Stuart Kind, Geographic Profiler.

[rjpalmer]

I still do not understand the logic behind the later TODs (in terms of time of day) being more likely to be closer to home.

I think all he is saying is that even a killer needs to go home at night.

If he's cruising Manchester to find a victim, he's not going to stay until 2 a.m., because he's got to allow time to get home. So, he's going to give up earlier.

On the other hand, he might commit a murder in Bradford or Leeds at 2 a.m., because the commute back to his digs is relatively short. So, if this logic is correct, the fact that the 'late fliers' are in Bradford and Leeds, suggests he is somewhere in those areas, closer to home.

I don't see how this could possibly be relevant to a set of murders happening in the same square mile in East London by someone on foot.

[Lombro2]

…the next biggest suburb or city, wouldn't work in geographical profiling for the simple reason there is no circle encompassing the base.

West Yorkshire is 780 square miles and 2.3 million people live in that area.

There were many places in West Yorkshire that were fortunate enough to not have an attack in the area. Kind could quite easily have come up with south of Leeds or somewhere like that.

As I said, that was impressive from Kind.

He gave extra weight to attacks earlier in the series and to those women attacked later in the night/early in the morning.

I find it very interesting that he believed the later TODs/attacks would be closer to home, and he was proven correct in that assessment.

I agree that Kind could have come up with south of Leeds. I would have gone with both North and South just to be in the safe side and go with the bigger urban centres as top possibilities such as Bingley and Wakefield to the north and south respectively. That’s just based on the theory of Critical Mass of Population.

North is still a good assessment. This is a commuter killer (I agree with Trevor that JtR is one too) so he’s usually outside the commuter kill zone and not in the middle but, with so many victims, it’s likely he’ll kill close to home at some point.

That however doesn’t turn the residence into a centred anchor point. Bingley isn’t really a central loci like most anchors. It’s a 5 miles or so distant town with thousands of people between Bradford and Keighley. All those victims and he never struck there once.

[JeffHamm]

Five to six murders aren't a sufficient number of 'data points' for any meaningful profiling analysis. Plus killer/victim behavior and external circumstances like the presence of witnesses in the street, lighting, etc. dictate when and where the killings occur.

Rossmo has said that in public at times, but I'm not sure it is based upon any empirical evidence. I've just had a look at the performance for routines based upon Rossmo's work, Canter's work, and my own, when I limit them to cases with only 5 or 6 offense locations. There's not that many, only 23 in total, but we can still have a look at the numbers.

The size of the search zone is determined based upon the offense locations, so the more spread out offenses are, the larger the potential search area. The first step is to define how large that search area is. That area applies to all of the routines, so they all search the same amount of "space" (though not necessarily the exact same locations, particularly around the edges of the search area, but there is a lot of overlap).

The search space is divided into 40 zones, with Zone 1 representing the "2.5% of the area that is most promising", Zone 2 is the next highest 2.5% of the search area, and so forth.

If you randomly searched the area, so just pick a spot at random (without replacement), you would, on average, have to search 1/2 of it before you uncovered the offender's location (sometimes more, sometimes less, but on average you have to search the equivalent of out to zone 20).

So, it depends upon what one calls a "meaningful" profile, but I will present a couple of possible definitions. First, what percent of the time does the offender's anchor point fall in "Zone 1". By random search, you would expect that to happen about 2.5% of the time (so with only 23 cases, that would be 0.575 cases (so, 0 or 1). A less stringent definition would be what percent of the time does the offender's anchor point fall inside of zones 1-20 (so better than a chance search). Another way to look at it might be to say "to what Zone do you have to search to have a 50% chance of locating the offender's anchor point"?, which is 20 for a random search, so "meaningful" would mean less than zone 20. (bracketed numbers are the number of cases out of 23)
.........................Rossmo.......Canter...... ........Hamm
Zone 1.........:.....30.4%(7)....52.2%(12)........52.2% (12)
<= Zone 20..:.....87.0%(20)..87.0%(20)........87.0%(20)
50% Zone #.:.........3...............1..................... .1

In other words, keeping in mind this is a very small sample, even with only 5 or 6 offense locations, the routines do better than chance. One thing to keep in mind, though, is that the estimated search area "expands" as the series length goes down because the fewer locations you have to work with the more those locations will underestimate the size of the actual criminal territory (the area in which the offender selects locations to offend). Also, given how few series we're dealing with here, I wouldn't make anything out of how Rossmo has fewer in Zone 1 and a larger 50% Zone number. Also, the very high "zone 1" rates look a bit non-representative to me, and I would expect if I found more series to add to the test set, the percentages would come down.

If instead, I only include a series with at least 7 and up to 25 offense locations (56 cases), we get this:
.........................Rossmo.......Canter...... ........Hamm
Zone 1.........:.....16.1%(9)....16.1%(9).........16.1% (9)
<= Zone 20..:.....85.7%(48)..85.7%(48)........91.1%(51)
50% Zone #.:.........4...............4..................... .4

and really, there's no difference between the routines there either.

