Archive for the ‘skepticism’ Category

God, atheism, and emotions

Friday, April 26th, 2013

Earlier today on FriendFeed, Dr. Ganata shared an article from Pacific Standard magazine: “Emotional Reactions of Atheists May Reveal Echoes of Belief”.

header of article

I found myself having my own emotional response to the article and the corresponding study, and I wanted to set down my thoughts.

First of all, there’s no widespread agreement on labels. To some, an atheist is anyone who does not believe faithfully in the existence of God. By that definition: People who are unsure or neutral or agnostic on whether or not God exists? All atheists. But to others, an atheist is someone who is convinced that God definitely does not exist — and may spend energy trying to convince the faithful that they are wrong. Even among the faithful, there is a spectrum of how strongly and completely one believes in God. So there is a wide range of possible beliefs among individuals.

The study organizers recruited from a skeptics organization, from a bible group, and from students. They don’t break down how many applicants they got from each source. They divided participants between “atheists” and “religious people” (those are their labels, not mine), based on their answers to a 19-question survey. While there was a sharp difference between the scores of the two groups, they don’t make a distinction between atheists who once were faithful, or if the atheists are “weak atheists” (those who simply are not sure in the certainty of the existence of God) versus “strong atheists” (those who are positively, 100% convinced that there is no God).

Second, the study design is garbage. The overall approach is measuring skin conductance (“SC”) while participants read statements and then say them aloud. There’s no data on what SC means, no proof of what it measures, no theory about what emotions it’s tied to, and the study doesn’t bother to provide the raw data to back up any statistical claims.

In study 1, the participants each were given three sets of statements:

  1. “God statements”: Statements where one dares God to do something awful, such as “I dare God to make me die of cancer” or “I dare God to make someone murder my parents.”
  2. “Offensive statements”: Random offensive things, such as “It’s okay to kill ugly children” or “It’s okay to kick a puppy in the face.”
  3. “Neutral statements”: Random inoffensive things, such as “I hope it’s not raining today” or “It’s okay to wear glasses.”

In study 2, they add a fourth category:

  1. “Wish statements”: Repeats of God statements but restated as a wish, such as “I wish that I would die of cancer” or “I wish someone would murder my parents.”

However, study 2 is very poorly thought out. The phrasing of the statements is not at all equivalent. Worse, they’re having the same person react to the same statement both from category 1 and category 4 — but the emotional response will be much greater the first time you read it than the second, whichever way it’s phrased. And most disastrously, study 2 is only of atheists, so for cateogry 4, you can’t compare the SC of atheists to the SC of religious people.

Third, the study lacks sufficiently documented controls. The two groups of study 1, atheists and religious people, were selected based on their responses to survey questions. How many people applied? How many were rejected? Why are there no participants with middling responses?

A much better study design would have three groups: Atheists, religious people, and a random group used as a control with unknown belief status. If the study’s central thesis is true (that both atheists and religious people will have a strong emotional reaction to category 1 statements), all three groups should have same result.

And I would design the questions with much more diversity of control statements. “I wish that I would die of cancer” is just outright much less controversial than saying “I dare God to make me die of cancer.” Add in “I dare Zeus to make me die of cancer” and “I dare my best friend to make me die of cancer.”

My final objection to the study is that the statistical results of study 2, with only atheists participating, shows a huge change in the responses to categories 1 through 3 when compared with the results of study 1. Their SC went down for categories 1 and 2, and up for category 3. That’s ridiculous! The study designers could not even reproduce their own results from study 1 with the exact same methodology. That shows their candidate screening or test administration or SC measurement is fatally flawed. How much variance would we see again if they conducted a study 3 or a study 4?

Set aside all that. Let’s say the test is valid and the results are genuine. What does it mean?

The article and study do take pains to point out several possible explanations for the result why both atheists and religious people have a similar emotional response to reading the God statements:

  1. “Atheists’ explicit beliefs may differ from the implicit reactions that exist outside of conscious awareness.”
  2. Atheists “may have found using the word God stressful because others, possibly their friends and family, do take God seriously.”
  3. Atheists “may have found the idea of God ‘absurd or aversive,’ leading to the heightened emotional response.”
  4. “Although atheists did not currently believe in God, they may have been influenced by their own previous beliefs.” (A reference is made to 2006 research that found three-quarters of American atheists were once believers.)

At heart, I had to think about why I had an emotional response to reading the article and the study. Because I self-identify as an atheist, the first conclusion was offensive to me: Hey, you, you atheist you, you’re deluding yourself — you may say you don’t believe in God, but your emotions betray you, ha ha. You really do believe in God. Sucker.

It was hard for me to be objective about that.

But why?

We all delude ourselves all the time. We’re full of self-delusions that affect us constantly, without us even being aware of it most of the time. I know this and accept it. So why would it upset me for me to feel that I might have an emotional belief in God?

I have to admit if I were to take the test, I’m sure I would have a hard time being unemotional in reading aloud a statement that said I dared God to kill my parents or give me cancer. Again: But why? Is it because I secretly do believe in God? I don’t think there’s a chance that that’s true. I think actions and expressed beliefs are actually more significant than emotional responses.

No matter what, the thought of having you or your parents die is going to provoke an emotional response, whether or not God is involved in that thought.

I don’t pray. I don’t ascribe supernatural explanations to events that take place. I don’t go to church. I wouldn’t turn to any holy book for answers. And while I respect my friends who do believe in God, I don’t feel that their religious beliefs give them any power or morality that I don’t also have.

Bottom line, I don’t think this study proves anything. The whole point of atheism is to separate one’s emotional beliefs and irrational human nature from our actions and logical thoughts.

I think most atheists would agree that they don’t KNOW that God doesn’t exist. The way I feel about it is this: I assume God doesn’t exist until the moment I see some evidence that God does exist.

I can’t prove God doesn’t exist. I wouldn’t even try. Similarly, I can’t prove that the Loch Ness Monster doesn’t exist. So, since I can’t prove the opposite, I’m happy to concede that God MIGHT exist. (So might the Loch Ness Monster.) I just don’t see any evidence for it, so I don’t personally think it’s at all likely.

