One Incontrovertible, Unequivocal, Undeniable Fact Which Refutes the Diary

[rjpalmer]

my statistical analysis is unimpeachable and is accurate, I promise you.

Hi Ike

As discussed yesterday, I think debating this matter further would be futile.

Still, if you will hear me out, last night I came up with a simple exercise that I think will help you visualize the error of your thinking.


1. On 9 March 1992, a sketchy Scouser named Mike Barrett contacts a literary agent with what purports to be the Diary of James Maybrick. The last entry reads, “I place this now”… and gives a date of 6 May 1889.

Later, it is realized that on the same day that Barrett contacted his agent, Maybrick’s old bedroom floorboards were lifted—an astonishing coincidence, it seems.

As our resident statistician, you calculate the odds thus: Maybrick died on 11 May 1889 (the last date he could have stashed the diary) and Barrett came forward on 9 March 1992---which is 37, 558 days later, counting an extra day when “Leap Year” occurs.

Thus, we are told, it is 37, 558 to 1 that the two events would occur through mere ‘coincidence.’ A fact that should shake any Ripperologist out of his or her complacency!!


2. Now let’s make some small, irrelevant changes, for the sake of clarity. On 9 March 1992, Mike Barrett shows up at a literary agent’s office in Rome, Italy, with what purports to be the diary of the Emperor Caligula. The last entry states, “I place this now…”

Later, it is realized that on that same date that Mike came forward with Caligula’s scrolls, the Emperor's bedroom chambers were excavated for the first time on Palatine Hill, and the flagstones lifted. Caligula died on 24 January A.D. 41. The span between that date and 9 March 1992 is 712, 635 days.

Thus, according to your methods, it would be 712,635 to 1 that those two events would coincide thru mere coincidence! Nearly twenty times more unlikely than in the case of James Maybrick's diary!


3. Final example. On 9 March 1992, Mike Barrett shows up in New Haven, CT, with what purports to be the secret diary of the American actress Sandy Dennis. The last entry states, “I place this now…”

Later, it is discovered that on the same date that Mike came forward, the floorboards in Dennis’s old bedroom in New Haven were lifted for repairs. Only this time, Dennis died only seven days earlier, March 2, 1992. (I looked it up).

According to your methodology, the odds of these two events coinciding through mere chance is only 7 to 1.


Seriously, Ike. You can’t see the flaw in your thinking?!?


The circumstances of these three events are identical.

Barrett comes forward with a sketchy diary of a famous person that has not been authenticated, but has occurred on the same date that their floorboards were lifted. The odds of this happening by chance should be the same. Yet, using your flawed methods, the “odds” range from a tepid 7 to 1, all the way up to a staggering 712,635 to 1.

Clearly, there is something seriously, seriously amiss with your reasoning.

The simple fact is that there is nothing wrong with your ability to count and to do simple arithmetic. You are simply analyzing the wrong data.

You are also basing your calculations on the ASSUMPTION that the diaries are genuine, and that you know where they came from, neither of which are true.

If Caz came up with a theory that the diary was a hoax created by Stephen Knight, and stashed under Maybrick’s floorboards on the last day of Knight’s life (25 July 1985 ) you’d be calculating the odds of this having occurred randomly on 9 March 1992 at 2420 to 1.

Clearly, the THEORY has changing the odds! Run that one by your statistician friend and hope that his head doesn't explode.

In short, get a grip. It’s bonkers.

Now, take the time to go back and re-think Lord Orsam’s calculations and try to fathom why they are far more legitimate. And please keep in mind that the diary’s origins are NOT known. You cannot calculate the odds from a position of faith.

With all good wishes

[Iconoclast]

Hi Ike
As discussed yesterday, I think debating this matter further would be futile.
Still, if you will hear me out, last night I came up with a simple exercise that I think will help you visualize the error of your thinking.

I haven't read it yet but I am 100% certain you will fail to find any error in my thinking.

Before we start, if we know that two events occurred on the same day (and - for sake of arguments here - neither had occurred on any earlier date), then the probability of their occurring on the same day is calculated as follows (in every case of two events 'randomly' occurring on the same day where neither occurred on any earlier day):

What was the first day they could have happened on the same day?
What was the final day they could have happened on the same day?
The sum of all possible days becomes the denominator.
The number of times the event occurred becomes the numerator (as we defined it as neither having occurred on any day earlier, this must be 1).

