Pawn tickets in Mitre Square

[David Orsam]

David

if we are not using all the letters i can find 2 senior police officials based outside london saying they are the killer and a senior police doctor as well.


And that is with out really trying, just showing if you have a name in mind you can probably fit it.
I asked Pierre if we need to use up all the letters, however he has not replied yet

Yes, indeed, if one doesn't use all the letters of the 'anagram' it's a pretty much ridiculous and worthless exercise, especially if one keeps expanding the number of letters involved. A suspect's name out of just "Emily Birrell" and "Jane Kelly", using all 21 letters but only those 21 letters, would have been impressive. But once you include the addresses as well, and don't have to use all the letters, it becomes silly.

Did the addresses on the pawn tickets also say something like "Dorset Street, Whitechapel"? If so, what would be the reason for excluding the word "Whitechapel"? And what about the item descriptions: "Man's flannel shirt" and "pair of man's boots" or whatever was actually on the pawn tickets? And is there any meaning in the name of the pawnbroker: Joseph Jones of Church Street, Spitalfields? Should we be looking for serial killer names, or even a message from the serial killer, in that set of letters too?

[Elamarna]

David

if we are not using all the letters i can find 2 senior police officials based outside london saying they are the killer and a senior police doctor as well.


And that is with out really trying, just showing if you have a name in mind you can probably fit it.
I asked Pierre if we need to use up all the letters, however he has not replied yet

s

[David Orsam]

What is the statistical probability that you will find a serial killerīs name in a mustard tin on a murder site?

Just as an example of how poorly phrased Pierre's question was.

If you take the names and words: "Emily Birrell", "Jane Kelly", "Dorset Street" and "White's Row" we can extract the name: Myra Hindley.

Separately, we can also get Rose West.

Separately, if we are looking for male serial killers, we can also extract the name: Robert Lee Yates who killed 16 prostitutes in Washington between 1975 to 1988.

So what is the statistical probability of finding a serial killer's name in a mustard tin on a murder site? From a single random sample: Very, very high. Indeed, one could almost say: inevitable.

[David Orsam]

Elmarna: If all the letters does that include white's row and dorset street ?

Pierre: Of course.

Ah, here we go. And the rules have suddenly changed.

Pierre, you expressly confirmed to me in #112 that the full and unabridged name could be found in the exact 21 letters comprising "Emily Birrell" and "Jane Kelly". Now you are adding additional words and letters.

And you do know that an anagram means that you must use ALL the letters, with none spare, don't you?

Here's the wikipedia definition:

"An anagram is direct word switch or word play, the result of rearranging the letters of a word or phrase to produce a new word or phrase, using all the original letters exactly once"

No, I bet you don't. All you are doing is adding as many letters into the equation until you have enough letters to make up the result you are looking for (i.e. this person's name) and you will ignore the letters that are left over.

So no wonder you have been getting so aggressively defensive in this thread as you must realise that you are deciding the outcome by the words you are choosing.

[David Orsam]

No, David. I am just getting tired from your constant expressing of total ignorance, even when it comes to scientific problems where you have no knowledge at all.

You have no statistical education and yet you think you know how statistics work and yet you accuse me in advance of doing things you have no understanding about.

And yes, David. One CAN SIMPLY IGNORE THE POSSIBILITY OF HUMAN BIAS AFFECTING THE RESULTS. Ever heard of a simple random sample? No, of couse not.

Pierre, please don't make factual statements about me about which you have no knowledge such as "You have no statistical education".

A random sample is not going to eliminate the human bias potentially involved here. Just to give you one example. If the name of your suspect was an anagram of "Emily Birrell" you would naturally ignore the "Jane Kelly" pawn ticket. And if the name of your suspect was an anagram of "Jane Kelly" you would ignore the "Emily Birrell" one. So if you know the name you are looking for you are setting your own parameters. A random sample is, therefore, not going to achieve anything in respect of eliminating the bias.

[Elamarna]

Of course.

Best wishes, Pierre

pierre

does the anagram use all the letters or are there spare letters.


Steve

[Pierre]

Wow another degree, in statistics this time??

No, GUT. It is integrated in one of my sociology degrees.

[Pierre]

Pierre

if all the letters does that include white's row and dorset street ?

s

Of course.

Best wishes, Pierre

[Elamarna]

Steve,

I meant all of the given names and the surname. And the use of all the letters in the mustard tin.

Best wishes, Pierre

Pierre

if all the letters does that include white's row and dorset street ?

s

[GUT]

No, David. I am just getting tired from your constant expressing of total ignorance, even when it comes to scientific problems where you have no knowledge at all.

You have no statistical education and yet you think you know how statistics work and yet you accuse me in advance of doing things you have no understanding about.

And yes, David. One CAN SIMPLY IGNORE THE POSSIBILITY OF HUMAN BIAS AFFECTING THE RESULTS. Ever heard of a simple random sample? No, of couse not.

Wow another degree, in statistics this time??

[Pierre]

The hypothesis is basically disproved by the evidence in the case which explains how Eddowes came to be in possession of the pawn tickets.

But anyway, right, so "Emily Birrell Jane Kelly" will, when rearranged, form this person's name, right?

Is it the name of your suspect, or someone else entirely?

If you read my discussion with Steve on the subject, you get the answers you need.

[Pierre]

I was attempting to answer your question about statistical probability, Pierre, and, in this case, one cannot simply ignore the possibility of human bias affecting the results for the reasons I explained. The fact you are getting so defensive about this is rather worrying and suggests that I might have touched a nerve here.

No, David. I am just getting tired from your constant expressing of total ignorance, even when it comes to scientific problems where you have no knowledge at all.

You have no statistical education and yet you think you know how statistics work and yet you accuse me in advance of doing things you have no understanding about.

And yes, David. One CAN SIMPLY IGNORE THE POSSIBILITY OF HUMAN BIAS AFFECTING THE RESULTS. Ever heard of a simple random sample? No, of couse not.

[Joshua Rogan]

"I'll be real Mr E in KY jelly"

[Joshua Rogan]

But anyway, right, so "Emily Birrell Jane Kelly" will, when rearranged, form this person's name, right?

This sounds fun.... I've got "I am J Belly Killer, neerly"

[Errata]

No, Pierre, I have not the foggiest idea what you are talking about and I wasn't accusing you of doing anything, I was pointing out that if one is trying to calculate the statistical probability of something one has to remove the possibility of human intervention affecting the results. Do you understand that?

Haha I don't know. Aren't you just straight calculating the popularity of the name? I mean, James Black. Super common first name, super common last name, ergo the probability of that name being on the ticket, and a name must be on the ticket, is much higher than the probability of say, Oleg Krzyzewski.

I mean, we aren't looking for the killer to be on the ticket. Just his name.