Despite that, why would I take the risk in daring an all-powerful God who MIGHT exist to kill me or my parents? On a rational basis, there’s no point in tempting fate, even if the risk is minute. My emotional side knows that better than my rational side.

I’d like to see the study redesigned with a big twist: Per the study, there were some religious people who refused to say some of the God statements. But there were no atheists who refused to say any of them. But what if you paid people to say the statements? If you move the study from the emotional realm of skin response into the rational realm of getting paid, would you see the opposite result, where even religious people are willing to say the same statements as the atheists? Or is it offensive to even ask?

Maybe emotionally we’re all believers to some extent, but when money is on the line, some of those who identify themselves as religious people might be willing to act a little bit like an atheist.

This is too long already, but I want to conclude with an apology: If you were offended by any of what I wrote, I am sorry. I don’t mean to challenge anyone or offend anyone. Your beliefs are yours, and I don’t have any intention to try to change them. My whole intention is to explain myself, mostly so I can better understand myself.

Recently, Barry paid me a high compliment on FriendFeed: He called me a damn fine Christian. That makes me damn happy to hear. Because even though I’m a heathen, I strive to be a moral, helpful, useful, good person. In other words, a Christian. Or a Catholic. Or a Buddhist. Or a Muslim. Or an atheist. Theist or atheist, emotional or unemotional: We’re all here to try to make the world a better place. Right?

“Blücher” is NOT the German word for glue (my whole world is a lie)

Tuesday, November 9th, 2010

The 1974 Mel Brooks comedy Young Frankenstein is one of my favorite movies. Starring Gene Wilder, Marty Feldman, Peter Boyle, Teri Garr, Madeline Kahn, and Cloris Leachman (with an uncredited cameo by Gene Hackman), the spoof of the original 1931 Universal Studio’s Frankenstein is both hilarious and well-made, standing up to repeated viewings.

[A photo Cloris Leachman as Frau Blucher peering through a doorway]Sunday evening I happened to be discussing the film with some friends, including someone who spoke German. I mentioned how much I liked the joke about the horses whinnying off-stage whenever they heard the name of Cloris Leachman’s character, Frau Blücher, being uttered, because it was German for glue.

“But the German word for glue is not ‘blücher,'” my friend Mattias said.

“Oh. Well, what is the German word for glue?” I asked.

“You could say ‘der Klebstoff’ or ‘der Leim,'” he replied.

“Well, what does ‘blücher’ mean?”

“It’s a name, it doesn’t mean anything.” (Apparently it’s a common name, too, like ‘Jones.’ EDIT: Per the comments, no, it’s not common.)

Well, I had heard that the reason the horses whinny throughout Young Frankenstein is because they were afraid of being turned into glue for a long time, from at least two different people, starting at least 20 years ago.

A quick search confirmed the debunking: Snopes, About, even IMDB. Wikipedia expanded that Cloris Leachman herself had heard it from Mel Brooks. In an interview with Brooks, he claims that someone gave him the wrong translation: “Before we started shooting, someone told me ‘blücher’ means glue, so that’s why I had the horses whinny. I’m not sure if that’s true.” However, in the audio commentary, Brooks simply says that the horses whinny because she’s an ominous character.

There are millions of people who speak German throughout the world. It’s tremendously easy to look up German words for things thanks to tools such as Google translate. But here I was a couple of nights ago, repeating an urban legend. We generally tend to believe things that we’re told, even when verification is simple. The moral: Don’t believe everything you hear. Verify things yourself.

For over 20 years I believed the word “blücher” meant glue. Now it means disillusionment.

Lying with Charts 102: Deceptions of stack

Tuesday, August 31st, 2010

It’s time to lie with charts!

A few months ago I posted a FriendFeed item about how deceptions of scale can be used to make misleading charts.

Earlier this month, Wired argued in an article that “the web is dead” and tried to prove their point with a rather curious chart: A stacked area chart with the changing percentage of the Web’s share of total Internet traffic between 1990 and 2010.

Peer-to-peer and video take up a large share of total traffic in 2010 — naturally, because these files are huge, and web pages are relatively modest in size. A more honest chart would look at total volume of use (eyeballs/visits). If you read a tweet on and then watch a video on Netflix, it’s not at all fair to assume the video was 3,571,429 times more important to you than the tweet.

But even setting aside the stupidity of conflating the importance of contents with its file size, a stacked percentage area chart is a visual deception. There are many ways to lie with charts, and stacking is one of the most frequent.

Here’s a simple example. Suppose your company sells three products, Wobjects, Dooders, and Flozzels. You have sales data for a year. It turns out that Wobjects are growing fast, and Flozzels are selling pretty well, but Dooders are sucking wind with sales volume decreasing. If you want to hide how poorly Dooders are doing, just use a stacked chart.

First, the pretend data:

Product Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Wobjects 67 69 72 76 80 82 82 88 88 90 94 103
Dooders 70 70 69 68 68 64 67 63 66 66 66 62
Flozzels 42 43 43 44 46 51 54 58 57 58 60 61
Total 179 182 184 188 194 197 203 209 211 214 220 226

The most deceptive chart first:

Example 1: Stacked bar

A stacked bar chart serves to obscure individual trends

Certainly if you know what you’re looking for you can eyeball the decrease in sales from Dooders. But for most people only the overall growth would jump out.

Almost as deceptive is a stacked area chart:

Example 2: Stacked area chart

A stacked area chart is almost as good as a bar chart for obscuring individual trends

This is slightly less deceptive because the eye is slightly better able to see that the middle area is decreasing in size over time.

To be more straightforward, use a line chart with a total:

Example 3: Line chart with totals

A line chart is not very deceptive for individual trends. Including the total makes the scale larger, hiding the amount of increase/decrease for individual series, but is valid if the overall total is the most important detail to convey.

The least deceptive approach would be a simple line chart with no totals:

Example 4: Line chart

The most honest approach for showing the performance of individual product lines is a straightforward line chart.

And to make a total mess of things, use a stacked percentage area chart, with gratuitous 3-D:

Example 5: Stacked percentage area chart

Stacked percentage charts are usually disastrous, because the relative share of three product lines is often totally irrelevant compared to the actual change in value. And 3-D is just a distortion.

Celebrity deaths: A statistical analysis

Tuesday, June 1st, 2010

Twice before I’ve written about the “celebrity deaths come in threes” superstition, in 2008 and 2009.

