John Richardson

[GBinOz]

There's no way to know, alas. But the relevant info - including descriptions of the victims, plus the name and location of the mortuary each had been taken to - was certainly published in the press before Packer was contacted by the private detectives.

As Jon says, the coroner uses the witness statements to question them at the inquest.

The police statements (taken 9 Nov) of witnesses who later appeared at the Kelly inquest are available if you have access to a copy of The Ultimate JTR Companion / Sourcebook (or the London Metropolitan Archives), along with the inquest statements themselves. Chapter 21 The Kelly Inquest, Maxwell's statement is on page 365-6 of my old copy.

Thanks Jon and Joshua, for that information.

Cheers, George

[Fleetwood Mac]

For 4 out of 4 to be wrong

I don't think the reasoning follows, i.e. simply applying %s without context.

For example:

We know either Cadosch or Long was 'wrong', because they both claimed to know the time.

We know Richardson misled the coroner with his knife tale, and therefore his entire statement is compromised.

We know Cadosch didn't claim to hear the murder, he claimed to hear a voice and a noise against a fence. What conclusion you draw from that is up to you, but Cadosch doesn't need to be 'wrong' about anything given he didn't claim to hear a murder.

We know Long, in her own words, didn't take much notice of the couple and didn't see where they went.

We know that from 5am to 5.30am the place is coming alive with workers and the pubs opening, and so for a 5.30am TOD the murderer and Annie have to get to the back unseen, murder, mutilation, the murderer leave the scene unseen by 6am.

In the context of what happened that morning, it's pretty easy to conclude that on balance they probably were all 'wrong'.

[Trevor Marriott]

Thanks Herlock.

We're dealing with information that's over 130 years old, long past the time we can follow up or seek clarifications. Moreover, the information that survives to this day was gathered by techniques far less honed than those used today. Even today there are many complications with dealing with eye-witness testimonies. How a witness is interviewed, even when done with the best of intentions (so setting aside cases where a police officer had developed tunnel vision and railroads a witness/suspect, which we also knows happens), small changes in word choice in how a question gets asked can influence the response such that the witness statement becomes less accurate than it would have been otherwise.

We're also dealing with far less developed forensic techniques, and so medical experts of the day simply were not equipped with the knowledge we now take for granted. It's not disparaging of them or their abilities, it is a simple matter of fact that we know more about the reliability of medical information now than was known then.

My own personal view, which I've said on a number of threads, is that there simply is not sufficient information to prove anything apart from there was a series of murders. We can't even be sure which victims are part of the series and which are not; and this is something that even in a series committed today is a very difficult question to answer. DNA is starting to make it possible to find definitive connections, but it is not always the case that DNA is found at a scene, and not all DNA found necessarily means that was the offender (particularly in cases where a suspect is known to a victim; is that suspect's DNA at the victim's house because they knew each other or because they are actually the offender? Pending upon how it was located, and where, can greatly influence it's evidential value).

So with the above opinion in mind, I think it is critically important that we recognize the insufficiency of the information we have to conclusively prove anything, and to simultaneously recognize that all statements of "fact" are associated with "measurement error" (for lack of a better term). Long's identification of Annie might appear to be a 50/50 thing; either she's right or wrong in her identification. But that isn't a proper evaluation of "measurement error". Rather, how probable she is to be right or wrong depends upon the specifics of how that identification took place. If she was shown a group of bodies (like a simultaneous line up), there is a very good argument to be made that she made a false positive identification, but there is also a good argument she may have made a correct one. And it's about 50/50 (even if there was more than 2 choices for her to select from). If, however, she was taken to multiple different locations, and had to make a decision each time, then the "measurement error" narrows and we become more inclined to think she may have correctly identified Annie, but we shouldn't ignore the fact that about 1:3 times that would also be a false positive. That's still a high failure rate, and one that cannot be ignored.