1. On 9 March 1992, a sketchy Scouser named Mike Barrett contacts a literary agent with what purports to be the Diary of James Maybrick. The last entry reads, “I place this now”… and gives a date of 6 May 1889.

Later, it is realized that on the same day that Barrett contacted his agent, Maybrick’s old bedroom floorboards were lifted—an astonishing coincidence, it seems.

As our resident statistician, you calculate the odds thus: Maybrick died on 11 May 1889 (the last date he could have stashed the diary) and Barrett came forward on 9 March 1992---which is 37, 558 days later, counting an extra day when “Leap Year” occurs.

Thus, we are told, it is 37, 558 to 1 that the two events would occur through mere ‘coincidence.’ A fact that should shake any Ripperologist out of his or her complacency!!

What was the first day they could have happened on the same day? May 12, 1889
What was the final day they could have happened on the same day? March 9, 1992
The sum of all possible days becomes the denominator. 37,618 (as I calculated it using Excel)
The number of times the event occurred becomes the numerator (as we defined it as neither having occurred on any day earlier, this must be 1). 1

Therefore, the likelihood that these two events (which could have happened separately on any day between May 12, 1889 and Mar 9, 1992) occurred on the same day by sheer chance alone is 1/37,618.

2. Now let’s make some small, irrelevant

The changes you are proposing are NOT irrelevant ...

changes, for the sake of clarity. On 9 March 1992, Mike Barrett shows up at a literary agent’s office in Rome, Italy, with what purports to be the diary of the Emperor Caligula. The last entry states, “I place this now…”

Later, it is realized that on that same date that Mike came forward with Caligula’s scrolls, the Emperor's bedroom chambers were excavated for the first time on Palatine Hill, and the flagstones lifted. Caligula died on 24 January A.D. 41. The span between that date and 9 March 1992 is 712, 635 days.

Thus, according to your methods, it would be 712,635 to 1 that those two events would coincide thru mere coincidence! Nearly twenty times more unlikely than in the case of James Maybrick's diary!

Correct. I am trusting that you have calculated your denominator correctly. If you have, then there was approximately 20 times more days than the first example so the the probability is 20 times more unlikely (to have occurred by mere chance). Therefore, the probability of these two events occurring on the same day is 1/712,635.

3. Final example. On 9 March 1992, Mike Barrett shows up in New Haven, CT, with what purports to be the secret diary of the American actress Sandy Dennis. The last entry states, “I place this now…”

Later, it is discovered that on the same date that Mike came forward, the floorboards in Dennis’s old bedroom in New Haven were lifted for repairs. Only this time, Dennis died only seven days earlier, March 2, 1992. (I looked it up).

According to your methodology, the odds of these two events coinciding through mere chance is only 7 to 1.

What was the first day they could have happened on the same day? March 3, 1992
What was the final day they could have happened on the same day? March 9, 1992
The sum of all possible days becomes the denominator. 7
The number of times the event occurred becomes the numerator (as we defined it as neither having occurred on any day earlier, this must be 1). 1

Therefore, the likelihood that these two events occurring by chance on the same day is 1/7.

Seriously, Ike. You can’t see the flaw in your thinking?!?

I can't see the flaw in my thinking because there is none.

The circumstances of these three events are identical.

Only if the same length of time had passed in which the two events in each example could have happened.

Barrett comes forward with a sketchy diary of a famous person that has not been authenticated, but has occurred on the same date that their floorboards were lifted. The odds of this happening by chance should be the same.

If I wasn't so bored of this my-ego-won't-let-me-be-wrong-so-I'm-just-going-to-wear-you-down-with-my-utter-shite, I would literally piss myself. I'm not sure if this conclusion is actually funnier than your dreadful 1/18 calculation.

Yet, using your flawed methods, the “odds” range from a tepid 7 to 1, all the way up to a staggering 712,635 to 1.

As you would expect when the situations are so different.

Clearly, there is something seriously, seriously amiss with your reasoning.

If you say so.