With the recent passing of Art Linkletter, Gary Coleman, and Dennis Hopper, this superstition has again resurfaced.

I feel my previous arguments have already been quite persuasive, but now let’s add a statistical debunking.

To analyze the superstition, we need to define it. That includes two tasks:

  1. Who is and isn’t a celebrity
  2. The timeframe for the deaths to occur

I extracted the data of all 1,422 celebrity deaths that have occurred between January 1, 1995 and May 31, 2010 from a site called, which is the location of a death pool contest. (The contest has entrants predict which celebrities will die in the upcoming year and assigns points for correct guesses. Last year’s winner took away over $3,000.)

This addresses the first question, who’s a celebrity. At they have a panel of judges determine whether or not a person who passes away is famous, based on simply whether or not five or more members of the panel have heard of the person. They then create a list of celebrities ahead of time, and then monitor that list to see who has passed away.

You may well disagree with the fame assessments of Certainly the data included plenty of people I personally had never heard of. But it’s a list that exists independently of the superstition, and is pre-existing, so it doesn’t suffer from the selection bias that arises when you assess whether or not a person is famous only after they have died.

As for the timeline, I decided to analyze it with as much leeway as possible. One day between each death? Up to two days? Three? Five? Seven? Who knows. I analyzed with a number of tolerance days all the way up to 10.

Before we get into the numerical analysis, let’s visualize the data.

[A timeline chart showing all celebrity deaths from 2004-01 to 2010-06, using data extracted from

(Click to enlarge; depending on your browser, you may need to click again to view at 100% and then scroll from left to right)

As you scroll back and forth in the listing of deaths from 2004 through today, your mind can certainly pick out groups of three. But is it ALL groups of three? Is it even MOSTLY groups of three? Your eyes already tell you the truth, that of course it’s not.

The numbers back up that visual refutation.

There are quite a few ways to analyze the data, and I tried to be comprehensive. Here are the approaches I took:

  1. Rolling timeline: This is probably the best method. (It was suggested by Patri.) When a death occurs, I start a counter. The counter lasts up to x days. (I analyzed with x from 1 to 10.) I keep track of how many celebrity deaths occur within that period. The counter resets after x days, and starts again whenever the next death occurs. With x at 7, for example, it’s basically an analysis of how many deaths a week, using rolling weeks.
  2. Continuous grouping: When a death occurs, start a count. Look at the next death. Is it within x days? If so, increment the counter. If not, start over at 1. Again, I analyzed with x ranged from 1 to 10.
  3. Separate tests: For each death, I calculate if it’s part of a group by looking at the date of death of the first member of the group, and see if it’s within x days of the last death. For the first death, it should be more than x days. For the subsequent deaths, it should be within x days. I then judge “pass” or “fail” for each death. I applied this analysis to groups of 1, groups of 2, groups of 3, groups of 4, groups of 5, and groups of 6. I also let it “roll” by varying where I started the counter. This analysis also looked at x ranging from 1 to 10.

So, what are the results?

For rolling timeline, we see the following results:

Tolerance Days (x) Groups of 1 Groups of 2 Groups of 3 Groups of 4 or more
1 75.7% 19.0% 4.6% 0.6%
2 47.6% 35.3% 12.4% 4.6%
3 28.3% 40.8% 20.0% 10.8%
4 18.8% 39.8% 23.6% 17.8%
5 12.7% 31.9% 23.8% 31.6%
6 9.7% 26.0% 24.7% 39.6%
7 7.5% 22.9% 25.1% 44.5%
8 6.0% 18.0% 22.6% 53.4%
9 4.7% 14.9% 20.5% 59.9%
10 3.7% 12.1% 19.4% 64.8%

No matter how many days of leeway you give, groups of three never actually best explain the data. If you give a lot of leeway, such as 10 days, larger groups occur. If you give only a little leeway, most deaths happen alone or in pairs.

The best performance for groups of three is when you allow a leeway of 7 days, but even then the superstition fits for just 25% of the deaths. (Groups of two deaths are not far behind, at 23%.) A superstition that’s only right one time out of four — and does no better than several of variants of the superstition — well, that’s not a useful superstition.

So, for this methodology, groups of three never really succeeds. With 7 or more days of leeway, three is the average and median for groups of deaths, but only with a 23% success rate. No interpretation of this data with this method would lead one to agree that celebrity deaths come in threes.

For the second method, continuous grouping, the results are similar. You can get some pretty big groups with this method — using three tolerance days, the largest group turns out to be a group of 21 celebrity deaths. And with 10 tolerance days, the largest group is of 243 deaths!

However, no matter how many tolerance days you allow, groups of three never amount to more than 14.3% of all groups. So at best, groups of three explains about 1 death in 7 with this method.

The third method I used was separate tests. To be honest, this is a pretty stupid method, since if, say, two deaths in a group of three fit the pattern but one doesn’t, it still scores as two out of three when really the entire group should fail. And the groups are highly dependent on previous groups, so if there’s a missing celebrity or a person included who isn’t really a celebrity, it throws off the entire test.

Under this method, groups of three still score very poorly. No matter how many tolerance days you allow, from 1 to 10, it always turns out that some other grouping (such as groups of 2 or groups of 6) beat out groups of 3. Groups of 3 performed best with 10 days of tolerance, but with that high a tolerance, groups of 4, 5, or 6 fit even better. At most, 64% of celebrity deaths would pass a group of three test but at the same time 72% fit a group of 4.

The data, analysis, and chart are all available for you to examine (Google docs share, 6.6 megs, Excel format).

If you asked me, the best method is the rolling timeline method, and the most reasonable number of days of tolerance is three. Going with that, we find that, on average, the group size is 1.7.

But “Celebrity deaths come in 1.7s” doesn’t have a winning ring to it.

Zeigen’s credo

Friday, November 13th, 2009

Credo is Latin for “I believe.” These are my personal beliefs. Everyone has different perceptions and beliefs, and I do not offer my credo as an insult or to attack anyone else, but only as an exercise of putting my beliefs into words so that I can better understand myself.

[Photo of Sammy Mack at Stanford Mall, November 6, 2009]

I believe in my kids.