But just because each individual bit of evidence is "noisy", and has some sort of error range associated with it, the more noisy bits of information one has to line up, the greater the chance a common solution emerges. In the case of Annie's murder, that is what I've been focusing on, the fact that all of the individually largish error ranges all tend to include a common time point, around 5:25. And when that happens, that is an indication that underneath all the noise there appears to be a common possible "truth" emerging. It is not, and never can be, proof that truth is True with a capitol T, but it does indicate that is the most supported interpretation (which is all I'm saying we have).

I have no problem with people who want to explore other possibilities, sometimes the rare event happens. But those other possibilities are, given what we have, less supported options. Nothing wrong with considering them all the same, but they are less consistent with the information we have.

As such, I tend to think our best option is to try and sort the order of the alternatives, which one does by looking at things under a systematic organization of the information.

Something like this:

Evidence:
Dr. Phillips, Richardson, Long, Cadoche, Davies (open door), leggings spring: -> ToD 5:25ish
Drop the leggings spring and we still get 5:25ish
Drop the leggings spring and the open door (Davies), we still get 5:25ish
Also drop Long, we still get 5:25ish
Drop Cadoche, we get sometime after 5:00 or so, which includes but does not specify 5:25ish
Drop Richardson (leaving only Dr. Phillips), we get a wide range spanning hours, centred around 4:30ish, which also still includes 5:25ish.

We could drop things in different orders than as I've listed above as an example, and basically, at no point do we end up being able to preclude 5:25, we only end up widening the window of time we have to consider as possible (so it may lower the probability of 5:25 being the "right time", but it never results in 5:25 being out of the running).

From the above, at least as I've presented it, Cadoche is the witness who has to be discarded to really remove 5:25ish; we could drop Richardson for example, and if we include Cadoche we still get around 5:25. Richardson alone simply means at some point after he left (around 5:00ish I believe), so if we drop Cadoche we can no longer point to 5:25 and maybe she was there since 5:05 type thing, bringing the ToD closer to Dr. Phillips' 4:30 estimate, which we know has such a wide range of error is unlikely to be bang on anyway.

While 5:25 appears to leave very little time for the murder and mutilations, ignoring the debate over organ harvesting, we know that very little time was available to perform even more extensive mutilations to Eddowes, which were performed in worse lighting conditions. So even that doesn't really present a problem (and it points to the over estimations made at the time as to how long it would have taken - because this was such a unique murder that the doctors at the time had no experience with, so their estimations for how long will be error prone, which we can see was the case after Eddowes' murder where there simply wasn't much time available).

Basically, in order to get to a ToD at, or earlier, than 4:30, we have to conclude that all the evidence should be discarded other than Dr. Phillips, and even then we cannot even exclude 5:25 entirely.

Therefore, I see it as being entirely irrational to conclude anything other than 5:25ish is the most supported time, that after 5:00 is the next most reasonable, and that prior to 4:30 is the least supported hypothesis.

That doesn't mean it's wrong to say "I'm betting on the least supported because I am comfortable setting aside all other information as being wrong". And yes, onehas to say it's wrong, not just possibly wrong because we know all the evidence is possibly wrong as stated - that's what the margins of error represent after all. I just don't see any compelling reason to conclude any of the evidence "must be wrong" to justify it being discarded. I can see the justification of considering the possibility some bit is wrong though.

We have 4 eyewitnesses (Richardson, Long, Cadoche, and Davies), all of whom have to be wrong before we open up the possibility of before 4:30 being the ToD. If any of them are right, though, 4:30 disappears. Let's say there is a 50% chance for each of them to be wrong. For all 4 out of 4 to be wrong that is 0.5 to the power of 4, or 6.25% chance of all 4 to be incorrect. And all it takes is for 1 of them to be correct to rule out that earlier time (which has a 93.75% chance of being the case).

I could be accused of being overly generous, though, and some may feel the odds of them being wrong is greater than 50%, say 80% likely to be wrong.

So let's say there is an 80% chance for each of them to be wrong (so they only have a 20% chance of being correct). For 4 out of 4 to be wrong, that is 0.8 to the power of 4, which is 40.96%, meaning, the probability of one, or more, of them, being correct is still greater than the probability of them all being wrong. And it only takes one of them to rule out 4:30 and earlier.