The simple fact is that there is nothing wrong with your ability to count and to do simple arithmetic. You are simply analyzing the wrong data.

You are also basing your calculations on the ASSUMPTION that the diaries are genuine, and that you know where they came from, neither of which are true.

The only issues I have factored in were given above. Genuine or hoax is irrelevant to the likelihood that these two events would happen by chance on the same day. I trust that you are understanding what we have been talking about, yes? We haven't been discussing the likelihood that these events would happen. We have simply backtracked from our knowledge that they did happen - and on the same day - in order to do a bread-and-butter calculation of the likelihood that (as they had happened on the same day) what were the odds of that occurring 'coincidentally' (that is, without connection). If you are as clever as you think, you will spot that I have just handed your ego a Get Out of Jail card here.

If Caz came up with a theory that the diary was a hoax created by Stephen Knight, and stashed under Maybrick’s floorboards on the last day of Knight’s life (25 July 1985 ) you’d be calculating the odds of this having occurred randomly on 9 March 1992 at 2420 to 1.

The principle doesn't change if Caz, Kaz, Baz, or Jaz produced a diary by Will, Bill, Phil, or Gill on the same day the floorboards were raised in Tutankhamun's tomb on any day you choose. Nor on his tailors, accountants, or cleaners tombs. None of this matters at all. All you need to know are:

What was the first day they could have happened on the same day?
What was the final day they could have happened on the same day?

Clearly, the THEORY has changing the odds! Run that one by your statistician friend and hope that his head doesn't explode.

The only thing that has changed are the number of days on which the two events could have occurred simultaneously.

In short, get a grip. It’s bonkers.

If you say so.

Now, take the time to go back and re-think Lord Orsam’s calculations and try to fathom why they are far more legitimate. And please keep in mind that the diary’s origins are NOT known. You cannot calculate the odds from a position of faith.

Ooh, will I Won't I?

What do you think, dear readers?

Ike
Lord Awesome

[harry]

What do you think,dear readers?
I think I am convinced.

[rjpalmer]

I haven't read it yet but I am 100% certain you will fail to find any error in my thinking.

Before we start, if we know that two events occurred on the same day (and - for sake of arguments here - neither had occurred on any earlier date), then the probability of their occurring on the same day is calculated as follows (in every case of two events 'randomly' occurring on the same day where neither occurred on any earlier day):

What was the first day they could have happened on the same day?

Ike, you have fallen at the first hurdle.

You do not KNOW when the first day "they could have happened on the same day" because you do not know where the diary came from, who wrote it, nor when it was created. It’s an unknown.

You are formulating your “denominator” based largely on your belief that the scrapbook is likely to be the genuine diary of James Maybrick’s and that it was placed somewhere in Battlecrease as early as March 11th 1889.

But since you are forced to make your calculations from a position of ignorance (because no one knows or at least agrees to the diary’s origins), it is entirely possible that your ‘earliest’ date is entirely wrong and flawed and an impossibility---yet you insist on using it.

In a sense, your calculations aren't so much wrong, as they are invalid, and incorrectly formulated.

In other words, by setting the first possible date as May 12, 1889, [strangely, Maybrick actually died on the 11th], you are, in effect, irrationally insinuating that it WAS placed under the floorboards on that date, and then you are simply counting forwards until you reach March 9, 1992 (the date of Barrett's call) which gives you the number 37,618, which you then convert to 37, 618 to 1.

No. No. No.

You are adding hundreds and thousands of what could be entirely imaginary ‘possible’ days to your denominator without having the faintest idea where the diary actually came from. And, because you are basing your calculations on an assumption, you are left with no way of knowing whether or not your resulting ‘denominator’ is accurate.

Your calculations are a group of goblins dancing on the head of an imaginary pin

But I’ve tried many ways to express this, Ike, and, with a heavy heart, I must give up.

Someone else will have to carry the baton, and I pray for your sake that they do not whack you over the head with it.

Good luck.

[Abby Normal]

Abandon all hope ye who enter here

[erobitha]

It is no shock that I understand and support Ike’s analysis. I also have a sneaking suspicion that he might be more qualified in the area of statistics than RJ or Lord O.

Perhaps RJ you should consider that.

[Iconoclast]

Ike, you have fallen at the first hurdle.

