One Incontrovertible, Unequivocal, Undeniable Fact Which Refutes the Diary

[caz]

Of course, I agree.
The odds of ten heads being spun in a row is just as likely or unlikely as any other combination of head and tails.

Not sure how this is relevant. Every coin spin is going to have a completely random outcome every time. We know there will be an outcome and it's always going to be one of two: heads or tails.

The odds of the winning lotto numbers spewing forth this Saturday in the following order: 1, 2, 3, 4, 5, 6 [bonus number 7] are identical to the odds of the winning numbers coming out as follows: 50, 8, 46, 12, 13, 4 [choose your own bonus number], or any other possible combination, emerging in any order. The order is irrelevant, or the jackpot would keep rolling over until we are all dead, but still, it would be very silly to pick the first combination of 1-6 - not because the chances of those numbers coming up are any different from any other selection, but because if they ever do come up, the jackpot will inevitably be shared among the thousands of equally silly people who pick their numbers according to a popular and predictable pattern. There's a reason for not choosing the PIN number 1234 or the password RJORSAM!

The trick is to be the only person to pick whichever six numbers come up on Saturday. But the chances of it actually happening to any one of us are very very very very very tiny indeed, and only very very very very very slightly more likely than if we don't buy a ticket, because the numbers always come up at random so there's no way of predicting the outcome.

What we were discussing here - or what I thought we were discussing - were the chances of anyone calling anyone about a diary signed Jack the Ripper, which turned out to identify James Maybrick as the supposed author, at any time from May 1889 to the 12th of Never, and this event actually happening on the same day - 9th March 1992 - that Maybrick's old floorboards came up, as part of a modernisation process which was likely to happen at some point while the old place was still standing. The diary, incidentally, covers just the period of Maybrick's brief occupancy of Battlecrease, from early 1888 to 8 days before he died there, in his own bed. It's very much the story of his life while in that house.

Nobody knew on 9th March whether such a diary would actually materialise, or even existed outside of Mike Barrett's fertile imagination. There were rumours in the wake of Florie Maybrick's conviction in 1889 that her diaries existed and someone had tried to publish them, but nothing ever turned up. Nobody as far as I know ever suggested that James had kept a diary too - until Mike produced the old book on 13th April 1992.

Starting with the presumption that the two events - the floorboards and the phone call - were NOT related [as I still think Ike was doing], and the bonus ball of the Saddle connection was therefore another coincidence, the trick would be to calculate the chances of those two events coinciding on the same day.

It's all very well for Lord Orsam to have worked out that in 1992 there was a 1 in 18 chance of Mike telling Doreen he had Jack the Ripper's diary, and this turning out to be supposedly by Maybrick, on one of the weekdays that year when 'work' [not confined to floorboard lifting in Maybrick's old bedroom] was being done in Battlecrease House. But either by accident or design he was missing the point that neither was an everyday event [one or both were unique], and neither need have happened at all, either in 1992, or before or since. It would have been a completely different matter if Orsam was merely calculating the chances of Mike talking to a man about a dog, on one of the days when a man living in Mike's neck of the woods was seen in the evening walking his dog through Anfield Cemetery.

Dog poo in, dog poo out.

Love,

Caz
X

[Yabs]

Not sure how this is relevant. Every coin spin is going to have a completely random outcome every time. We know there will be an outcome and it's always going to be one of two: heads or tails.

The odds of the winning lotto numbers spewing forth this Saturday in the following order: 1, 2, 3, 4, 5, 6 [bonus number 7] are identical to the odds of the winning numbers coming out as follows: 50, 8, 46, 12, 13, 4 [choose your own bonus number], or any other possible combination, emerging in any order. The order is irrelevant, or the jackpot would keep rolling over until we are all dead, but still, it would be very silly to pick the first combination of 1-6 - not because the chances of those numbers coming up are any different from any other selection, but because if they ever do come up, the jackpot will inevitably be shared among the thousands of equally silly people who pick their numbers according to a popular and predictable pattern. There's a reason for not choosing the PIN number 1234 or the password RJORSAM!