This is what I mean when I say that 5:25ish is the most supported hypothesis, given that we have to deal with probabilities.

And to be clear, it is still possible that all 4 were wrong. Even if they have a 50/50 chance of being correct, sometimes all 4 will be wrong (it just happens to be about 6.25% of the time, so it's far less likely). And if we go with 80% likely to be wrong, then we're still betting on the less probable horse, but the odds are not so long (sorry for the pun, not entirely unintentional).

- Jeff

Jeff
You have again made a sterling effort, but may I just say that murder investigations do not revolve around the use of percentages and your post clearly shows why I keep using the word "unsafe" when discussing the witness testimony

https://www.trevormarriott.co.uk

[Herlock Sholmes]

I don't think the reasoning follows, i.e. simply applying %s without context.

For example:

We know either Cadosch or Long was 'wrong', because they both claimed to know the time.

But they could have been wrong on their time and not on what they saw or heard.

We know Richardson misled the coroner with his knife tale, and therefore his entire statement is compromised.

Simply untrue.

We know Cadosch didn't claim to hear the murder, he claimed to hear a voice and a noise against a fence. What conclusion you draw from that is up to you, but Cadosch doesn't need to be 'wrong' about anything given he didn't claim to hear a murder.

But if he hadn’t heard noises connected to the murder and we assume that Phillips was correct then someone was moving around in that yard seemingly oblivious to the fact that there was a mutilated corpse there.

We know Long, in her own words, didn't take much notice of the couple and didn't see where they went.

Which still doesn’t mean that she couldn’t have been correct.

We know that from 5am to 5.30am the place is coming alive with workers and the pubs opening, and so for a 5.30am TOD the murderer and Annie have to get to the back unseen, murder, mutilation, the murderer leave the scene unseen by 6am.

They might have been seen by Long. When the killer left all that we have to accept is that for a period of around 5 seconds no one was looking at the door to number 29. After 5 seconds or so he’s 20 yards away and walking along Hanbury Street with absolutely no reason to have given him a second glance.

In the context of what happened that morning, it's pretty easy to conclude that on balance they probably were all 'wrong'.

Balance doesn’t come into it where you’re concerned. Its only easy to conclude that they were all wrong if you are, for some inexplicable reason, determined to make them wrong. Jeff has provided a very clear, balanced and thought out assessment there and, although I don’t know what Jeff’s job is, he certainly has a deep knowledge of statistics, but as you’ve previously claimed greater knowledge than all of the modern day forensic experts it would be par for the course for you to disagree with Jeff on statistics. All that you need to do next is to disagree with Trevor on modern day police procedures.

[Fleetwood Mac]

Balance doesn’t come into it where you’re concerned. Its only easy to conclude that they were all wrong if you are, for some inexplicable reason, determined to make them wrong. Jeff has provided a very clear, balanced and thought out assessment there and, although I don’t know what Jeff’s job is, he certainly has a deep knowledge of statistics, but as you’ve previously claimed greater knowledge than all of the modern day forensic experts it would be par for the course for you to disagree with Jeff on statistics. All that you need to do next is to disagree with Trevor on modern day police procedures.

Sherlock, I no longer have the will to go 'round the houses of invention, "we just don't know" and statement manipulation with you.

We'll simply have to agree to disagree.

As for statistics, see my previous post as to why it is a flawed approach in this situation.

[harry]

Would you say Jeff,the examples you use,prove beyond a reasonable doubt?,Before the usual shout of I am being too precise is raised,it is a real series of murders,and real witnesses,that are being discused.Models may be effective up to a point,but investigations are not work study exercises.and as Trevor points out,witness testimony is a weakness that cannot be brushed aside.

[Herlock Sholmes]

Sherlock, I no longer have the will to go 'round the houses of invention, "we just don't know" and statement manipulation with you.