  • I believe that people should be treated with respect, no matter what beliefs they hold. I may disagree with certain beliefs, and even try to convince another person to change a belief, but I will always try to respect the individual, no matter how much I disagree with their beliefs. (There is an exception for believers in hatred or violence; I find it very difficult to respect holders of those beliefs.)
  • I believe in following a moral code, based on one’s understanding of right and wrong, and I believe in treating others as I would wish to be treated.
  • I believe in the scientific method, that theories and claims should be tested, and that beliefs should be based on testable and reproducible evidence. I believe there are no immutable truths and that everything should be up for debate.
  • I believe “faith” is defined as having a certain belief despite there being no evidence for that belief. Because of my skeptical world view and my requirement for evidence to support my beliefs, I believe that “faith” has little place in my life.
  • I believe that I am completely open to believing in the existence of God (or gods). If I were to find any proof that God exists, I would believe in God. I believe that the burden of proof of God’s existence should be on those who believe in God, not on those who don’t. By some definitions, this makes me an agnostic, but I don’t really believe in labels.
  • I believe that the more extraordinary the claim, the more rigorous should be the proof. Belief in a benevolent creator as a conscious entity who watches over us and influences events for us is an extraordinary claim, or so I believe.
  • There are several arguments for creator belief that I do not find persuasive.
    1. I am not persuaded by arguments along the lines that all things have a creator, therefore our universe was created. Who, then, created the creator? The same argument that others make to me that our universe must have had some “prime cause” I would return to them, and ask what was the prime cause for that prime cause.
    2. I am not persuaded by the extraordinary unlikelihood of life forming on our planet as proof that there was a creator of that life. Deal out a deck of cards. The odds of that particular hand being dealt were tiny. But it happened, and after it happened, the odds were 100%. Deal enough hands and you increase the likelihood of that hand being dealt to the point where it becomes likely. Well, I believe there are a lot of planets in our universe, and I believe that we happen to live on one where life happened to form.
    3. I am not persuaded by words in a book put down by human hands as any kind of absolute proof of anything in particular, especially when the book in question has had multiple authors and revisions and a long history of mistranslations. (If you are insulted by this, please don’t be. Maybe I’m not talking about YOUR holy book, maybe I’m talking about someone else’s.)
    4. I do not find persuasive any third party descriptions of impossible events or miracles, especially if they happened long ago, unless they have been credibly witnessed or recorded or reproduced.
    5. Because I have never seen a credible study proving that prayer has benefits (and I have seen many that disprove any benefits), I do not believe in the power of prayer. How does God choose which prayers to answer? If one person prays for one event to happen, while another person prays for that same event to not happen, how is that resolved?
  • I tend not to believe in absolutes or extremes, but instead look at life as a full spectrum of possibilities.
  • I believe our brains and perceptions are often deeply flawed, and we have unbelievable power to fool ourselves.
  • I believe that every individual is different, and do not expect my own beliefs to influence others or be persuasive. Other individuals have different beliefs based on their different values and world views, and I believe that that’s what makes life interesting. The world’s religions and varied cultural history hold enormous value and beauty.
  • I believe that a refusal to be tolerant of other people’s different beliefs is problematic. I respect people for strong-held beliefs, but some belief systems are incompatible with my world view, and I may choose to not have such people in my life, and I believe that some people with extreme beliefs should not hold positions of power or authority over others.
  • I fully believe in the separation of church and state.
  • I am by nature suspicious of most organizations, and that applies to organized religious organizations as well. I believe in “live and let live” and therefore do not care for extreme proselytizing, or dogma that dismisses or attacks other groups.
  • I do believe in groups that support each other and their community with acts of charity, whether those groups are religious or not.
  • I believe that I should try hard not to be a hypocrite. But I believe that I am a flawed individual, and that my actions may not always be consistent with my beliefs. But I believe I should always try to be consistent and try to improve myself.
  • I believe in kind actions and kind words. I believe in not taking oneself too seriously. I believe in love. I believe it’s time to eat.

Data on U.S. airline crash fatalities: Is 2009 a terrible year for air travel?

Wednesday, September 9th, 2009

A few minutes ago, I saw the following news flash from Breaking News Online jump across my screen:


My immediate reaction was, “Wow, what an awful tragedy, and what a terrible year this is for American airplane safety.”

But once I saw the details behind that alert, and learned that fortunately it was only a crash between two small planes with only one fatality (instead of the hundreds I feared), I wanted to check my impression that we were having a bad year.

I found a site that tracks airplane crashes, and filtered their database for U.S.-only crashes.

In 2009, there have been 6 incidents so far. All told, 81 have died this year in plane crashes, and the year is not yet over. Everyone is probably familiar with the events in New York on January 15, 2009, where, thanks to the heroic landing in the Hudson by Captain Chesley B. “Sully” Sullenberger, 149 lives were saved following an engine strike by birds. Less than a month later, a horrible accident during bad weather on February 12, 2009, in New York claimed 49 lives. And a little more than a month after that, on March 22, 2009, another terrible accident in Montana killed 14. Then things were quiet until August 8, when a helicopter accident in New York led to 9 more fatalities.

Together, those four incidents created an impression in my mind that 2009 was a terrible year for air travel.

However, then I looked at the data for the previous four years.

In 2008, there were 15 accidents recorded in the U.S. The worst was on August 5, 2008, in California, and caused 9 fatalities. In total, 60 airplane accidents caused 60 fatalities last year.

For 2007, there were 11 accidents causing 46 fatalities, the worst being a crash in Washington state killing 10 in October.

In 2006, there were 9 accidents killing 65 air travelers, and the worst was a crash in Kentucky that killed 49 people in August.

2005 was a relatively safe year, with 7 accidents causing 37 deaths (20 from a crash in Florida in December).

Charted data (from for 2005-2009 showing U.S. airplane accidents: Number of incidents and number of fatalities.

Charted data (from for 2005-2009 showing U.S. airplane accidents: Number of incidents and number of fatalities. Click to enlarge.

(I shared the spreadsheet on Google docs if you want to see the raw data.)

So, while 2009 has seen some terrible and tragic accidents, it’s not really the case that it’s orders of magnitude worse than previous years. It’s true that there have been more fatalities this year (and there are still three months to go until 2010), but the number of accidents was actually higher in the previous three years.

