EP178: Unlikely

[vilate]

I think the idea was interesting, but the oh-so-sweet love story really got to me by the end.  I like love stories sometimes but I with the main character suddenly "opening his mind" to the possibility of being with Tuesday really annoyed me.  I think I would've preferred an ending that didn't involve some convenient opening of the heart that leads a character to realize they've been attracted to another character from the beginning.

[Windup]

I kept thinking that surely a philosophy professor could put up a better fight in terms of the correleation/causation business than, "it's all bullshit!"  I mean, come on, wouldn't this guy "hold forth" on the subject, give examples, discuss spurious correlation, etc.?

Hurt my already-strained suspension of disbelief...

[gregS]

OK, i liked this story but i had to post because of the podcast i listened to before this one. I had just finished listening to NPR's Science Friday podcast.   (http://www.sciencefriday.com/program/archives/200810037)   There was a 12 minute story about how people create superstitions when they are stressed to gain a sense of control. I thought the this story about somebody not creating superstitions in spite of evidence was a wonderful insight into human psychology. People are not willing to uproot their belief system unless there is some serious pressure.

[Praxis]

This might be a bit 'out there' but this story seems more like the statistics was used to show what the world is really like than that the statistics or 'science magic wand' was the thing that actually caused the world to be this way.  Like, the story shows this fact of the world through stats/science spectacles so it becomes a-story-with-science whereas if this fact of the world was found differently in the story, there would be other spectacles used.

It could, with not a lot of tweaking, have been a magical realism story, with the "statistical expert" being replaced with "someone with a sense and intuition of things" and the basic premise and even the character development kept as they are already.

Maybe I just want a big space rocket in the story somewhere.

[ajames]

After Friction, how unlikely was it that I wouldn't love the next story of Will McIntosh's? Unlikely.

Well, I didn't love it, but I liked it. The correlation is causation bit bugged me a bit, but I'll take Phronk's explanation and also allow that in this world correlation can be causation.

I'm trying to figure out if Steve was right or wrong in his outro. Is "correlation is not causation" a hypothesis that can be rejected? If so, this is indeed a scientific story. Or is it a basic principle of scientific research? If so, then this story is actually anti-scientific. Not sure which is which, but perhaps that's a sign that I should brush up on my philosophy of science...

I didn't listen to Steve's outro carefully, apparently. But "correlation is not causation" has been demonstrated again and again, and is a basic working assumption of science. That's why we experiment - if correlation was causation, we'd simply need to observe. Since its not, we need to manipulate variables to determine the functional relationships. This story is scientific in the sense that Dr. Barry attempts to manipulate some variables. The finding that in some cases correlation appears to be causation could mean that the story has moved into a world that doesn't operate like ours, or it could mean the the cause is yet to be determined (in the ice cream sales/crime rate correlation, it might be possible to do some manipulations that appear to confirm that the correlation is causal. You could, for example, shut down the ice cream parlors on several random days, and find that the crime rate goes down on these days. It might be the case, though, that by chance the randomly selected days were all very cold and rainy).

[Edited to add footnote below]

I went back and listened to Steve's outro again, and, as usual, he was spot on. As Hume pointed out, observation is one thing, explanation another. On a philosophical level, there's a lot of interesting things to be said on causation and explanation. On a practical level, though, I'll stick with what I said above.

[WillMoo]

I personally think that the best readings are when the reader doesn't attempt changes, or large changes I should say, in the voices. Old time radio had two or three actors in the studio. Even if they had only two actors they usually had a male and a female if the parts called for it. That way you didn't get the silly falsetto male voice for the woman or the equally silly woman doing a male voice. Also, the stories were, most often, written as screen plays so it was more obvious who was speaking. Personally, I think that when reading a story, subtle changes to differentiate the characters works best.

All of that aside, thanks Steve, for bringing these stories to us!

[eytanz]

I'm trying to figure out if Steve was right or wrong in his outro. Is "correlation is not causation" a hypothesis that can be rejected? If so, this is indeed a scientific story. Or is it a basic principle of scientific research? If so, then this story is actually anti-scientific. Not sure which is which, but perhaps that's a sign that I should brush up on my philosophy of science...