We'll simply have to agree to disagree.

As for statistics, see my previous post as to why it is a flawed approach in this situation.

I read your previous post. The usual biased stuff.

You are the one making things up. You are the one who, for some strange reason, is conveniently averse to admitting that there are things that we just have no way of knowing. This is what you call logic.

Everyone of your imaginings have been unravelled. If you want to claim the impossible for a Doctor then deliberately try to create mysteries where none exist then it’s up to you. I’m tired of your ducking and diving. The evidence points to Chapman being killed at around 5.20-5.25. If you put that in front of any jury it would get a unanimous

[JeffHamm]

Jeff
You have again made a sterling effort, but may I just say that murder investigations do not revolve around the use of percentages and your post clearly shows why I keep using the word "unsafe" when discussing the witness testimony

www.trevormarriott.co.uk

Hi Trevor,

Oh, absolutely agree. Murder investigations must get things to as close to zero probabilities as possible, that's the whole basis of "beyond reasonable doubt", while that criterion allows for some small percentage of possibility "the other way", it means must be very very small. Viewed in that context, even 6.25% is too large to be called "beyond reasonable doubt"; there are 1000s of cases in the courts each year; if we allowed 6.25% of them to be wrong convictions, that's 62.5 miscarriages of justice per year per 1000!.

I'm only talking about how we, this distant from the events, can sort the hypotheses available. I don't think we could ever reach the criterion necessary to bring charges, let alone a conviction, if we view this case as a modern police investigation. Seriously, you of all people know that the investigative procedures (setting aside expert testimonies/opinions) were so primitive they would be laughed out of court before opening statements. If we look at the evidence as if it has to meet modern standards, we might as well just pack up and go home now. There's no point.

I know you view things from the perspective of a modern investigation, that's your expertise. While it may not often be apparent, I do respect your background and knowledge. I do get what you mean by "unsafe", and I agree with you when you say we can't base things on X, Y, or Z, entirely. That's what I'm saying when I talk of probabilities, I just use different language to say the same thing based upon my expertise and background. You and I tend to lock horns when it appears to me that you're then going off on an idea that may be an alternative but it appears you are presenting it as if it is a more probable alternative; as if being able to say "The reason idea A is unsafe is because B is possible" somehow leads to it being acceptable to say "therefore B must be true!" I agree that B is possible, but I tend to push back if it appears someone is arguing the "therefore B is true" conclusion. The former may be correct, but the latter does not follow, particularly if option B appears less probable than the original option A. I'm interested in ordering the probabilities, accepting that we can probably never know if the big T Truth is, indeed, the most probable option or one of the less common, but sometimes occur, options.

Anyway, it's a different way of pushing the same cart sometimes. I prefer our current interactions, though, and I'm glad to see we're finding common ground. I hope my above comments aren't taken as sneaking in a poke at you, rather they are just intended as an explanation of where I'm coming from. Like you, I won't back down just because someone pushes back, but I'm glad to seeyou also will agree with someone even when that has at times been less frequent.

- Jeff

[JeffHamm]

Would you say Jeff,the examples you use,prove beyond a reasonable doubt?,Before the usual shout of I am being too precise is raised,it is a real series of murders,and real witnesses,that are being discused.Models may be effective up to a point,but investigations are not work study exercises.and as Trevor points out,witness testimony is a weakness that cannot be brushed aside.

Good God no. We do not have the information available to reach "beyond reasonable doubt". I've always argued that point, and have always suggested the best we can do is try and order the probabilities of the various explanations. We won't, given how probabilities work, always order them with "Truth" at the top, but we are more likely to put Truth near the top, if we do this. And I think that's the best we can do with the JtR murders. (Note, by the time we get to specific suspects, I think we're so far removed from any evidence that the ordering is random - meaning, I have no suspect in mind, I am more interested in just trying to work out what happened, rather than who was responsible. That's a question, while interesting, I think we have insufficient information to even sensibly order. At best, we can order who different people thought was a good suspect, but that doesn't mean squat as to who really committed the crimes unless we have the evidence they used to get there so we ourselves can examine it with a modern perspective based upon any new things we've learned).