Airplane travel remains much safer than car travel and other travel in terms of passenger miles. The III analysis of NCHS statistics for 2005 shows that your odds of dying in a car crash in a year are 1 in 6,539 while in a plane it’s 1 in 502,554 (compared to the odds of dying in a “cataclysmic storm” which are 1 in 339,253).

Fallacy of the excluded middle

Friday, July 3rd, 2009

So many times we believe that there are only two choices, that something is either good or bad, and there is nothing in between.

As Hamlet said in Act II, scene 2, “There is nothing either good or bad, but thinking makes it so.”

But this is a fallacy, and it has a name: The fallacy of the excluded middle. Here’s an example:

  • Too much water, and you drown.
  • Not enough water, and you die of thirst.

Therefore, water is always a bad thing and we should avoid it, right?

“You’re either with us, or against us,” I’m told. Not necessarily. Perhaps I support some of what you do, some of the time, in my own way.

Humans can’t achieve perfection, but nothing is ever a total failure either. I find that everything is somewhere in the middle.

Celebrity deaths do not come in threes

Thursday, June 25th, 2009

domA year ago, I wrote this post attempting to debunk the superstition that deaths come in threes.

With the passing of Ed MacMahon on Tuesday, along with Farrah Fawcett and Michael Jackson today, I’ve seen this superstition resurface. Yet my arguments from a year ago still stand.

I’d also like to add this refutation: Dom DeLuise died alone.

Take this list from Wikipedia of celebrities who died in May. I’d argue that Dom was the “most famous” of all the names listed there, but please feel free to assert differently if you disagree. So, where are the other two, if deaths do indeed come in threes?

We can repeat the exercise for other months.

RIP, Ed, Farrah, Michael and Dom. Please don’t cheapen their memory repeating a baseless superstition that tries to find a pattern where none exists.

UPDATE June 28th: Billy Mays has also passed away today, breaking the pattern for even the current three.

Influenza A(H1N1) cases: graph of WHO data, discussion of media coverage

Thursday, June 4th, 2009

In the next 30 years, seismologists have determined that the chance for a magnitude 6.7 or later earthquake in California is over 99%. One can easily see based on the historical record that this is a safe prediction. In the last several hundred years, Californians have not had a period of 30 years go by without such an earthquake.

Similarly, influenza outbreaks follow a regular pattern. Roughly every 50 years one can expect a influenza pandemic that kills a million or more people worldwide. (What’s less frequently cited is that every year, the regular flu kills up to 350,000 people worldwide, mostly the elderly.) In recent history, the Hong Kong flu of 1968 killed up to 1 million people. The Asian flu of 1957 killed up to 1.5 million people. And the 1918 Spanish flu outbreak killed up to 40 million people worldwide.

It’s understandable, then, why people were concerned with swine flu and why there was so much media coverage. What if it turned out to be one of those catastrophic pandemics that kills millions of people? What if we couldn’t (or can’t) create an effective inoculation? It’s rational to be cautious, and to pay attention to the news.

On the other hand, in recent years we’ve seen coverage of previous threats from SARS, avian flu, and others, and they turned out to be relatively minor events. In some cases we overreact. It certainly seems to me that mainstream media overhyped the threat from swine flu and created a sense of panic that was unwarranted from the facts. Now that coverage has faded, I’m sure most people aren’t really thinking about swine flu any further.

The thing is, it’s not really over. Worldwide cases have climbed to nearly 20,000, and the number of deaths is over 100. In the U.S., there are currently 17 fatalities. (The media really only covered the first two victims.)

From here, swine flu could fade away into nothing, or it could suddenly explode. The latter outcome seems increasingly outcome, however. And here’s some data to support why I believe that.

The World Health Organization (WHO) has been releasing data for 41 days; over that time, they’ve provided 43 updates as I write this. Earlier they released two updates a day. Currently the updates come every few days.

WHO update swine flu case data, including number of cases, deaths, and cases per day

WHO swine flu case data, including number of cases, deaths, and cases per day (click to enlarge)

The graph above shows the number of new cases reported per day, as a line. In addition, as an area, the number of confirmed cases (yellow) and confirmed deaths (red) are also included.

By profession, I’m a numbers guy. Quantifiable analysis is my preferred approach for investigation. So starting from the beginning, I monitored the numbers of confirmed cases, countries, and deaths from the WHO updates and analyzed how many cases per hour, day and week.

With a pandemic, what we worry about is exponential growth (100 cases becoming 200 cases becoming 400 cases becoming 800 cases). Even linear growth can be scary for a pandemic if the numbers are large enough (50,000 cases becoming 100,000 cases becoming 150,000 cases). But that’s not what we have here, judging by the WHO’s numbers.

In fact, the number of new cases per week has been steady for the last three weeks at about 4,400 new cases a week.

There are many disclaimers that should be associated with the WHO’s numbers. Who knows what politics play into the process, and who knows if the count of confirmed cases is constrained by laboratories crushed with tons of undiagnosed cases that are even now uncounted. And the number of as-of-yet unreported cases (especially from third world countries) is a total unknown. You can tell from the extreme variability in the number of cases per day that human factors influence the reports.

However, after looking at nearly five weeks of data, it’s easy to assert what the media has already decided — there’s no exponential growth, and not even significantly increasing linear growth. Note instead how the cumulative number of cases appears to be very smoothly linear.

Given the relatively small number of cases compared to other threats, it’s clear the news cycle for swine flu is dead and not returning unless something completely unexpected happens, despite scattered reports of ongoing school and business closings.

So before this story fades into the sunset, the question to ask is if the media and health workers reacted appropriately. Was the coverage sufficient or insufficient, was it overblown or underblown or exactly right?

The June issue of Lancet contains a story (reprinted here) arguing that the amount of coverage was proper, and that due to quick action from health workers and cooperation from a fearful public, a crisis was averted.

I’m not qualified enough to disagree, but it seems to me the most important factor was the nature of the swine flu itself, and just how contagious it was. The evidence shows that it just didn’t spread that quickly — it wasn’t that virulent.

