Saturday, November 01, 2008

Three Bloggers

On the left, yours truly. In the center, T. Ryan Gregory of Genomicron. On the right, the world's most famous science blogger, P. Z. Myers of Pharyngula. We were all in Guelph for P. Z.'s talk sponsored by CFI and the University of Guelph Skeptics.

The talk was well-attended, as P. Z. and his daughter discussed a variety of different subjects, including science education, the upcoming election, and strategies for fighting the foolishness of creationism and intelligent design.

The highlight for me, however, had to be the fellow who during question period insisted that there had to be something outside scientific inquiry, and gave as his prime example (and no, I am not making this up) the 2004 World Series victory by the Boston Red Sox. He claimed that the Sox's improbable finish, including victory during a total lunar eclipse, was proof of supernatural intervention. The majority of the audience laughed, because I suspect they know what I know: that the 2008 World Series Victory by the Phillies is the ultimate, unquestionable proof of a deity.


JP said...

I am still laughing over lunar eclipse guy's comments. Wow.

T Ryan Gregory said...

Sadly, I am really only photogenic in person.

S.M. Elliott said...

Well, in LOST, the Others used video footage of that anomolous Red Sox victory to prove they had off-island contacts, so I guess it does qualify as evidence in some areas. ;)

Garkbit said...

Looking at the three of you together, I think I see some sort of theme or commonality ...

Anonymous said...

The best part was when the lunar eclipse guy said "I am a pretty reasonable guy" and the entire audience laughed.

will m said...

What is the actual probability of a team coming back from a 3-0 standpoint to win a series and that coinciding with a Lunar Eclipse? How often would we expect that to happen per century?

Jeffrey Shallit said...

Will M:

One can easily estimate this probabiity, but before I do so, let me point out the logical flaw here: during any sporting event, some extremely improbable event is likely to happen. Maybe a pitcher produces a perfect game (happened in 1956), or maybe a pitcher hits a home run (happened this year) or maybe there is an earthquake during the game (happened on October 17, 1989). Why don't we calculate the probability of these events, show it is extremely low, and then conclude these other events had a supernatural cause?

Because the space of odd events is extremely large, and during any World Series, some odd events are likely to happen. You can't arbitrarily pick and choose the space of events that concerns you after the fact and come to any reasonable probability conclusion. I can pick an entirely different set of events and come to a different conclusion. This is the same conceptual error made by people pushing intelligent design.

That having been said, let's look at the probability that a team comes back from a 3-0 deficit. Even this simple calculation is fraught with difficulty. First, we need to know the relative strengths of the two teams: if they played a very large number of games against each other, what is the rate at which one teams beats the other? Of course, we don't know that, because a World Series typically matches teams that have hardly ever played each other, and because personnel changes make even those data suspect. But let's estimate this by saying the teams are evenly matched, so it is 50%. Next, we need to know exactly what probability you are talking about: is it Pr[winning World series| already 3 games behind] or is it Pr[winning World series by losing the first 3 and then winning the next 4]? I'll be generous and assume you had the second in mind.

Then the probability of 7 specified outcomes, each with 50% probability, is 2^(-7), or 1 in 128.

Now, how about the probability of a lunar eclipse? Actually, they're surprisingly common: there are between 4 and 7 eclipses a year, and approximately half of them (on average) are lunar. Furthermore, a lunar eclipse is visible from most of one hemisphere. So, roughly speaking, from any given location, you can see approximately 1 lunar eclipse per year. (That's assuming perfect weather.)

So the probability of both events is approximately
(1/128)*(1/365), or about 1 in 50,000. As far as unusual events go, this is not very unusual at all. A really remarkable coincidence would have probabiilities on the order of 1 in a billion or more. See Warren Weaver's book on probability for a catalogue of some truly remarkable coincidences with far lower probabilities.

But again, the real question is, why did you restrict yourself just to those two events? You could have made the probability as low as you like by factoring in even more events. (To get your desired result, you can even ignore independence of events.) After all, the 2004 Sox victory was the first in 86 years; that's a low probability event right there.

All these calculations are, to my mind, bogus.

Anonymous said...

Wasn't the example of the Boston Red Sox winning the World Series brought up by Kenneth Miller in his testimony in the "Dover" case?

Tom S.

will m said...

He did use it. Here's the quote from the transcript:

Q. So supernatural causation is not considered part of science?

A. Yeah. I hesitate to beg the patience of the Court with this, but being a Boston Red Sox fan, I can't resist it. One might say, for example, that the reason the Boston Red Sox were able to come back from three games down against the New York Yankees was because God was tired of George Steinbrenner and wanted to see the Red Sox win.

In my part of the country, you'd be surprised how many people think that's a perfectly reasonable explanation for what happened last year. And you know what, it might be true, but it certainly is not science, it's not scientific, and it's certainly not something we can test. So, yes, those rules certainly apply.

Shane Caldwell said...

I saw good evidence at this talk of why I don't want to get too involved with the Skeptic movement. While someone's views may be odd and outlandish, if they do no harm, then I don't feel it's appropriate to ridicule that person. I came out of the meeting feeling that the others there were cold, arrogant and cruel.

Jeffrey Shallit said...


Maybe you could explain why you thought Lunar Eclipse Guy was being ridiculed, as opposed to his claims.

Look at the very first comment here: JP is "laughing over lunar eclipse guy's comments". Not laughing at him, laughing at his comments.

Shane Caldwell said...

Fair enough. I guess I just found it difficult that he was being respectful and acknowledged that his view was unorthodox, yet from the beginning no one really heard him out because he presented a view that opposed that of the community, regardless of the logical flaws. He had a valid question that no one really seemed to address because of the distraction.

His question was along the lines of "How do you explain the existence of things that science and reason cannot explain?" and perhaps presented a bad example.

I would have liked to hear an answer along the lines of "Of course there are things we cannot explain at present, but we believe that if there is no scientific evidence at present, in the future there will be." and discuss how you cannot prove something is correct in science, just determine the most likely explanation. This reasoning has not yet been proven wrong, and if it were the scientific establishment would fall apart.

Unfortunately everyone got too hung up on the flawed logic of the example and never addressed the real question.

Note that I understand his example about the world series is easily explained through probability and chance, but to answer his actual question about unknown phenomena, I would hope the above answer might be appropriate.

Jeffrey Shallit said...


You may be right - I don't recall the first part of the question at all. I don't think it's fair to blame people because the questioner had a ridiculous example. A ridiculous example undermines one's argument, and it's only natural that the bigger picture would be forgotten.

As Martin Gardner once said, "One horselaugh is worth a thousand syllogisms."