You do not KNOW when the first day "they could have happened on the same day" because you do not know where the diary came from, who wrote it, nor when it was created. It’s an unknown.

Statistics is primarily the analysis of probability and - at its simplest level (which we are at, by the way) - probability calculates from the premise of possibility.

There are more esoteric calculations which attempt to factor in the unknown but - for obvious reasons - they produce less reliable results. Interestingly enough, this is not because the statistical tests themselves are unreliable. I once wrote an essay on The robustness of the t-test in the face of the violation of one or more of its assumptions. The literature (at that time) seemed to show with some level of confidence that you could violate any number of the t-test's assumptions (for example, that the sample derives from a population demonstrating a normal distribution curve) and it would still produce very reliable results regardless.

What we are talking about here is simply Statistics 101. Given what we know for certain, what is possible? That's how we derive our denominator. What we don't get to do - unless we do not know what we are doing or if we are simply being mendacious - is to add in variables for which we do not have known data or exclude certain facts because we don't like the look of them. So, by way of an example, we can't exclude Sundays or Bank Holidays just because we have a hunch the 'double event' could not have occurred on one of those days (or we can but we have to alter our hypothesis to explicitly state this so that there are no surprises when our denominator falls sharply and therefore our probability increases sharply). Nor do we get to say we can't start the analysis because we don't know when variable two (someone contacting a literary agent over the mooted diary of Jack the Ripper) 'started'. As I was at pains to say in my previous post, it doesn't matter one jot if the diary is authentic or a hoax. We do not need to know that information to calculate what were the possible days it and Maybrick's floorboards would go on the record as being referred to (in the case of Mike's call - I am careful with my choice of words here as I wouldn't want to fudge things by using a term like 'existing') and on the record (in the case of Maybrick's study floorboards being raised).

The list of the possible is May 12, 1889 (or May 11, 1889 or May 10, 1889 or May 3, 1889 if you want to make your odds even more unlikely by a day each time) through to when you know the two things finally happened (an eye-watering 37,618 days). Most statisticians would discount (without certainty, note) the role of randomness when the odds get to as little as 1/20 (or 5%, which gives you the familiar expression 'p <= 0.05' in statistical analysis). More rigorous studies will insist on 1/100 (or 'p <= 0.01'). Therefore, if you give a statistician a number like 1/37,618, he or she will immediately say to you that they are 'almost certain' that chance alone was not the reason for the 'coincidence' which was observed. Personally, I would be happy to bet my life if there was a 1-in-37,618 chance of my dying by taking whatever action I'm being asked to take. Hence, I was more than happy to take the two AZ jabs even though there was an infinitesimal possibility of blood clots possibly forming as a consequence.

You are formulating your “denominator” based largely on your belief that the scrapbook is likely to be the genuine diary of James Maybrick’s and that it was placed somewhere in Battlecrease as early as March 11th 1889.

Authentic or hoax, someone could have contacted a literary agent with a mooted diary of Jack the Ripper (aka James Maybrick) on May 12, 1889 and on every one of 37,618 days thereafter. Technically, it could have happened before May 12, 1889, but 37,618 is a Really Big Number so it will do for me on this occasion. Instead, what happened was that 37,618 days passed, World Wars, fall of communism, etc., and eventually the two events decided to happen on the same day giving a 1-in-37,618 possibility the status of staggering coincidence (unless the two events were linked).

But since you are forced to make your calculations from a position of ignorance (because no one knows or at least agrees to the diary’s origins), it is entirely possible that your ‘earliest’ date is entirely wrong and flawed and an impossibility---yet you insist on using it.

And that is because I understand statistics and you do not.

In a sense, your calculations aren't so much wrong, as they are invalid, and incorrectly formulated.

There is a danger that you are becoming a dog barking at the moon now.

In other words, by setting the first possible date as May 12, 1889, [strangely, Maybrick actually died on the 11th], you are, in effect, irrationally insinuating that it WAS placed under the floorboards on that date, and then you are simply counting forwards until you reach March 9, 1992 (the date of Barrett's call) which gives you the number 37,618, which you then convert to 37, 618 to 1.

Okay, maybe I was wrong - it sounds like you've finally got it?