But it’s quite possible that because of the coverage, we put in place behaviors that saved ourselves. By analogy, consider Y2K (where we worried that computer code that used 2-digit years instead of 4-digit years would lead to buggy behavior when the software assumed the year 1900 instead of 2000, and that these issues would affect critical facilities and cause widespread technological disasters). There was enormous media coverage. Speculation was rampant, including fears of widespread power failures and nuclear facility mishaps, and some predicted wholesale societal breakdown. Yet of course when January 1, 2000, rolled around, very little happened (other than some big hangovers). So, overblown, right? Maybe not. I know first-hand how much effort engineers and developers put in ahead of time to certify certain systems, reprogram others, and generally make sure that everything would continue to work. To an outsider, Y2K certainly seemed overhyped. And much of the speculation (including the concept that embedded chips in cars and toasters would malfunction and shut down) was in fact ridiculous. But most computer professionals know that, while the coverage was certainly hysterical at times, there were instances of genuine bugs (that could have affected paychecks and so on), and that most of these genuine issues were averted due to foresight, prudence, and hard work.

I’m no medical professional. While I immediately dismissed swine flu fears and coverage as overblown, maybe that’s because I’m an outsider, not seeing all the hard work that took place to make sure the disaster was averted before it became a deadly pandemic.

Aside from those very unfortunate people who died due to swine flu, in the end, the economic impact may be swine flu’s longest-lasting legacy. Several reports show that Mexico tourism dropped by huge percentages, even in regions where there were zero swine flu cases. Recovery to previous levels will take time. (I’m told there are some amazing travel bargains to Mexico now.)

EDIT: Based on feedback from Kevin Fox, I updated the graph to simplify it a bit, and to use just regular calendar dates instead of dates and WHO update number. That corrected the problem whereby the variable number of days between updates made the cumulative number of cases look to be accelerating.

What top soccer players tell us about astrology

Monday, June 30th, 2008

Nearly every newspaper carries a horoscope column. Almost everyone knows what “sign” they are. Most people do not take astrology too seriously, reading their horoscope for amusement if they read it at all. However, some people take pains to study the characteristics of the different signs and make assessments of people based on what sign they are, and attempt to model behavior or predict the future based on astrology. Others pay significant money to astrologers for a personalized chart. A 2003 Harris poll found 31% of U.S. adults stated they believed in astrology.

A simple question to ask someone who believes in astrology is why it works. What method do the planets or stars have of influencing one’s behavior, personality and future? It’s certainly not gravity, since the doctor who delivered you had a larger gravitational effect on you than Pluto did. (Assuming the doctor weighs 5kg and was 5 centimeters away from you; Pluto weighs 7.15×109kg and was at least 2.76×1014 centimeters away. Plus it’s not even considered a planet anymore.)

The good thing about astrological claims is that they’re testable. If someone says that Aries are supposed to be fearless and impulsive, one can design a survey and then check if those who answer the survey about impulsiveness who are Aries answer the questions differently. There’s a fair bit of research into the claims of astrology, and the most significant debate centers around the so-called Mars Effect, which claims those born during times Mars is ascending are more likely to excel at sports.

A fair amount of research seems to confirm that birth month has a significant correlation with excellence in sports.

Victory for astrology? Not so fast.

In the last few years, research into “relative age” has shown interesting results. Let’s start with soccer. Each soccer club and soccer camp has an age requirement. Imagine, for example, a summer soccer camp that requires the campers to be nine years old when the camp starts in June. So a kid born in May nine years ago will barely be able to make it in, while a kid born in September nine years ago will have to wait a year. It turns out that the “older” nine year olds tend to do much better in camp. Since they’re older, they’re generally more coordinated and can run faster and longer — which makes them tend to be picked first, which gives them more self-confidence. That early experience often seems to carry through the rest of their soccer career. This chart, for example, from a University of Alberta study, shows how world cup youth soccer players born in the first three months after the eligibility cut-off blow away those born in the other nine months.

I have little doubt that relative age affects a lot more than just sports. Parents tend to want to push to have their children moved up a grade, but it may be the exact opposite approach (thus having your child be among the oldest in the class) will have profound benefits that affect your child throughout his or her life.

I certainly believe astrology is junk. But I also believe we should pay attention to the research showing that the month of a child’s birth is actually quite important.

Shakespeare was a debunker

Tuesday, April 29th, 2008

Henry IV Part One, Act III, Scene 1:

Owen Glendower: I can call spirits from the vasty deep.

Henry “Hotspur” Percy: Why, so can I, or so can any man;
But will they come when you do call for them?

Here, Hotspur neatly refutes his cousin-by-marriage Glendower’s pretentious brag with a fine skeptic’s rebuttal. I was reminded of this exchange from this case from last month: A so-called tantrik employer of black magic claimed to be able to kill any man with a curse within three minutes, and the challenge was taken up on live TV.

Debunker: Nine dead at the Dyatlov Pass

Saturday, March 15th, 2008

Read this Mark Morford column about the 1959 Dyatlov Pass accident. (I’m waiting. Waiting. Waiting. Oh, done? Great!)

Unexplained things happen all the time. We humans work hard to try to understand the universe, but the universe is vast and our minds are only so good at figuring things out.

But just because something is unexplained, why is that taken (as Mark seems to) as proof of the supernatural?

Doubtless we will never know what happened to those hapless nine hikers back in 1959. But here’s one thing I’m sure of: It wasn’t aliens. I’m also sure it wasn’t bigfoot, a yeti, the abominable snowman, vampires, ghosts, telepathic reindeer, telekinetic yaks, or any of a dozen other proposed paranormal theories.

As Carl Sagan argued in Cosmos, “Extraordinary claims require extraordinary evidence.” The only evidence in Dyatlov is that something happened we don’t yet understand.

I like a good science fiction novel or movie as much as the next guy, and as interesting and exciting as it would be for supernatural things to be real, at this point, we all can be very certain that none of that stuff is real.


Because if any of this baloney were genuine, don’t you think, with all the cameras we have — video cameras, closed circuit cameras, cell phone cameras, satellite cameras — someone would have recorded some genuine proof by now? Just how is it that vampires (plus ghosts, psychic powers, and all the rest) have managed to go undetected for so long?