The trick is to be the only person to pick whichever six numbers come up on Saturday. But the chances of it actually happening to any one of us are very very very very very tiny indeed, and only very very very very very slightly more likely than if we don't buy a ticket, because the numbers always come up at random so there's no way of predicting the outcome.

What we were discussing here - or what I thought we were discussing - were the chances of anyone calling anyone about a diary signed Jack the Ripper, which turned out to identify James Maybrick as the supposed author, at any time from May 1889 to the 12th of Never, and this event actually happening on the same day - 9th March 1992 - that Maybrick's old floorboards came up, as part of a modernisation process which was likely to happen at some point while the old place was still standing. The diary, incidentally, covers just the period of Maybrick's brief occupancy of Battlecrease, from early 1888 to 8 days before he died there, in his own bed. It's very much the story of his life while in that house.

Nobody knew on 9th March whether such a diary would actually materialise, or even existed outside of Mike Barrett's fertile imagination. There were rumours in the wake of Florie Maybrick's conviction in 1889 that her diaries existed and someone had tried to publish them, but nothing ever turned up. Nobody as far as I know ever suggested that James had kept a diary too - until Mike produced the old book on 13th April 1992.

Starting with the presumption that the two events - the floorboards and the phone call - were NOT related [as I still think Ike was doing], and the bonus ball of the Saddle connection was therefore another coincidence, the trick would be to calculate the chances of those two events coinciding on the same day.

It's all very well for Lord Orsam to have worked out that in 1992 there was a 1 in 18 chance of Mike telling Doreen he had Jack the Ripper's diary, and this turning out to be supposedly by Maybrick, on one of the weekdays that year when 'work' [not confined to floorboard lifting in Maybrick's old bedroom] was being done in Battlecrease House. But either by accident or design he was missing the point that neither was an everyday event [one or both were unique], and neither need have happened at all, either in 1992, or before or since. It would have been a completely different matter if Orsam was merely calculating the chances of Mike talking to a man about a dog, on one of the days when a man living in Mike's neck of the woods was seen in the evening walking his dog through Anfield Cemetery.

Dog poo in, dog poo out.

Love,

Caz
X

Hi Caz.

It’s not relevant at all. It was a post as part of a brief tangent after Erobitha posted that only the first spin of a 10 spin coin toss has 50/50 odds.
Then we moved on, so it’s only relevance at the moment is that you’ve quoted me and brought it up again.

Cheers

[Iconoclast]

What we were discussing here - or what I thought we were discussing - were the chances of anyone calling anyone about a diary signed Jack the Ripper, which turned out to identify James Maybrick as the supposed author, at any time from May 1889 to the 12th of Never, and this event actually happening on the same day - 9th March 1992 - that Maybrick's old floorboards came up, as part of a modernisation process which was likely to happen at some point while the old place was still standing.

Correct, Caz. Go to the top of the class. Pass Go. Collect £200.

Starting with the presumption that the two events - the floorboards and the phone call - were NOT related [as I still think Ike was doing], and the bonus ball of the Saddle connection was therefore another coincidence, the trick would be to calculate the chances of those two events coinciding on the same day.

£400.

It's all very well for Lord Orsam to have worked out that in 1992 there was a 1 in 18 chance of Mike telling Doreen he had Jack the Ripper's diary, and this turning out to be supposedly by Maybrick, on one of the weekdays that year when 'work' [not confined to floorboard lifting in Maybrick's old bedroom] was being done in Battlecrease House. But either by accident or design he was missing the point that neither was an everyday event [one or both were unique], and neither need have happened at all, either in 1992, or before or since.

As I keep saying, you can pursue other probability analyses, but you absolutely can’t do so in order to contradict some other probability analysis. So, what is the actual question being posed by Lord Orsam’s extraordinary 1-in-18 stat? It’s this:

What was the likelihood that someone would contact a literary agent with James Maybrick’s confession on a day in 1992 when any work whatsoever was being done in Maybrick’s old house?

