Numbers are hard

Tom Mitchell tells us (#10827):

"we know that at its very simplest statistical level, the odds of those two things happening by chance alone were 1-in-37,000".

What two things is he talking about?

Well, as far as I can make out:

1. Work being carried out in a house once lived in by James Maybrick, including the floorboards being lifted, on 9 March 1992; AND
   

2. Mike Barrett telephoning a literary agent to offer them the diary of Jack the Ripper which would turn out to have been written by James Maybrick.

Right, so we can see that there is a coincidence there, with both things happening on the same day, but how does Tom get to 1 in 37,000?

Well, although he subsequently claimed  (#10837) that, "I didn't mention the date of Maybrick's death", he then went to mention the date of Maybrick's death, explaining that:

"The only cold, hard mathematical equation was March 9, 1992 minus May 11, 1889 in terms of number of days."

That is 103 years, of which Tom said

"the 103 years I did mention was illustrative of the implausible nature of the double event which unfolded on March 9, 1992"

Tom believes that 103 years needs to be converted into days, which apparently amounts to 37,000 days, so that, by virtue of two things happening on 9 March 1992, this means, in his mind, the "odds" of those two things happening are 1 in 37,000.

I'm going to ignore the fact that a calculation of odds, which is something very specific that bookmakers do, is not done in this way. Let's assume he is talking about the chances rather than the odds. Is he in any way correct?

The answer is that no, he is not. The date of 11 May 1889 can have no possible bearing on whether the two events which occurred on 9 March 1992 were a coincidence, occurring by pure chance, or were connected. This can be demonstrated very simply.

As, I assume, even Tom accepts, the diary could be a fake. It could, if one ignores the language anachronisms which prove the contrary, be a fake that was created shortly after May 1889, perhaps later that year or the next year. Or it could be a fake that was created during the 1930s or 1960s or even on 8 March 1992. In all cases, after having been created, it could have been placed under the floorboards of Battlecrease to be discovered by electricians on 9 March 1992 and immediately sold to Mike Barrett later that day, meaning in all cases that there was no coincidence in the fact that Mike made a telephone call on the same day offering it for sale because the two events must have been connected.

But if the diary was a fake created (whether placed under the floorboards or not) on 8 March 1992, how could the date of Maybrick's death, more than one hundred years earlier, possibly factor into a calculation of whether or not Mike's telephone call on 9 March 1992 was a coincidence in respect of the electrical work on that day? For, in that case, there would be only one day separating the creation of the diary from its discovery.

The only sensible answer is that it can't have any possible bearing on it.

In fact, the date of Maybrick's death cannot have any bearing on the chances of coincidence, if the diary was a fake created at any time after 11 May 1889.

I mean, if the diary was created in the 1890s, the 1930s or the 1960s what possible bearing can the date of Maybrick's death of 11 May 1889 have with any calculation of coincidence as to what occurred on one day in 1992?

lf the diary is a fake, but was nevertheless found under the floorboards of Battlecrease on 9 March 1992, the coincidence of Mike telephoning a literary agent to offer that same diary for sale on that same day is exactly the same as if the diary was genuine, regardless of when it was actually created (and, in the scenario I've just outlined, there is no coincidence involved at all because the two events must be connected). The date of Maybrick's death can play no possible part in any calculation of the chances of whether the two events happened by coincidence or not. It's utterly irrelevant.

I might add that if the diary is a fake created between 1 and 13 April 1992, the date of Maybrick's death also cannot possibly factor into a calculation of whether or not Mike's telephone call on 9 March 1992 was a coincidence.

Where Tom has confused himself is that he seems to have started on the basis that the diary is genuine and thus couldn't have been created after 11 May 1889. But of course it could. The only issue with which mathematics can theoretically be said to relate to is whether the diary is connected with the electrical work carried out on 9 March 1992. It can't assist as to whether the diary is genuine or not. Although it's my own personal belief that if it had been found under the floorboards it must be genuine, that is not a mathematical certainty and cannot be factored into a mathematical calculation of probability.

If one starts from a strong belief that the diary wasn't a fake created after Mike's telephone call on 9 March 1992, as Tom Mitchell does, one will feed all kinds of nonsense into the formula to calculate "the odds" but Jeff Hamm, an actual expert in statistics, who Tom expressly invited to comment on his analysis back in 2021, has already demolished Tom's nonsense. It's kind of incredible that Tom has seen his arguments taken apart by people who know what they are talking about but simply repeats them three years later.

Of course, as RJ Palmer explained in the Casebook thread, the lifting of the floorboards also cannot enter into the calculation because, even if it was a fact that they were lifted on 9 March 1992 (about which no actual evidence, as opposed to inference, has yet been provided), it remains nothing more than a theory that the diary was found under the floorboards. It could equally have been found hidden in a different location in the house. Or it could have been found in a completely different building a few days earlier before being passed on to Mike Barrett. Or it could have been forged after Mike heard some electricians discussing their Battlecrease work on 9 March 1992. Or the forgery could have had nothing to do with the electrical work done in Battlecrease.

NUMBERS REALLY ARE HARD

A reminder of Tom Mitchell's proficiency in maths:

MEMORY IS ALSO HARD

Tom Mitchell said (#10869):

"You see, RJ has made the same mistake which Orsam previously made (with his leaves falling off trees analogy)."

WTF? No leaves falling off trees analogy has ever been made by Lord Orsam. I don't even know what he's talking about. The guy must live in a perpetual state of confusion.

LORD ORSAM

2 May 2024