I didn't listen to Steve's outro carefully, apparently. But "correlation is not causation" has been demonstrated again and again, and is a basic working assumption of science. That's why we experiment - if correlation was causation, we'd simply need to observe. Since its not, we need to manipulate variables to determine the functional relationships. This story is scientific in the sense that Dr. Barry attempts to manipulate some variables. The finding that in some cases correlation appears to be causation could mean that the story has moved into a world that doesn't operate like ours, or it could mean the the cause is yet to be determined (in the ice cream sales/crime rate correlation, it might be possible to do some manipulations that appear to confirm that the correlation is causal. You could, for example, shut down the ice cream parlors on several random days, and find that the crime rate goes down on these days. It might be the case, though, that by chance the randomly selected days were all very cold and rainy).

[Edited to add footnote below]

I went back and listened to Steve's outro again, and, as usual, he was spot on. As Hume pointed out, observation is one thing, explanation another. On a philosophical level, there's a lot of interesting things to be said on causation and explanation. On a practical level, though, I'll stick with what I said above.

Oh, I agree with Steve entirely about science. What I'm not entirely sure about is how to take this story with respect to science. The story has three characters - Sam, Dr. Barry and Tuesday. I think it's possible to take each of them as standing for a different viewpoint of the relationship of explanation and observation. Sam (at least in the beginning) is irrationally biased towards rejection of data- for him, observation without explanation is insufficient. He may recognize that the data exists, but he refuses to grant any validity to the results without knowing why the correlation is there. Dr Barry is a empiricist - he sees correlations, and he constructs an experiment that tests if these correlations persist. He does not provide explanations, he lets the data speak for itself. Tuesday is a mystic; she believes that the explanation lies in a spritual force, the numenous.

The reason I'm not sure about whether to take this story as pro- or anti- science is twofold. The first is that the story seems more sympathetic to Tuesday than to Sam. And the ending implies that Sam came around to her point of view, while she didn't need to make any adjustment to hers. The second is that, as Steve observes, the story does not provide a voice to scientific theory. As Steve (and you) also observe, this is not necessary. But given that msyticism is given a voice, this creates an imbalance.

That said, I don't think this is an entirely natural reading of this story; but then again, I don't think that it is any less natural than arguing that it is a story supporting science by rejecting the view that everything must be explained to be valid. If I were to guess, I'd say that the author's primary concern was mostly Sam's emotional journey, and that the philosophical implications of the way this world works on how we view the real world were incidental. But to the degree that we do go down that route, I would say that the story is  ambiguous in its stance.

[bedlamite9]

All of that aside, thanks Steve, for bringing these stories to us!

I wholeheartedly agree. I love this podcast and am a regular listener who will frequently share my favorites with friends and family. I meant no disrespect to Steve by criticizing his voices, but merely hope that he will take it as constructive feedback for future readings. The stories are great, just tell the story.

[SFEley]

I wholeheartedly agree. I love this podcast and am a regular listener who will frequently share my favorites with friends and family. I meant no disrespect to Steve by criticizing his voices, but merely hope that he will take it as constructive feedback for future readings. The stories are great, just tell the story.

Oh, certainly, no offense taken.  If no one criticizes it's hard to improve.

For whatever reason, I found Tuesday's character unusually difficult to read with a neutral affect.  I actually tried a couple of takes on this story; be glad I didn't go with one of the earlier attempts.  It's likely that I should have given this story to someone else to read.  Learning experience.

[Phronk]

As a psychology major (like the author), I just feel the need to clarify something.  We have to be careful about taking "correlation does not equal causation" too far.  It's true that a correlation does not NECESSARILY mean there is causation.  However, whenever (*) there is causation, there IS necessarily a correlation.  In the ice cream example, the correlation between ice cream sales and crime does not mean there is causation.  But the fact that high temperatures causes both ice cream sales and crime DOES mean that there is a correlation between temperature and sales, and between temperature and crime.  And a good way to discover this causal connection is to first notice the correlation.

The only difference between the story's world and the real world is that there are causal connections between things (like the two characters being together), which aren't explained in the story but may be perfectly natural.  The story does NOT ask us to abandon any laws of statistics, the scientific method, or rationality.  In fact, manipulating the variables by having the two people go out together (and apart) would be the best way to test the connections first indicated by the correlations.  This little bit of fantasy thrown into a solid scientific framework is one of the things I loved about this one.

Sorry for getting geeky.  Wait...I'm on a science fiction podcast forum....I shouldn't be sorry.



(*) Almost.  There are exceptions, but let's not complicate things.

[ajames]

Good points, eytanz. I am curious about one thing. You wrote

... Sam (at least in the beginning) is irrationally biased towards rejection of data- for him, observation without explanation is insufficient. He may recognize that the data exists, but he refuses to grant any validity to the results without knowing why the correlation is there...