Well, how many days was any sort of work at all done in Battlecrease House in 1992? I understand that the answer to that is 14.

How many days were there in 1992 minus weekends (or just Sundays?) and Bank Holidays? The answer to that (just based on Orsam’s 18-to-1 odds and without working it out) must be 14 x 18 so 252.

Therefore:

What was the likelihood that someone would contact a literary agent with James Maybrick’s confession on a day in 1992 when any work whatsoever was being done in Maybrick’s old house?

We’ll (drum roll!), that would be 14/252 or 1/18 or a p value of 0.05555 - which (cunningly) would ensure that it wasn’t statistically significant! A truly atrocious attempt to deflect away from the staggeringly unlikely events which coincided that fateful day in March 1992.

Loved your ‘Fido’ pun earlier, by the way, Cazzykins.

PS I think I deleted the bit about ‘chance’ from your original post, but - yes - you are absolutely right, the 1/37,557 odds are only relevant if this all happened purely by mind-blowing chance. If not, they were linked and therefore you don’t need to invoke the role of chance. That’s statistics for you!

Ikeypoos

[caz]

Hi Caz.

It’s not relevant at all. It was a post as part of a brief tangent after Erobitha posted that only the first spin of a 10 spin coin toss has 50/50 odds.
Then we moved on, so it’s only relevance at the moment is that you’ve quoted me and brought it up again.

Cheers

I thought erobitha merely said that a single one-off toss would be 50/50, not that a theoretical second or subsequent toss would not have the same odds each time. I took the subsequent 50/50 comments to be the product of reading between erobitha's lines and making 1 + 0 = 2 or more.

But maybe I got that all wrong.

Love,

Caz
X

[erobitha]

I thought erobitha merely said that a single one-off toss would be 50/50, not that a theoretical second or subsequent toss would not have the same odds each time. I took the subsequent 50/50 comments to be the product of reading between erobitha's lines and making 1 + 0 = 2 or more.

But maybe I got that all wrong.

Love,

Caz
X

You didn’t t. That’s exactly what I said.

[JeffHamm]

Hi all,

Ok, first, just to be clear, I have no dog in this race with regards to Diary debates. Moreover, even if we were to throw out the entire line of argument based upon probabilities, that's not going to change anything Diary wise, other than to illustrate why this approach doesn't work the way it may feel like it does.

The idea of the chance combination of two events is just the product of their individual probabilities. Let's keep the "calling the publisher" at whatever low rare probability one feels appropriate. The other event is the lifting of the floorboards at Battlecrease. If that event is truly rare, the chance probability for the combination of two extremely rare events would be very low. That gives the impression that the two events must be somehow related (this is why it's an inferential test; we're trying to infer a causal link between the two events).

Ike has suggested the chance probability for of the floor boards being uplifted has a 1 in 3000+ chance (forgive me if I don't go back and get the exact number, it doesn't matter for our purposes here) based upon the number of days between two events. That, however, is not correct for the following reason. That presumes the probability of repairs to Battlecrease are equally spread over those days. But, as a building gets older the probability of it requiring repairs increases, so the probability of requiring a repair increases with time, making the days at the end of the series more probable to require floorlifting than the days early in the timeline.

Second, repairs are something pre-planned, you don't just call a tradie and they show up that day. So the probability that the floorboards were going to be raised that day become very high, approaching 100% because no doubt that day was arranged in advance.

If the floorboards had a near 100% chance of being raised on that day, though, which they did because the repairs had to be pre-arranged, and the probability of repairs required is much larger after such a long time, then the combination of the two pretty much leaves the combined probability at the same level as the probability of just calling the publisher by itself.

We cannot distribute the probability equally over time for repairs to Battlecrease, because repairs are not a random event. They are events that increase in probability over time, and given the floors hadn't been worked on in such a long time, the liklihood they needed repairs would be high, which, in fact, is why they were worked on. And if the probability of the repairs is high, then the combination of the two events is not the combination of two improbable events, they are the combination of an improbable event with one of high probability. Moreover, the fact that both Mike and the repair fellow were regulars at the same pub means there would be a high probability on any random day they would both be there, making that another high probability chance event that requires no causal explanation.