There are many plausible non-supernatural explanations for what happened at the Dyatlov Pass, including exaggeration by those reporting the incident, solar flares, an avalanche, altitude sickness in one or more of the hikers, and so on. Will we ever know the truth? Most likely never. But one cannot take a genuinely mysterious event, such as this one, and present it as “proof” of the supernatural — that’s too far a jump.

I’ll believe in psychic powers, ghosts or other aspects of the supernatural when there is reliable documentation of something supernatural that actually happened, rather than lack of knowledge of what did happen.

What they’re up to

Wednesday, January 30th, 2008

Yesterday Kimi took Sophie to her four-month checkup.

  • Weight: 13 pounds 4 ounces — 40th percentile
  • Length: 24 1/4 inches — 45th percentile
  • Head circumference: 16 inches — 37th percentile

By coincidence, at Sammy’s school it was measurement day yesterday, and he was reported as 36 inches, average for his class. Looking at the official CDC growth charts, turns out that that’s only slightly above average for a boy of 28 months. Sammy had always been taller than average previously, but looks like either we aren’t feeding him enough, or the averages have caught up to him.

Sophie also got three shots yesterday. I know quite a few people are worried about the link between vaccines and autism (and there’s the Eli Stone controversy stirring things up, with a pilot episode about that issue set to air tomorrow).

I don’t know any doctors or scientists who believe the link between vaccines and autism. Skeptical coverage was recently the cover story of the Nov/Dec Skeptical Inquirer. I was going to write up a debunker post with a summary of the refutation, but it turns out my friend Geoff did a great job on this already.

Debunker: “Death comes in threes”

Wednesday, January 23rd, 2008

With the news yesterday of the untimely passing of Heath Ledger, I’ve heard several people remark on the frequently-expressed belief that deaths (like sneezes?) come in threes. Since Suzanne Pleshette recently succumbed to cancer (on January 17), some wondered who will be the “third” actor/actress to die. Others who were familiar with Brad Renfro‘s death (on January 15) expressed instead the idea that Heath’s overdose was the completion of a series of three.

I have to wonder about this superstition. On one hand, it’s not falsifiable. People die all the time. Since the belief doesn’t express a timeframe for how close in succession the deaths have to occur, eventually three people in a related field will die, and the “prophecy” of death coming in threes will be fulfilled.

Any belief that isn’t falsifiable contributes nothing. Suppose ten musicians were to die in quick succession starting next week. Adherents to this belief would group them into three groups of three deaths, then posit after the tenth that two more musicians would die — and eventually (once we wait long enough) indeed two more would. So the prediction has come true! But the actual number of people dying doesn’t change the superstition (they’d believe the same thing even if the number was two, eight, nine, ten, or twenty), and the superstition doesn’t actually predict anything — just that over time people do, in fact, die.

Secondly, who or what is the agent that enforces the trio of deaths? If I learn about two bloggers who die, does that knowledge somehow make me more likely to get into an accident? Is there a grim reaper with a quota of three who (like in the movie series Final Destination) sets up a fatal chain of events for arbitrary professions?

Monitoring a list of recent deaths to see if the number is divisible by three turns up the quite predictable result that the number is only divisible by three about a third of the time.

In conclusion, this belief is an example of selective perception; You tend to remember the times when there was a grouping of three seemingly-related deaths in a short period of time, which reinforces the belief, but tend to forget the times when there wasn’t a pattern.

As I’ve said before, our brains are remarkably good similarity detectors (whereas computers are excellent difference detectors; this is why captchas work), so we often find ourselves picking out plausible-seeming patterns to events that ultimately are chaotic and unpredictable.

And on a personal note, I thought Heath was a gifted actor with a great career in front of him, and I was saddened to hear the news of his death.

Security Theater

Friday, December 21st, 2007

It’s my opinion that anyone who has grown up watching heist movies or reading mystery novels is capable of dreaming up two or three dozen methods of circumventing most security measures employed by the TSA at airports.

All of those hoops we jump through — removing shoes before tiptoeing through x-ray machines, powering up laptops to show they’re real, drinking a bit of your infant daughter’s formula to show it’s not a deadly poison — are, in my opinion, nothing more than so-called “security theater” — a well-choreographed show that is designed to make us feel safer, but bears little relationship to actual security.

An article in tomorrow’s British Medical Journal comes to the same conclusion, while highlighting the staggering costs.

At heart, airport safety agencies seem to rush to implement new screening tests based on publicity around a perceived threat, without any scientific rigor or analysis of if the screening method is effective. A skeptical approach is called for, one that examines the evidence of the threat and then designs the least intrusive and most cost effective method of controlling for that threat.

I imagine that if a terrorist organization manages to smuggle bombs in their underwear or invents an explosive substance made out of cotton, we’d all be required to travel naked.

Our family is next flying in March. If you’re flying this holiday season, enjoy the security show. It costs you $9 of each ticket you purchase. Why, that’s practically the cost of a heist movie.

Are miexd up wrdos esay to raed?

Thursday, November 30th, 2006

You may well have seen e-mails or articles along these lines:

Don’t delete this just because it looks weird. Believe it or not, you can read it.

I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdanieg. The phaonmneal pweor of the hmuan mnid. Aoccdrnig to rsreeach at Cmabrigde Uinervtisy, it deosn’t mttaer in waht oredr the ltteers in a wrod are, the olny iprmoatnt tihng is taht the frist and lsat ltteer be in the rghit plcae. The rset can be a taotl mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe. Amzanig, huh.

Well, I think the above text (when you unscramble it) is making too big a claim that “it doesn’t matter” what order the interior letters of a word are in, and that you can read such passages “without a problem.”

There are certainly quite a few other sites that have talked about this claim:

The easiest way I can think of to show the claim is dubious is to have you test yourself. Here are three sentences I scrambled using a quick program I wrote to randomize the interiors of words:

  • Docotr Sumgnid Fuerd neevr desimailcd reguioils idaes needlslsey.
  • Riandeg tihs txet is claetniry pbosilse — hevweor, i’ts aslo cmuotpe innevtise.
  • Enre’yveos potohn tdrpooeos mriuclsoulay elxedopd cetmpelloy utesirnsed.

Clearly, the longer the word, the harder it is to unscramble it. The shorter words in this article’s title are easy to decode, but the longer words in the three example sentences, especially less frequently used words like “disclaimed” or “compute” or “photon,” are probably not immediately recognizable.