Do you think Sam's rejection of Dr. Barry's hypothesis is irrational? Or are you basing your description of Sam on this and other information we receive about Sam throughout the story? As you point out, Sam doesn't reject the data, but he does reject the interpretation of the data. To me this seems rational, for to accept the interpretation without any further explanation would be to accept a radically different universe than the one I know. Before I made such a leap, I would want more data.

[stePH]

Question: Near the end I kept hearing (I listened twice) “The pink sweater soiling his dashboard.”  What did I mishear?

You didn't mishear anything.  I clearly heard "sweater" too, when from the context it obviously should have been "sneaker".  When I heard that, I wondered whether the error was in the text, or in the reading.  Steve, can we get a ruling?

[wintermute]

My main issue with it was the distinction between correlation and causation; given enough data to mine though, you can almost certainly find a random correlation between two unrelated sequences (say, violent crime rates and average age of men on bicycles) without there actually being any kind of link at all. True, they come up with experimental data that suggests there a genuine correlation and not just noise, but I have difficulty imagining it getting quite that far...

While what you say is true, it's also true that if there truly is a causal connection between two things, there will also be a correlation.  It's clear that in this story's world, there are genuine causal connections between unusual things, so data mining would be a good way to detect them (along with some inevitable spurious correlations).

Well, yes. Causation implies correlation, but that doesn't mean that correlation implies causation. And even if the correlation does imply a causation, it doesn't say which way it goes; perhaps, if the universe is left to itself, the proximity of these two people is caused by the low accident rate rather than the other way around.

I just have difficulty imagining Dr Berry walking into the Mayor's office and saying something like "there's a very strong correlation between shark attacks and ice cream sales, so we should ban ice cream to stop people getting eaten by sharks!" and not getting laughed out of there...

[wintermute]

Sam (at least in the beginning) is irrationally biased towards rejection of data- for him, observation without explanation is insufficient. He may recognize that the data exists, but he refuses to grant any validity to the results without knowing why the correlation is there.

Sam, I think, accepts the data that accidents are less common when the two of them are together, but rejects the hypothesis that their proximity affects the accident rate, because it has a very low prior plausibility. That is, it contradicts what we already know about how the world works.

That, of course, doesn't mean that it's wrong; just that it's an extraordinary claim that requires extraordinary evidence. The experiment they're part of is a good step towards that but it would take quite a long study, with the times they're together or apart properly randomised in order to overcome the possibility of mere statistical noise or another, unaccounted-for, variable.

It's not at all irrational for Sam to consider the hypothesis vanishingly unlikely, especially before any actual data has been collected.

[Swamp]

Very good and fun story.  I don't have a problem with the two characters falling in love at the end.  I do however wish that Sam hadn't become "a believer".  I think Sam should have been able to love Tuesday without buying into the premise of the statistics.

I liked the scene before the crash when Sam slowly peeled off his bias and began to truly picture himself with Tuesday in incremental degrees of intimacy.  I think we have all done that in various ways.  He realized how lonely he was; and here was a woman who wanted to be with him who he had come to know and appreciate.  Once he let the barriers down and let go of the spite, i figure he was thinking, Who cares about the statistics and other people's expectations.  I'm willing to love this woman.  Then not being too experienced at the relationship thing, he just went for it full bore with the kiss.  I can totally see this as a real reaction.

From Tuesday's perspective, she was open to the relationship.  She was more free-loving to begin with and her previous husband was an older man.  Sam's age may have even been a comfortable aspect of the relationship for her.

Anyway, aside from Sam becoming a believer instead of simply falling in love, I really liked this story.  The freakenomics experiments were fun too.

[eytanz]

Good points, eytanz. I am curious about one thing. You wrote

... Sam (at least in the beginning) is irrationally biased towards rejection of data- for him, observation without explanation is insufficient. He may recognize that the data exists, but he refuses to grant any validity to the results without knowing why the correlation is there...

Do you think Sam's rejection of Dr. Barry's hypothesis is irrational? Or are you basing your description of Sam on this and other information we receive about Sam throughout the story? As you point out, Sam doesn't reject the data, but he does reject the interpretation of the data.

I was basing my description on the information we receive later - while it may seem rational to us to be skeptical, Sam's motivation seems to be more about not wanting it to be right than healthy doubt.

To me this seems rational, for to accept the interpretation without any further explanation would be to accept a radically different universe than the one I know. Before I made such a leap, I would want more data.

