a way to make a truly unscripted random generator?

Is all phenomena some up-scaled fractal self-similarity of all the quantum buzzing?

Jun. 16 2009 04:15 PM
Layla

I'm excited about the new format! Looking forward to hearing new episodes.

Jun. 16 2009 04:10 PM
Soren

Steve,
Right, the the math being discussed was for EITHER OR ... not for the specific instance of 7 tails.

This is something our guest, Jay Koehler, pointed out to us, it just got lost, as I said, in our editing.

Jun. 16 2009 02:44 PM
Soren

P.S. Sorry about the confusing math terms ... I was flipping back and forth between two different notations. Just to be clear: .0078 probability is the same as .78% probability. So .0078 + .0078 gets you .0156 or so, which is the same as 1.56%.

Jun. 16 2009 02:42 PM
Steve

Sorem

Thanks for the reply.

So do I understand then that is it a 1.5% chance that you would get 7 heads or 7 tails in a row (.78*2) and the math being discussed during the podcast was for either or, not the specific instance of 7 tails in the story?

Jun. 16 2009 02:42 PM
Soren

Steve -
You're pretty much right. The chance of flipping 7 HEADS is .0078 and the chance of flipping 7 TAILS is .0078 ... but since either one would've been freaky, to get the true probablility of Jad and Robert having that 'crazy' experience, you need to add the two probabilities together ... which gets you 1.5 something. And that, in our admittedly loose language came out as a 'a little over 1%.'

Unfortunately, somewhere in our editing ... trying to cut through some of the nuance of that conversation, we didn't quite get this point to the surface.

So our final number is right, but the logic of the conversation is off.

In the end, you're right though. We've should've made that point clear.

-Soren.

Jun. 16 2009 02:36 PM
Steve

I think some of the math may have been slightly off....

Podcast Time Point 13 min 41 seconds;

7 tails in a row is 0.78125% not "Just over 1%"

1 coin flip = 50%
2 coin flips = 25%
3 coin flips = 12.5%
4 coin flips = 6.25%
5 coin flips = 3.125%
6 coin flips = 1.5625%
7 coin flips = 0.78125%

Jun. 16 2009 02:21 PM
Ben

Bring back the old format! Where are the segment summaries??

Jun. 16 2009 01:35 PM
nandorocker

Haven't heard the episode yet, but I can tell you as your #0 fan that I'm STOKED about this new format idea and hope to hear you guys with more regularity.

Jun. 16 2009 01:30 PM
Radio Lab

Hey @Bubble Buster. Your observations about the wind speed/balloon and individual's perspective of a random event were on the dot. We'd love to get in touch with you. Can you send us email (radiolab@wnyc.org) with your contact details ? :)

RL team

Jun. 16 2009 11:46 AM
Connor Walsh

Guys, you made an old lady cry!

But we forgive you. Thanks for hour-long episodes.

Huzzah!

Jun. 16 2009 10:46 AM
travis

allow me to relay my own stochastic experience of today: this morning a certain song randomly popped into my head. in fact, i specifically imagined hitting shuffle on my ipod an hearing that song. so later i created a playlist with over 100 songs (including the song i was thinking of), hit shuffle, and indeed the song i was thinking of played first! what is even stranger is that i have had several other instances of playlist shuffle precognition. what does it mean? probably nothing but it is interesting nonetheless and made a fitting prelude to hearing this show today!

Jun. 16 2009 09:33 AM
Nevin

@Bubble Buster: Isn't the point that the probability of no unlikely coincidences EVER happening to you is very low?

I liked listening to this because it gave me this thought: People generally are in awe of the fact that life has evolved on Earth. When applying this type of thinking, this Earth-life situation gets reframed as the probability of life evolving on some planet somewhere out of however many planets are in existence. When framed this way, it seems that the probability must be extremely high!

Then I thought: What are the chances that I am here among that life? This is of course silly... I mean, what are the chances I would be alive somewhere devoid of life? :)

Jun. 16 2009 04:48 AM
Bubble Buster

As a profession bubble-buster, I have to throw in my 2 cents. The first cent will add one more (little) bubble-buster; with the second cent, I'll attempt to redeem myself.

1. As an atmospheric scientist, I can add that the 'prevailing' wind can be very misleading when you compare wind direction at ground level and above (where the baloon was)--the two are not always the same and often quite different.

2. One point you made lightly at the end of the first segment is crucial here: the perspective of the definition of 'randomness' is crucial. If statisticians are only looking at the perspective of the event, and not at the perspective of the individual, they are missing the point. It is indeed unlikely that this not-so amazing coincidence will happen to ME.

Jun. 15 2009 11:49 PM
Aaron

That change in format is a little stochastic... heh. But really, it's a good idea. Aside, I can't wait to listen to the episode!

Jun. 15 2009 11:11 PM

Leave a Comment

Register for your own account so you can vote on comments, save your favorites, and more. Learn more.
Please stay on topic, be civil, and be brief.
Email addresses are never displayed, but they are required to confirm your comments. Names are displayed with all comments. We reserve the right to edit any comments posted on this site. Please read the Comment Guidelines before posting. By leaving a comment, you agree to New York Public Radio's Privacy Policy and Terms Of Use.