- Jeff

[FISHY1118]

I don't think the reasoning follows, i.e. simply applying %s without context.

For example:

We know either Cadosch or Long was 'wrong', because they both claimed to know the time.

We know Richardson misled the coroner with his knife tale, and therefore his entire statement is compromised.

We know Cadosch didn't claim to hear the murder, he claimed to hear a voice and a noise against a fence. What conclusion you draw from that is up to you, but Cadosch doesn't need to be 'wrong' about anything given he didn't claim to hear a murder.

We know Long, in her own words, didn't take much notice of the couple and didn't see where they went.

We know that from 5am to 5.30am the place is coming alive with workers and the pubs opening, and so for a 5.30am TOD the murderer and Annie have to get to the back unseen, murder, mutilation, the murderer leave the scene unseen by 6am.

In the context of what happened that morning, it's pretty easy to conclude that on balance they probably were all 'wrong'.

Indeed Mac , Its really very simple , with the amount of uncertainty unprovable beyond a doubt , and ambiguious nature of all the witness and medical testimony its a miracle some would have others belive the time of Annie Chapmans murder was a dead certainty between 5.20 and 5.30am !

Im sure ill get the usual response like,'' but hey fishy but all the evidence points to it being more likely tho'' . Well more likely doesnt cut it im afaide, as some people think the evidence points to a more likely ''earlier t.o.d'' . [refer first paragraph as to why]

Just for once id like to see some one from the other camp just come out and say , ''yes based on all the evidence and confusion with witness testimony there ''Could'' have been an eariler t.od ''

Good on you, have your own theory im happy for you either way , but please dont try claim that 5.30am is the ''only'' possible outcome in regards to Annie Chapmans t.o.d.

1747 debated post says its not as clear cut as it seems .

[Fleetwood Mac]

Indeed Mac , Its really very simple , with the amount of uncertainty unprovable beyond a doubt , and ambiguious nature of all the witness and medical testimony its a miracle some would have others belive the time of Annie Chapmans murder was a dead certainty between 5.20 and 5.30am !

Im sure ill get the usual response like,'' but hey fishy but all the evidence points to it being more likely tho'' . Well more likely doesnt cut it im afaide, as some people think the evidence points to a more likely ''earlier t.o.d'' . [refer first paragraph as to why]

Just for once id like to see some one from the other camp just come out and say , ''yes based on all the evidence and confusion with witness testimony there ''Could'' have been an eariler t.od ''

Good on you, have your own theory im happy for you either way , but please dont try claim that 5.30am is the ''only'' possible outcome in regards to Annie Chapmans t.o.d.

1747 debated post says its not as clear cut as it seems .

Aye, you simply cannot apply mathematical statistics when there are so many variables involved, e.g. the intricacies of witness statements.

As said, we know there are contradictions in their statements, and so something is clearly wrong there; and Cadosch does not need to be wrong about anything as he never claimed to hear a murder. These events, which on their own are fairly insignificant, have been lumped together with Richardson underpinning the whole thing.

So, we have Richardson, whose entire statement is compromised due to misleading the coroner, Cadosch who heard some noises at a time when people are coming alive again; and Long who in her own words didn't take much notice of the couple, didn't see where the couple went and sees lots of couples at that time of the morning.

Considering all of this in isolation, i.e. leaving Dr Phillips' examination aside, it's not particularly convincing.

Then we have Dr Phillips, who leaves us actual evidence, e.g. 'rigor commencing of the limbs'; upon which we can draw a conclusion underpinned by some actual evidence given by someone who did not compromise his entire statement by misleading the coroner.

I would say the strong case rests with an earlier TOD, and I reckon it's nowhere near a close run thing.

My bet would be 3am to 3.30am.