Sam, I think, accepts the data that accidents are less common when the two of them are together, but rejects the hypothesis that their proximity affects the accident rate, because it has a very low prior plausibility. That is, it contradicts what we already know about how the world works.

Yes, I agree. But note that you are projecting Sam's response onto our universe, where we have no evidence that anything like the events in the story could ever happen. What we know about how our world works seems to be a misguided presupposition in Sam's world. Of course, having incorrect assumptions is not always irrational.

On a slightly different note, one thing occured to me - note how everyone, both this thread and in the story (myself included) never questioned the direction of causality. Why did Dr. Barry deduce that Sam and Tuesday being in proximity reduces car accident rates, rather than deducing that accidents reduce the chance that Sam and Tuesday will be in proximity? Or is causality bi-directional in this world? If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

[MacArthurBug]

This dosn't replace friction for me, that's for sure. I liked it. However, wher as friction made me... think in new ways- this gave me brief interesting ideas then wandered off distracted. Great potential in this story, but it seemd to go off track somewhere- mayhap that's just me.

[wintermute]

Yes, I agree. But note that you are projecting Sam's response onto our universe, where we have no evidence that anything like the events in the story could ever happen. What we know about how our world works seems to be a misguided presupposition in Sam's world. Of course, having incorrect assumptions is not always irrational.

Arguable. My interpretation of the story is that everyone in the fictional world assumes that it works the same way our world does, otherwise there'd be nothing new or interesting about this study.

On a slightly different note, one thing occured to me - note how everyone, both this thread and in the story (myself included) never questioned the direction of causality. Why did Dr. Barry deduce that Sam and Tuesday being in proximity reduces car accident rates, rather than deducing that accidents reduce the chance that Sam and Tuesday will be in proximity? Or is causality bi-directional in this world? If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

You missed where I pointed out that causation might run in the opposite direction, then?

Well, yes. Causation implies correlation, but that doesn't mean that correlation implies causation. And even if the correlation does imply a causation, it doesn't say which way it goes; perhaps, if the universe is left to itself, the proximity of these two people is caused by the low accident rate rather than the other way around.

[alllie]

If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

I thought the old men riding bicycles was more a result of reduced crime than a cause. Surely old men would be less likely to ride bicycles in a crime ridden area because they would recognize they would be more likely to get knocked off their bikes and robbed in such areas. Or maybe old men riding bicycles automatically act as an unofficial neighborhood watch, calling the police if they see any crime in progress with the result that criminals would learn to avoid neighborhoods where they were common.

I can't figure out how proximity between Samuel and Tuesday would result in less accidents unless one or both of them was consciously or unconsciously causing accidents but when they were together tended to concentrate on each other in a way that defused their accident causing vibes.

[wintermute]

If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

I thought the old men riding bicycles was more a result of reduced crime than a cause. Surely old men would be less likely to ride bicycles in a crime ridden area because they would recognize they would be more likely to get knocked off their bikes and robbed in such areas. Or maybe old men riding bicycles automatically act as an unofficial neighborhood watch, calling the police if they see any crime in progress with the result that criminals would learn to avoid neighborhoods where they were common.

I can't figure out how proximity between Samuel and Tuesday would result in less accidents unless one or both of them was consciously or unconsciously causing accidents but when they were together tended to concentrate on each other in a way that defused their accident causing vibes.

It's magic.

That's why Sam doesn't believe that it can work, and why Tuesday credits it to "the numinous".

[eytanz]

If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

I thought the old men riding bicycles was more a result of reduced crime than a cause. Surely old men would be less likely to ride bicycles in a crime ridden area because they would recognize they would be more likely to get knocked off their bikes and robbed in such areas. Or maybe old men riding bicycles automatically act as an unofficial neighborhood watch, calling the police if they see any crime in progress with the result that criminals would learn to avoid neighborhoods where they were common.

I can't figure out how proximity between Samuel and Tuesday would result in less accidents unless one or both of them was consciously or unconsciously causing accidents but when they were together tended to concentrate on each other in a way that defused their accident causing vibes.

I think you may need to relisten to the story - the fact that unlikely factors reduce bad things in the city without a logical explanation is sort of the whole point. As for the old men in bicycles - they tell Sam quite directly that they ride the bicycle as part of one of the city's other "talisman" projects, and that by them doing so violent crime rates go down.

[wintermute]

If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

I thought the old men riding bicycles was more a result of reduced crime than a cause. Surely old men would be less likely to ride bicycles in a crime ridden area because they would recognize they would be more likely to get knocked off their bikes and robbed in such areas. Or maybe old men riding bicycles automatically act as an unofficial neighborhood watch, calling the police if they see any crime in progress with the result that criminals would learn to avoid neighborhoods where they were common.