Finally, if I increase the "short series" sample, to include all series with 3 to 6 offense locations, we end up with 54 series, so similar sample to the 7+ version, we get:
.........................Rossmo.......Canter...... ........Hamm
Zone 1.........:.....24.1%(13)....33.3%(18).........29. 6%(16)
<= Zone 20..:.....83.3%(45)....77.8%(42).........83.3%(45)
50% Zone #.:.........4...............5..................... .6.5

So on the whole, performance (in terms of search area reduction) for short series isn't really noticeably different from longer series.

However, one very big thing to keep in mind, is that the shorter the series, the more it has to be inflated (the "base size" is the area of the smallest circle that surrounds the offenses; then you do a profile and demark your zones until you're jeopardy map covers that amount of area - you don't limit the map to that circle, that's just used to get the search area size). Then you "expand" that area as a function of the number of offenses, with greater expansion the fewer the number of offenses.

Anyway, that all being said, given that I have to source many of the series data from published articles on spatial analysis of offense locations, there is a good chance I have a biased sample to work with. On the other hand, I do have some series data from a New Zealand study on serial arson, which I applied these routines to (and this data includes all cases of serial arson over a period of a couple years - not a huge number of cases, and a lot are short, but at least it can't be viewed as a biased set), and performance is similar (although the 50% zones for this set of 36 series shifts out to 7 type thing, but in the end it is the sort of variability one would expect for small samples like this).

I would like to get my hands on a large set of data, that includes all series of 3+ offenses (so no selection bias), that is large enough that I could really examine some of these sorts of questions. One can only dream.

- Jeff

[GBinOz]

I wasn't very clear, while there are serial killers who do find their victims in pubs, with JtR I was thinking more that there is a good chance he spent time in the pubs before going out and looking for victims, which he then found walking the street.

- Jeff

Hi Jeff,

"There are no less than 146 registered lodging-houses, with a number of beds exceeding 6,000. Of these 1,150 are in Flower and Dean-street alone, and nearly 700 in Dorset-street. Some of the houses contain as few as four beds, whilst others have as many as 350."

If Jack was randomly selecting women he found walking the street, as far apart as Mitre Square, Dutfields yard, Buck's Row and Hanbury St, what would be the odds that they would all be living in the Flower and Dean/Dorset street lodging houses, which comprised less than a third of the total. What additional effect would there be on those odds that the profiler programs show the same areas as hot-spots for Jack's anchor points.

Best regards, George

[JeffHamm]

Hi Jeff,

"There are no less than 146 registered lodging-houses, with a number of beds exceeding 6,000. Of these 1,150 are in Flower and Dean-street alone, and nearly 700 in Dorset-street. Some of the houses contain as few as four beds, whilst others have as many as 350."

If Jack was randomly selecting women he found walking the street, as far apart as Mitre Square, Dutfields yard, Buck's Row and Hanbury St, what would be the odds that they would all be living in the Flower and Dean/Dorset street lodging houses, which comprised less than a third of the total. What additional effect would there be on those odds that the profiler programs show the same areas as hot-spots for Jack's anchor points.

Best regards, George

Hi George,

My understanding is that the Flower & Dean/Dorset street area had a high concentration of "unfortunates" relative to other, even fairly nearby, areas. Not sure where I get that from though? If unfortunates were equally distributed over the whole area, then the odds of getting 5 from the same 1/3 area, would be 1/3^4 (4 not 5, as the first victim could come from anywhere, after that the other 4 have to come from the same "1/3" area as the first), which is about 1.2%. But if "unfortunates" clustered in that area (woman tended to group together in those doss houses), then the odds could be higher. Did Tabram, McKenzie, Coles, or some of the other victims, have connections to that area as well? I can't recall. If so, it may be that it isn't so unlikely after all.

Regardless, it's the fact that all lived in that area that makes me wonder if the hot spot from the profiling is picking up on that commonality of space between victims. Obviously, JtR has to be there too, but it does leave open the very real possibility that the hot spot may not be something like his residence per se, but rather if there is something there other than "where he goes to find victims, knowing there are many in that vicinity", that leads me to think that places like pubs would be the sort of thing to look at. We know it is the hot spot, but figuring out why it is the hot spot is where real police work comes into to. They are the ones who look there to see what sort of evidence can be found.

It is, of course, possible that he did reside in that area. I believe the police even did a concentrated house to house in and around there as well, but if anyone seemed of any interest to them at the time that information appears to have been lost.

I do think the spatial analyses outputs could have been put to use, though I would caution against viewing them as indicating his residence or work place for the above reason. Again, I wouldn't preclude those as possibilities but would caution against tunnel vision and thinking those must what one is looking for. Probably the best thing would be to conduct a concentrated surveillance of the nighttime activities, looking for someone who seems to fairly regularly be "out and about" on their own, and who walks around the area, my guess is probably along Commercial Street, along Whitechapel Road, and circles up and back along Hanbury street (perhaps including the other east/west roads as well - look for someone sort of "patrolling" the whole area on their own, who might explain themselves as out for air, etc). I would expect JtR spent a lot of time doing such things, and was out looking for victims much more often than on the nights he succeeded.