[Herlock Sholmes]

Indeed Mac , Its really very simple , with the amount of uncertainty unprovable beyond a doubt , and ambiguious nature of all the witness and medical testimony its a miracle some would have others belive the time of Annie Chapmans murder was a dead certainty between 5.20 and 5.30am !

Im sure ill get the usual response like,'' but hey fishy but all the evidence points to it being more likely tho'' . Well more likely doesnt cut it im afaide, as some people think the evidence points to a more likely ''earlier t.o.d'' . [refer first paragraph as to why]

Just for once id like to see some one from the other camp just come out and say , ''yes based on all the evidence and confusion with witness testimony there ''Could'' have been an eariler t.od ''

Good on you, have your own theory im happy for you either way , but please dont try claim that 5.30am is the ''only'' possible outcome in regards to Annie Chapmans t.o.d.

1747 debated post says its not as clear cut as it seems .

Fishy, I’m not going to go back and count how many times, but I must have said that the possibility exists that Phillips got it right and that there might have been an earlier TOD at least 20 times on this very thread. How many times does it need saying?

Its also a fact that for around 100 pages we’ve had posters repeatedly and emphatically stating that because Phillips was a competent doctor then that would have made it likely that he was correct. An ongoing refusal to accept the facts which only now appears to be finally accepted by all but 2 people. This is what we have been up against.

Then we’ve been up against people disputing the English language by trying to deny what Doctor Phillips actually said.

Then we’ve had people claiming a non-existent discrepancy between what Richardson and Chandler said at the inquest. A figment of the imagination.

Then Fishy, we’ve had you claiming that 3 woefully and provable inaccurate rough sketches are valuable evidence when the facts tell us otherwise.

All the sense, reason, fair-mindedness, common sense, logic and respect for evidence has come almost solely from the one side.

As Jeff has shown, without going anywhere near to certainty, the evidence clearly favours a TOD of around 5.20/5.20. Personally, my opinion is a minimum of 90%. A doctor guessing and 3 witnesses pointing to a later TOD? Not a difficult decision.

[Herlock Sholmes]

Aye, you simply cannot apply mathematical statistics when there are so many variables involved, e.g. the intricacies of witness statements.

Unless the statistics had favoured your side of course then you’d have been all over them.

As said, we know there are contradictions in their statements, and so something is clearly wrong there;

Apart from the time issue in regard to strong all other ‘contradictions’ are inventions.

and Cadosch does not need to be wrong about anything as he never claimed to hear a murder.

He heard a noise coming from a yard where you believe that a mutilated corpse lay.

These events, which on their own are fairly insignificant, have been lumped together with Richardson underpinning the whole thing.

A strange and meaningless sentence.

So, we have Richardson, whose entire statement is compromised due to misleading the coroner

This is an invention by someone who is biased.

, Cadosch who heard some noises at a time when people are coming alive again;

That’s a lie where you are assuming an earlier TOD.

and Long who in her own words didn't take much notice of the couple, didn't see where the couple went and sees lots of couples at that time of the morning.

Which doesn’t mean that she was wrong in the real world.

Considering all of this in isolation, i.e. leaving Dr Phillips' examination aside, it's not particularly convincing.

Leaving aside your lies and manipulations….yes it is.

Then we have Dr Phillips, who leaves us actual evidence, e.g. 'rigor commencing of the limbs'; upon which we can draw a conclusion underpinned by some actual evidence given by someone who did not compromise his entire statement by misleading the coroner.

Continue defending Gandalf Phillips. You are pretty much the last man standing on this point. What a joke.

I would say the strong case rests with an earlier TOD, and I reckon it's nowhere near a close run thing.

Miles ahead for the other side. Your biased, illogical and a stranger to honesty.

My bet would be 3am to 3.30am.

And considering your thinking on this issue I think that we can safely dismiss that.

[Herlock Sholmes]

Anyone that still tries to dismiss the witnesses based on Phillips is not worth listening to.

[Aethelwulf]

And considering your thinking on this issue I think that we can safely dismiss that.

here here. this FM joker is clearly one of the conspiracy theory 'light' brigade.