No. No. No.

Ah ...

You are adding hundreds and thousands of what could be entirely imaginary ‘possible’ days to your denominator without having the faintest idea where the diary actually came from. And, because you are basing your calculations on an assumption, you are left with no way of knowing whether or not your resulting ‘denominator’ is accurate.

Your calculations are a group of goblins dancing on the head of an imaginary pin

If only any of this were relevant!

But I’ve tried many ways to express this, Ike, and, with a heavy heart, I must give up.

You put me in mind of the American summer camp owner I worked for in New York State many decades ago. He didn't like me. I didn't like him. He had guns. I kept quiet. One day, the smart arse challenged me to a Chess game where the prize was my entire summer's wages ($1,000, as I recall). If I lost, I had no money to travel after camp ended and would have to go home. If I won, I'd have $2,000. I was a member of my university's Chess club (I think I forgot to put that bit on my CV) so I immediately said "Yes" which instantly spooked him (it was a lovely moment). Anyway, after about five moves, I knew I was going to win so I started to toy with him. I would make my move and explain to him why I was making it. After about ten moves, I decided not only to tell him why I was making my move but also advise him which move he should make to stop me from winning (made the game far more interesting for me). After about fifteen moves, I politely offered him a draw (because I honestly didn't believe the slime ball would be honourable enough to pay me my winnings) and he accepted immediately and stormed out of his office with a huge pet lip on. So funny! Good thing was, he stopped dicking me around after that.

Someone else will have to carry the baton, and I pray for your sake that they do not whack you over the head with it.

I can't promise they won't because I will take no precautions to stop them (unlike my circumspection with the guy with the guns), but I am happy to take on the challenge from anyone. If anyone can get a credible statistician on here to contradict my calculations, I will happily oblige. No fears. Not one.

Good luck.

I think it would be safe to say that it isn't down to luck.

Cheers,

Ike

[Iconoclast]

Abandon all hope ye who enter here

Abby, it isn't often that you and I agree, but I'm right there with you, kidda.

[Iconoclast]

It is no shock that I understand and support Ike’s analysis. I also have a sneaking suspicion that he might be more qualified in the area of statistics than RJ or Lord O.

Perhaps RJ you should consider that.

I agree with your assessment, erobitha, although personally I might go a hint further than a 'sneaking suspicion' ...

[Iconoclast]

It couldn't be more unlikely if it tried ...

[rjpalmer]

It is no shock that I understand and support Ike’s analysis. I also have a sneaking suspicion that he might be more qualified in the area of statistics than RJ or Lord O.

Perhaps RJ you should consider that.

Alas, no.

If you think you "understand" Ike, then you're making the same fatal error of logic that he is. You've followed him Through The Looking Glass.

The simple mental error that Ike is making is that he isn't basing his calculations on what is actually known. He's throwing in irrelevant and unconfirmed and theoretical factors.

What is actually known?

It is known that Barrett called Crew on 9 March 1992.

It is also known that Dodd had work done on Maybrick's old house on 9 March 1992.

Now take a deep breath and clear your head.

It is NOT known that the Diary is the work of James Maybrick (and, in fact, it isn't, but that's another question) so throwing in the date 11 March 1889 as one of the factors immediately make's Ike's equation invalid.

To use Lord Byron's phrase, Ike's calculations are "mental masturbation" based on the Diary actually being the work of James Maybrick, thus shoving the 'possibilities' back to 11 March 1889.

Find a grade-school math teacher to explain it to you guys, I'm done!

The question that Lord O asks, as bizarre as it seems to you two, is the relevant one. How often does Dodd have work done on his house? And really, how often does anyone connected to James Maybrick have work done on their house, because you two would be grasping at those straws, too, had they happened to have coincided with Barrett calling Crew on the telephone, despite you not knowing whether the diary is old or not, or whether it actually came from any of those places or not.

Extend my warm wishes to Ike, I can waste no further time on him.

R P

[rjpalmer]

By the way, Ike is also wrong when he implies that 'Ripperologists' having no interest in statistics, and this is why his great discoveries have gone unnoticed.

If Ike ever strayed outside the padded community of Maybrick Hatch, he would realize this.

