Probabilities and nincompoops

        Well, now we know why that fellow Mark Ward, attender at the Leeds Conference, exclaimed "Get this... the chances of the comet's data being naturally occuring coincidences is 44 billion to one!" He was citing Richard Hoagland's presentation on Comet Elenin, of course.

        While I was on my Spanish beach, on August 29th, Hoagland went through the pseudo-arithmetic that allowed him to make that claim, on Coast to Coast AM (hour 4.) He did it by cascading the probabilities of a given apparent magnitude, a given date of perihelion, a given anniversary, a given time of closest approach to Earth, and a given orbital inclination. Here are his figures. (CAUTION: Reading this may cause severe nausea and heart palpitation in anyone familiar with mathematics and/or statistical analysis. Proceed at your own risk.)

STEP 1: Apparent magnitude
        The first recorded magnitude, by Leonid Elenin himself, was 19.5. Oh, what a gift for the "theory" of hyperdimensional physics—even though, of course, an astronomical magnitude is totally irrelevant to an angle measured in degrees.

        Hoagland asks "WHAT ARE THE ODDS of a comet having this particular magnitude?" He answers himself "The brightest comet ever recorded had a magnitude of -17. The dimmest was +28.2. That's a range of 45.2. So the odds are 1 in 45.2."

STEP 2: Date of perihelion
        Hoagland still insists that the date is September 11th even though it's the 10th everywhere except Japan and Australia. He says the odds of that date occurring are 1 in 365.

STEP 3: Anniversary
        Even though he has the wrong date, he still insists that the attacks on 9/11/01 are important. He says that fact that this is not JUST ANY anniversary, but the tenth, allows him to factor in odds of 1 in 10.

STEP 4: Time of closest approach to Earth
        The time is 19:50 UTC on October 16th. Hoagland confuses the time of 19:50 with the decimal 19.5, but let that pass. He asks WHAT ARE THE ODDS of that particular time occurring and, multiplying 24 x 60, he arrives at 1 in 1440.

FINAL STEP: Orbital inclination
        Elenin's inclination is 1.84°. Hoagland figures the odds of that occurring are 360/1.84 or 1 in 195. (and of course he can't resist pointing out the magic 195 again.)

The odds are...
        Hoagland multiplies 45.2 x 365 x 10 x 1440 x 195 and arrives at 45,000,000,000. Stunning. Or not.

What's wrong with this picture?
        I feel sure many readers of this bloggery are ahead of me, and WAY ahead of Hoagland. Let's see what would happen if the figures were different. How about if the magnitude were 14.6, the date of perihelion were 3rd July, the anniversary were the 20th, the time of closest approach were 07:21 and the inclination were 2.02°?

Let's see... tap-tap-tap-tap... Oh, tens of billions.

        How about if the magnitude were -2.0, the date of perihelion were 10th June, the anniversary were the 21st, the time of closest approach were 21:21 and the inclination were 11.7°? Wow, tens of billions again.

        Hoagland has fallen slap bang into the ex post facto fallacy that so many other untrained and ignorant people have fallen into before him. This is like throwing snake eyes in a casino and claiming that your throw is unique in human history. It's true that you only have one chance in 36 of throwing that result, but the odds of the other 20 possible results are also long. WHATEVER you throw has a low probability—until you throw it, then the probability is 100%. Likewise, the parameters of Elenin have to have some values, and multiplying them out as though they were unique is going to yield nothing useful.

        Every single time someone throws a pair of dice, a long shot is pulled off, in other words. Using Hoagland's own perverse ideas, every single time a comet comes through, an event whose odds are 40 billion to one against also happens.

        Are you listening, Mark Ward? IT'S RUBBISH, MARK. HE'S A NINCOMPOOP AND YOU WASTED YOUR MONEY.