And again, to infer a causal link between those events, we have to consider all possible causal links of stories that would create a link between Battlecrease and those involved with the Diary, and that creates a near infinite number of possible stories, each of which might appear extremely improbable, but if we get to pick and choose what we use to build those stories, we can inevitably come up with something, making it unsurprising that a story of connection can be produced.

We could view this as a hypothesis though. In which case we could now make predictions, such as it should be possible to show a known social link between Mike and the repair man; or evidence of money changing hands between them; or looking for touch DNA on the Diary that shows the repairman has handled and gone through the Diary while it was in his possession (though by now I suspect that's not possible); or the Diary would have sawdust/dust etc in creases etc, from Battlecrease that matches substances found under the floors there, and so forth. Those then become low probability bits of information that we now predict, and we then test them because their chance combination is apriori low. Once we know an event has happened, like the floorboards, we have to look at the events leading up to that event to determine if it was a low probability event for it to occur at that time, and given the age of the property, the fact the floors had not been worked on before, and the fact those repairs had been pre-arranged, means there was a high probability of the floorboards being raised that day, leaving our only low probability event being the calling of the publisher, and the combined probabily not changing as repairs on that day were, in fact, highly probable.

Had we not known the floorboards were raised, but predicted that Mike must have gotten the Diary from someone who did repairs on the house that day, and then went and found that was the case, then that would have been different (the probability of repairs around that time would still be fairly high, but because I'm talking about a situation where we didn't know they actually had happened, then our estimates would be different, and we would have found information not previously known but that we predicted, and our prediction was supported. That's not the case here, we have information we know, and we find a way to create a story, or a hypothesis if you will, about how those known events are causally connected. That's post-hoc reasoning, and it's fine for creating testable avenues of investigation, but it is not proof of itself, that's circular reasoning - I have A and B, I hypothesis A and B are related, my proof that A and B are related is that I have A and B).

Anyway, I see what people are trying to do, and it's a good idea in principle to try and apply some objective approaches. However, this one does not work the way it feels like it should. Statistics are shockingly counter-intuitive sometimes.

That's really all I feel I can contribute, though.

- Jeff

[erobitha]

Had we not known the floorboards were raised, but predicted that Mike must have gotten the Diary from someone who did repairs on the house that day, and then went and found that was the case, then that would have been different (the probability of repairs around that time would still be fairly high, but because I'm talking about a situation where we didn't know they actually had happened, then our estimates would be different, and we would have found information not previously known but that we predicted, and our prediction was supported. That's not the case here, we have information we know, and we find a way to create a story, or a hypothesis if you will, about how those known events are causally connected. That's post-hoc reasoning, and it's fine for creating testable avenues of investigation, but it is not proof of itself, that's circular reasoning - I have A and B, I hypothesis A and B are related, my proof that A and B are related is that I have A and B).

Hi Jeff,

Just for clarity, knowledge of Mike calling the publisher came before anyone knew the floorboards were lifted the same day. That is important to consider. Also, I don't recall anyone claiming the hypothesis was a smoking gun calculation that was to be used as absolute proof.

The scenario posed in the original premise (from my understanding) was a very simple calculation based on no other influencing data. If you were to take those events on their face value, the probability of the occurrence being the same day remains statistically hugely unlikely. Even if the repairs of the house probability increases due to age, a suggestion of 18-1 is not correct. The odds surely still remain very high? What would you consider as the odds at this point?

Obviously, other data must be considered for an absolute scientific result, but that was not the intention. There are many factors as you outlined if we were to stress-test the basic premise into a workable hypothesis that could be used for the purposes of prediction. This is simply to amplify "isn't that interesting" and no more than that from what I can tell.