It wasn't long ago that Christer Holmgren gave the odds of the Ripper being anyone other than Charles Lechmere, based on Lechmere's supposed connections to the murder sites. I can't find the relevant post at the moment, but his colleague estimated it as something like 3 million to 1, or 5 million to 1.

And then we had Russell Edwards calculating how likely it would have been for any other London woman in 1888 to have had the same mDNA profile as Kate Eddowes (based on a sample extracted from the questionable shawl) and came up with a number of 290,000 to 1.

Many years ago we had Alex Chisholm discussing Bruce Paley's efforts to calculate how likely it would have been for 6 random murders to have occurred in the East End in 1888 without there having been a connection.

I don't think the lack of interest people have in Ike's theories has anything to do with them not being interested in such matters. He simply fails to convince them that his methods are legitimate.

[erobitha]

By the way, Ike is also wrong when he implies that 'Ripperologists' having no interest in statistics, and this is why his great discoveries have gone unnoticed.

If Ike ever strayed outside the padded community of Maybrick Hatch, he would realize this.

It wasn't long ago that Christer Holmgren gave the odds of the Ripper being anyone other than Charles Lechmere, based on Lechmere's supposed connections to the murder sites. I can't find the relevant post at the moment, but his colleague estimated it as something like 3 million to 1, or 5 million to 1.

And then we had Russell Edwards calculating how likely it would have been for any other London woman in 1888 to have had the same mDNA profile as Kate Eddowes (based on a sample extracted from the questionable shawl) and came up with a number of 290,000 to 1.

Many years ago we had Alex Chisholm discussing Bruce Paley's efforts to calculate how likely it would have been for 6 random murders to have occurred in the East End in 1888 without there having been a connection.

I don't think the lack of interest people have in Ike's theories has anything to do with them not being interested in such matters. He simply fails to convince them that his methods are legitimate.

Honestly RJ, I am saying this from a point of respect. Re-read exactly the exact premise he outlined. It has nothing to do with fake or hoax. That is as relevant as the colour of the book itself. That is not what is being discussed. I’m wondering if whether the attempt of muddying the waters is deliberate, because I think most people reading it will grasp what he is saying.

Perhaps seek the expertise of someone who has a strong academic background in mathematics, statistics or physics. I’m sure once they read the premise and the explanation they can perhaps persuade you that what he is saying is accurate.

I fear this has the tendency to descend into personal attacks and I for one would hate to see that happen. Although I fear we are already borderline.

[Iconoclast]

Honestly RJ, I am saying this from a point of respect. Re-read exactly the exact premise he outlined. It has nothing to do with fake or hoax. That is as relevant as the colour of the book itself. That is not what is being discussed. I’m wondering if whether the attempt of muddying the waters is deliberate, because I think most people reading it will grasp what he is saying.

Perhaps seek the expertise of someone who has a strong academic background in mathematics, statistics or physics. I’m sure once they read the premise and the explanation they can perhaps persuade you that what he is saying is accurate.

I fear this has the tendency to descend into personal attacks and I for one would hate to see that happen. Although I fear we are already borderline.

What he said ...

[caz]

There's nothing coincidental about a bandwagon hoax. One event inspires the other, so it doesn't qualify as a coincidence.

But I wasn't talking about a bandwagon hoax. That's your theory and of course it would be no coincidence if the emerging diary news in 1993 had indeed inspired someone to put the scratches in the watch. I was referring to documented events, which you have to believe were entirely coincidental: the watch itself being put on sale, with no known details of its previous owner [just a stranger off the street], at the same time Mike was trying to flog his diary [claiming its previous owner was someone who had conveniently died the year before]; and the double event of 9th March, which preceded the watch and diary sales, when floorboards were lifted in Maybrick's old bedroom, and the Maybrick diary got its first documented mention - as JtR's diary - courtesy of Mike's call to Doreen.

I'm going to leave Dr. Fuller alone. It's too tedious.

I don't blame you. I wouldn't want the uphill task either, of trying to explain how the Barretts managed to quote Fuller's words from 1889 directly and accurately, with only Bernard Ryan's narrative to guide them. Stick to the low hanging fruit. Try to forget Fuller, and the sensible brother Johnson, and keep looking towards the dodgy one.