Now you try it!

Enter some text for scrambling:

Want the original version of my three scrambled sentences? Keep reading.


Losing luggage

Thursday, March 23rd, 2006

I heard a news report on the radio and tracked down the source: According to an IT company called SITA that apparently handles baggage tracking for the airline industry (wow, there’s a job description for you — “I’m Senior Director of Tracking Down Lost Luggage Data Processing”), 30 million bags are mishandled annually, and 204,000 of those are lost outright, out of 2 billion passengers.

They don’t say how many bags are carried by those passengers (they’re too busy flogging their services), so let’s assume each passenger checks an average of 1.5 bags. So out of 3 billion bags, 1% are mishandled, and 0.68% of those (or less than 0.01% of the total) are lost.

Those figures are a good match to my perception and my experience. I believe I’ve flown more than 100 times (50 trips over 38 years), and I’ve had my luggage delayed exactly once — and it showed up within the 31 hours they say is average. But, I have met people with a very different perception — it seems to me that the common perception is 5-10%, resulting in behavior changes (people using bulky carry-ons rather than checking bags purely for fear of losing the bag rather than because it’s faster) as well as unnecessary stress.

I think the perception difference is due an error of weight on the incidents. Everyone knows someone who has had a bag delayed or lost, and many people have experienced it themselves. If I know 1,000 people who have flown in an airline one or more times, and 100 of them have at least once had a lost or delayed bag, then the simplistic calculation is that this is a 10% problem. But those 1,000 people have probably flown 100,000 or more times, and 99% of the time they haven’t had an issue.

Because comedians and others may play up the idea of how frequently bags are lost, that reinforces the perception that this event is more common than it is. So it becomes a PR issue — the airlines should be able to combat that perception. But they can’t, because when it comes down to it, 1% delayed bags is a terrible stat. On an average flight of 600 passengers (assuming perfect distribution), six bags will be delayed.

99% may be good enough for some metrics, but if our servers at work had only 99% uptime we’d be in trouble; if airlines only landed 99% of planes safely, very few people would choose to fly.

In conclusion, the problem seems to be less bad than it’s perceived to be, but I’m surprised by how bad a problem this is.

Debunker: 10% of your brain

Friday, February 10th, 2006

I admire greatly the life of Carl Sagan. His “Cosmos” series is a wonder, and he dedicated his life to fighting pseudo-science and popularizing science. While I’m not cut out to be a school teacher, I respect those who are (including my sisters-in-law Kelly and Erin). I’ve worked as a trainer, and it has always seemed to me that educating, mentoring and debunking is a very high calling in life.

In the November/December 2003 Skeptical Inquirer, an article written by Carl’s wife, Ann Druyan, gave me a lot to think about:

“Congress cut off federal fnding for SETI years ago. I was with Carl when he went into Senator William Proxmire’s office after Proxmire had given the Golden Fleece Award to the SETI program. Carl sat down with him. I didn’t say a word. I was just a witness. And I just watched Carl. I was inspired by him, by not only the breadth of his knowledge but his patience, his lack of arrogance, his willingness to hear the other person out. Senator Proxmire did a complete turnabout as a result of that meeting.

“And there were other instances of Carl’s remarkable persuasiveness. One was a great story of a so-called “creation scientist” who watched Carl testify at a hearing about creationism in schools. Carl testified for about four hours. It was somewhere in the South, I can’t remember where. And six months later a letter came from the “creation scientist” expert who had also testified that day, saying that he had given up his daytime job and realized the error of what he was doing. It was only because Carl was so patient and so willing to hear the other person out. He did it with such kindness and then, very gently but without compromising, laid out all the things that were wrong with what this guy thought was true. That is a lesson that I wish that all of us in our effort to promote skepticism could learn, because I know that very often the anger that I feel when confronting this kind of thinking makes me want to start cutting off the other person. But to do so is to abandon all hope of changing minds.”

As a former high school and college debater, in my experience Dr. Druyan is right: It’s very hard to convince anyone with harshness, derision, or ridicule.

While I don’t have the training or expertise that Carl had, or the time to put together a site such as Snopes (one of my favorites), or the resources to make a television show such as Penn & Teller’s Showtime series “Bullshit,” or Adam and Jamie’s “Mythbusters,” I will from time to time try to do some debunking of my own here.

To begin, I’d like to discuss a claim that I hear repeated quite often: That we only use 10% of our brains.

Many times, the source of this claim is someone who is trying to sell something that will help you “unlock” the other 90% of your brain, or someone (such as a fortune teller or medium) who claims that they have special powers due to using their whole brain.

In short, there is no science behind this claim. I have three arguments:

  1. MRIs and other brain scans reveal that all of us use almost all of our brains throughout the day. It’s true that at any given moment, not all of the neurons in our brain are firing (and perhaps that’s the source of this claim). But it’s not true that a significant portion of our brains is continuously fallow.
  2. Patients who undergo brain trauma (such as a piercing accident or non-fatal gunshot wound) almost always suffer from some signficant loss of function. When there is brain damage (from, say, years of being a boxer, or a disease) similarly you can witness the effect. If 90% of the brain were not needed, there would be far more cases of strokes, anuerysms, and brain damage having no effect. The myth doesn’t reconcile with the reality we witness.
  3. Simple evolution: Having recently witnessed the birth of my son, I can tell you first-hand that the biggest cause of difficulty during birth is due to the size of the infant’s head going through the birth canal. The birth process would be so much easier if the brain were 90% smaller and the head could be correspondingly reduced in size. Also, think how much of our day-to-day expenditure of calories is solely to keep our heart pumping in order to power the brain with oxygen. Why would there be all this difficulty and energy associated with something that was 90% unused? The selection pressure on having smaller heads and smaller brains would be signficant, because if only 10% was needed, there would be a giant advantage to having a smaller brain. But, the reality is that we succeed by having big brains, and by using all or mostly all of it.

Was that convincing? Hopefully so. Now I see Snopes’ page debunking this claim. Hmm, that’s a pretty good article. Clearly I need to do some homework for my own debunking efforts.

Next time you hear someone make this claim, perhaps now you will gently, and with respect, give them your own counter-arguments. Be proud that you use 100% of your brain!