Regards,

Ero

[Iconoclast]

Hi Jeff,

Just for clarity, knowledge of Mike calling the publisher came before anyone knew the floorboards were lifted the same day. That is important to consider. Also, I don't recall anyone claiming the hypothesis was a smoking gun calculation that was to be used as absolute proof.

The scenario posed in the original premise (from my understanding) was a very simple calculation based on no other influencing data. If you were to take those events on their face value, the probability of the occurrence being the same day remains statistically hugely unlikely. Even if the repairs of the house probability increases due to age, a suggestion of 18-1 is not correct. The odds surely still remain very high? What would you consider as the odds at this point?

Obviously, other data must be considered for an absolute scientific result, but that was not the intention. There are many factors as you outlined if we were to stress-test the basic premise into a workable hypothesis that could be used for the purposes of prediction. This is simply to amplify "isn't that interesting" and no more than that from what I can tell.

Regards,

Ero

Bloody Hell, ero b, I think I’m just going to leave this debate to you and Caz as your recent posts are right on the statistical money, and this one saved me the time and effort of replying to Jeff. Nice one.

What I was planning to reply to Jeff was - as you say - an acknowledgement that the likelihood that the floorboards would come up probably did increase with age (he makes a good point there that I hadn’t thought about) but I think you are right, the increase in probability would be right there at the very margins and any increase we might factor in would be nominal in relative terms.

I was also going to say - as you said - that this should not be viewed ‘experimentally’. Jeff - I suspect - is a researcher first and a statistician second. That’s not to say he doesn’t know both his onions, but that his eye is immediately on the ‘experiment’ and whether it is sufficiently powered to support any given hypothesis. But - as you again point out - this was simply a basic probability analysis with only the core elements assessed and no hypothesis formed. Crucially - and again you state it - none of us knew of any statistical elephant in the room until Keith Skinner (waiting patiently for Bruce Robinson to publish We All Love Jack the Spratt) was free to reveal what he had first uncovered in something like 2004 - namely, that the smoking gun of Maybrick’s study floorboards being raised had gone off on the very day that Bongus Barrett had made his fateful ‘phone call.

Thus, we looked back some two decades and asked ourselves how likely it was that those two events would happen on the same day, and that was ‘number of times either happened’ over ‘how many times either could have happened’. By the way, the odds were not - as Jeff states - the product of two probability events (that would make the odds 1/37,557 (or 2/37,557 which is 1/18,779 if the demolishing of Knowsley Buildings is accepted) multiplied by 1/37,500 (or 1/18,779) which is a frighteningly small probability but actually is not required: the odds are 1/37,557 (or 1/18,779) because we do not know that either event would ever have happened and could only address the probability of them both happening when at least one of them had happened so we can’t treat both events as individually very unlikely so we shouldn’t sum the product of the two. Lifting the floorboards was no ‘biggee’ - it just had never happened on the record until 37,557 days after Maybrick’s death. Bongo’s call to Rupert Crew was arguably very unlikely (Maybrick simply isn’t an obvious candidate for Jackus Sprattus) but that’s for someone else to calculate because doing so serves effectively no purpose for us (and would only strengthen our argument if we attempted to do so) - what we have been debating is simply the likelihood that two events would coincide on the same day after 37,557 days where they could have done, and we could only do so in the retrospect of some two decades because we didn’t know that Maybrick’s floorboards had come up on March 9,1992 until the 2010s.

So this statistical elephant in the room was not to be analysed as an a posteriori event - we looked back with a priori only in our eyes and said “Let’s just do the maths here, pure and simple - no need to invoke ‘inference’ based upon experience or even explicit experimentation with its eager demand for hypotheses, both null and predicted”. And when we did that - twenty years post-hoc - we fell back on simply probability theory and skipped the subtleties of aging houses and which day the workmen were booked to turn up. For us it was all down to heads and tails without needing to factor in which way the grass was lying or whether one side has a greater likelihood of coming up because it is weighted differently by the impression on it.