I can't figure out how proximity between Samuel and Tuesday would result in less accidents unless one or both of them was consciously or unconsciously causing accidents but when they were together tended to concentrate on each other in a way that defused their accident causing vibes.

I think you may need to relisten to the story - the fact that unlikely factors reduce bad things in the city without a logical explanation is sort of the whole point. As for the old men in bicycles - they tell Sam quite directly that they ride the bicycle as part of one of the city's other "talisman" projects, and that by them doing so violent crime rates go down.

Actually, so far as I recall, the only project where we're given any kind of indication of its success is Sam and Tuesday, where Sam sees Dr Berry's data, which shows a clear correlation across the time of the experiment, suggsting that their being together does reduce accidents.

The old men on bicycles, we're told, are part of a similar experiment, to see if increasing the average age of male cyclists would reduce violent crime; I don't recall anyone commenting on the success or otherwise of that hypothesis, though I may have missed it. Personally, I agree with alllie that there's good reason to assume that the causation goes the other way, so that experiment would be more likely to be shot down in the planing stages. For a higher cost (and therefore more obvious) example, consider that there's a strong correlation between violent crime and the number of people admitted to the local hospital's ER, and then see if shutting down the ER reduces the crime...

An interesting data point I just came across: In US presidential elections, the heavier candidate is twice as likely to be elected than the lighter candidate. Does this mean that Americans prefer fat presidents? Should McCain go on a lard-only diet for the next month to make himself moer electable? Or is it that weight correlates with height, and there is a tendency to prefer tall candidates, all other things being equal? Or is the whole thing just statistical noise resulting from a limited dataset?

[ieDaddy]

If someone goes on a violent crime rampage, would old men all around the city discover that they had flat tires on their bicycles all of a sudden? I wonder...

I thought the old men riding bicycles was more a result of reduced crime than a cause. Surely old men would be less likely to ride bicycles in a crime ridden area because they would recognize they would be more likely to get knocked off their bikes and robbed in such areas. Or maybe old men riding bicycles automatically act as an unofficial neighborhood watch, calling the police if they see any crime in progress with the result that criminals would learn to avoid neighborhoods where they were common.

I can't figure out how proximity between Samuel and Tuesday would result in less accidents unless one or both of them was consciously or unconsciously causing accidents but when they were together tended to concentrate on each other in a way that defused their accident causing vibes.

I think you may need to relisten to the story - the fact that unlikely factors reduce bad things in the city without a logical explanation is sort of the whole point. As for the old men in bicycles - they tell Sam quite directly that they ride the bicycle as part of one of the city's other "talisman" projects, and that by them doing so violent crime rates go down.

Speaking of relisten... The old men riding the bikes are told it's part of a tourism project - Sam makes the jump to them being a talisman project and calls the Doc to confirm it, demanding to know if they are the only experiment or not.  At which point the Doc fesses up that there are a dozen or so projects.

[eytanz]

I think you may need to relisten to the story - the fact that unlikely factors reduce bad things in the city without a logical explanation is sort of the whole point. As for the old men in bicycles - they tell Sam quite directly that they ride the bicycle as part of one of the city's other "talisman" projects, and that by them doing so violent crime rates go down.

Speaking of relisten... The old men riding the bikes are told it's part of a tourism project - Sam makes the jump to them being a talisman project and calls the Doc to confirm it, demanding to know if they are the only experiment or not.  At which point the Doc fesses up that there are a dozen or so projects.

You're right that I was wrong about who tells Sam about the men being a talisman, but if we're going to nitpick each other's descriptions of small details, note that the story itself didn't say whether the men themselves know why they're on bicycles or not. They tell Sam it's part of a tourism project, but it's not clear whether they believe that, or whether it's just a cover story they give out.

[Windup]

Hey! I really liked this one! Yes, it could be more fleshed out but then it is a short story. Wasn't the "luck" statistic thing dealt with briefly in Ringworld?

Birthright lotteries, with Puppeteers manipulating the outcome to breed lucky humans, the end result of which Teela Brown; it's strongly implied that she'll be lucky enough to avoid dying until the heat death of the universe.

At the risk of a thread hijack, I thought Teela turns out to be the least lucky of those with the luck gene; that's why she's the one that winds up getting transformed into the Protector for those that are genetically similar.  Or does Niven change his mind about that in one of the later novels?  I can't remember anymore which was the last one I read.