And we got what we got. And no amount of post-hoc rationalisation of our simple methodology can alter that. Nor should it. Others are welcome to check the elephant out from different perspectives, but we’ve just walked in the room and said “elephant - we wouldn’t have expected that”. Nor would we have expected the Spanish Inquisition, but - Lord - we have certainly suffered it.

In fairness, no-one expects the Spanish Inquisition, but maybe next time we will …

Ike

[Iconoclast]

Hi Ike.

[Ike: Edited from post #8260]

I don't have the time, nor the inclination, to get into this at the moment, but as a demonstration of how little we have progressed over the past 15 years, let me reprint a post of mine on this topic dating to 18 June 2005. Please study it closely and see what points are raised.

* * * *

As far as I can fathom ---and the truth is far from clear--something rather strange is going on. If it was known that typescript was created "by agreement" with Crew, why is Keith asking Anne Graham about it? One can only conclude that there was still some mystery as to when and why the typescript was created. Hence the questions put to Anne in 1995.

Yet, it certainly seems that Graham's version of events directly contradicts Barrett's earlier version as reported by Montgomery. Montgomery clearly believed Barrett bought the word processor specifically to transcribe the diary and study it; Graham was now stating that the typescript was not made until the Diary was ready "to go" to a literary agency.

Is there any reason why we should now accept Anne's revision?

* * * * *

The above is slightly edited, but one can find the entire conversation here:

Casebook: Jack the Ripper - Message Boards: Where do we go from here?

Hi RJ,

Apologies for the delay in replying to your post #8260, but I wanted to get an answer from Keith himself. He has updated me as follows:

“Nothing at all strange is going on. On May 31st 1995, I took Anne to dinner as she was in London. Martin Howells was with me having flown over from New Zealand at Paul Feldman’s behest. Feldy was tremendously excited by Anne’s story of the diary allegedly having come down through Anne’s family and wanted to make a follow up documentary. Hence the reason he had flown in Martin to write and direct it. Martin was not at all convinced by any of Feldy’s ideas and did not believe Anne’s story. It was an opportunity therefore for Anne to tell Martin the story herself rather than Martin being dependent on what he had been told by Feldy. I invited another person along who had fixed ideas about the diary being a modern hoax so that he could judge for himself Anne’s story and veracity. What I wanted Anne to explain were the reasons she allegedly gave the diary to Devereux to give to Mike. And what had happened in the Barrett family home from the moment Mike allegedly had taken possession of the diary circa Spring 1991 up to the point he took the diary to London on April 13th 1992. At that point I had no idea about the potential significance of the March 9th 1992 date. That came 12 years later. Also, I had no knowledge about Mike’s affidavit of January 5th 1995. Had I been aware of its existence I would have openly asked about it in front of Martin and the other person. I did not sit there during dinner taking notes and neither did I have a tape recorder running. I don’t recall there being huge discussions going on in the background about the transcript and agreements with the Rupert Crew Agency. What I do remember is the huge divide between Robert Smith, Shirley Harrison, Doreen Montgomery and Paul Feldman. At this distance in time I cannot even recall if the transcript was uppermost in my mind or whether I asked Anne about it. I think the reason I wrote that one note later on was because of the little details Anne provided. I’m not even sure now what I meant by “go situation” and as to precisely when that was? I have the feeling I equated it with Mike taking the diary to London for the first time but, in hindsight, it could have referred to Doreen having arranged for publishers to see the diary and bid for publication rights. If on May 31st 1995 it had been crucial for me to know about the history of the transcript, I would have asked Shirley or Doreen as to when they first had sight of it. What impressed me most when Anne was telling the story was her feeling of resignation at having to accept that events had overtaken her – and if the diary was being taken seriously then the transcript should be presented professionally. So Anne started reading it whilst Mike typed but he was so hopeless with the word processor that the roles were reversed – Mike orally transcribing, Anne typing. It was just that little detail Anne added of constantly checking back with the original to see how a word was spelt that impressed me – hence the reason for my note.”

Hope this helps,

Ike

[Iconoclast]

As I keep saying, you can pursue other probability analyses, but you absolutely can’t do so in order to contradict some other probability analysis. So, what is the actual question being posed by Lord Orsam’s extraordinary 1-in-18 stat? It’s this:

What was the likelihood that someone would contact a literary agent with James Maybrick’s confession on a day in 1992 when any work whatsoever was being done in Maybrick’s old house?

Well, how many days was any sort of work at all done in Battlecrease House in 1992? I understand that the answer to that is 14.

How many days were there in 1992 minus weekends (or just Sundays?) and Bank Holidays? The answer to that (just based on Orsam’s 18-to-1 odds and without working it out) must be 14 x 18 so 252.

Therefore:

What was the likelihood that someone would contact a literary agent with James Maybrick’s confession on a day in 1992 when any work whatsoever was being done in Maybrick’s old house?

We’ll (drum roll!), that would be 14/252 or 1/18 or a p value of 0.05555 - which (cunningly) would ensure that it wasn’t statistically significant! A truly atrocious attempt to deflect away from the staggeringly unlikely events which coincided that fateful day in March 1992.

Hi Dear Readers,

I have tried valiantly to locate Lord Orsam's original work of statistical genius on his website for those of you who wanted to check what he actually argued, but after a lot of clicking on various articles, I had to give up, so here's a link to his website itself and - if you're interested - the fun can start from there ...

Home - Orsam Books

I assume it comes under the 'Articles' section, but - Lord - there's too many articles and too many links for me on a scorching Saturday evening.

Cheers,

Ike

[Scott Nelson]

So, what if Mike had the diary for some time prior to 9 March 1992? Maybe it was given to him by Devereux? So during a chance encounter between Eddie and Mike in the Saddle pub on that March day, Eddie recounts that he had heard of a book found somewhere in the house in the past. Mike assumes it's the diary in his possession and since he then had a usable provenance, he goes home and contacts a publisher.

Now, what are the odds of that happening?

[Iconoclast]

So, what if Mike had the diary for some time prior to 9 March 1992? Maybe it was given to him by Devereux? So during a chance encounter between Eddie and Mike in the Saddle pub on that March day, Eddie recounts that he had heard of a book found somewhere in the house in the past. Mike assumes it's the diary in his possession and since he then had a usable provenance, he goes home and contacts a publisher.

Now, what are the odds of that happening?

Hi Scott,

You’ve come up with a scenario (based on RJ’s recent scenario perhaps where Mike is in leagues with Eddie to pretend a diary was found?) which may or may not have happened. This one is for you to calculate because it doesn’t lend itself to simple probability theory.

Good luck, though. Do let us know what you come up with.

Cheers,

Ike

[Scott Nelson]

Thank you, Ike. The odds part was tongue-in-cheek, as it's impossible to calculate odds on anything like this, as I tried to exemplify earlier.

[Iconoclast]

Thank you, Ike. The odds part was tongue-in-cheek, as it's impossible to calculate odds on anything like this, as I tried to exemplify earlier.

Strange.

I thought your example earlier was something about some guy throwing nine heads and then a tail (or something like that)? If that’s what it was, I could work out the odds in seconds.

Or was it that the 10th throw was 50-50, as were the other nine throws, as were every throw of a coin there’s ever been? That bit is too obvious to imagine you meant, but it’s possible you did I suppose.

And yet you say your example was impossible to calculate the odds for? You’re going to have to remind us all what your example was, Scott, because the two above were about as easy as burping after a glass of Coke.

Ike

[Iconoclast]

Probability example: A guy tosses a coin in the air 9 times. 8 times it comes up heads, 1 time tails. What are the chances the 10th toss will come up heads?

Exactly 50-50.

Is this the one? It’s the only one I could find.

Obviously the odds of the final throw are the same as any throw there’s ever been in the history of coin throwing. It’s 50-50 for that single incarnation of a throw.

As I think I said in reply originally, what’s your point, caller?

Edit: Hold on, surely you’re not saying that every probability is 50-50 for any set of possible events, ever?

Ike