If he did live outside the area, which I do think is sufficiently likely that it needs considering, I don't think he lived very far outside. My first guess is that he entered the area in the vicinity of the Hanbury and Commercial hot spot area. So perhaps he lived North of there, and enters into the area via Brick Lane, or Wilkes Street, or from further North along Commercial. That, of course, is just a hypothesis. If the hot spot is primarily about the commonality of spatial information arising from the victims, then JtR could easily enter from any direction.

There are some interesting ideas that follow from some of the options to be considered. If one believes the apron piece was deposited as JtR fled the Eddowes scene (rather than the option where he "goes home comes back out"), then the implied direction of travel does have him heading back towards Commercial, which would be consistent with a desire to then head North on Commercial (if he was going to head south, then why turn "north" on Goulston to drop the apron?). Of course, if that assumption is wrong, and the apron was dropped there much later, then that would point to someone probably living fairly close to the GSG, and that works with the hotspot too, but probably places him internal to the region, rather than where I'm running to at the moment.

The other idea of interest follows from going with a later ToD for Chapman (I know you lean away from that, but bear with me), which would imply that JtR was willing to commit this murder as the sun was coming up. Obviously, that means more people around, increased risk of being spotted with blood on him, etc. If he was willing to take such a risk, that means something must be mitigating that risk assessment, and I would suggest that mitigating factor is that he lived close enough to #29 that he felt the short exposure time was acceptable to him. And his willingness to get quite bloody at the Kelly murder (he reaches up under her ribs to remove her heart after all), could also suggest he felt that he was close enough to home that the risk of extra "potentially visible evidence" was reduced by a short return travel time. And of course, for those who favour Kelly being murder the next morning, that gets magnified.

As an initial idea, a search roughly in the red rectangle below might have been worth conducting, perhaps checking out the Pub circled in orange (the lower right corner is Commerical and Hanbury by the way). And to be clear, this is nothing more than running with an idea, and I myself would not be at all surprised if it is completely wrong. Fortunately, as we don't know who JtR is, nobody can prove I am! ha ha Sadly, I can do no better. There are, of course, many assumptions being made to get to the point where this suggestion is even worthy of mentioning, which I've tried to make clear above is what I'm doing - I'm choosing a set of the possibilities we have to consider, and seeing where that particular combination might lead us. The more "wrong guesses" I've made above, the less this idea need be considered. Even if by some stroke of luck I've guessed them all correctly, this is still just a case of offering a "well, let's consider this idea then" as one has to start somewhere.

- Jeff

[GBinOz]

Hi George,

My understanding is that the Flower & Dean/Dorset street area had a high concentration of "unfortunates" relative to other, even fairly nearby, areas. Not sure where I get that from though? If unfortunates were equally distributed over the whole area, then the odds of getting 5 from the same 1/3 area, would be 1/3^4 (4 not 5, as the first victim could come from anywhere, after that the other 4 have to come from the same "1/3" area as the first), which is about 1.2%. But if "unfortunates" clustered in that area (woman tended to group together in those doss houses), then the odds could be higher. Did Tabram, McKenzie, Coles, or some of the other victims, have connections to that area as well? I can't recall. If so, it may be that it isn't so unlikely after all.
I believe Tabram lived further south in the St George Yard buildings, but McKenzie lived in Gun St, and Coles in Thrawl-street and Flower-and-Dean-Street.

Regardless, it's the fact that all lived in that area that makes me wonder if the hot spot from the profiling is picking up on that commonality of space between victims. Obviously, JtR has to be there too, but it does leave open the very real possibility that the hot spot may not be something like his residence per se, but rather if there is something there other than "where he goes to find victims, knowing there are many in that vicinity", that leads me to think that places like pubs would be the sort of thing to look at. We know it is the hot spot, but figuring out why it is the hot spot is where real police work comes into to. They are the ones who look there to see what sort of evidence can be found.

It is, of course, possible that he did reside in that area. I believe the police even did a concentrated house to house in and around there as well, but if anyone seemed of any interest to them at the time that information appears to have been lost.

I do think the spatial analyses outputs could have been put to use, though I would caution against viewing them as indicating his residence or work place for the above reason. Again, I wouldn't preclude those as possibilities but would caution against tunnel vision and thinking those must what one is looking for. Probably the best thing would be to conduct a concentrated surveillance of the nighttime activities, looking for someone who seems to fairly regularly be "out and about" on their own, and who walks around the area, my guess is probably along Commercial Street, along Whitechapel Road, and circles up and back along Hanbury street (perhaps including the other east/west roads as well - look for someone sort of "patrolling" the whole area on their own, who might explain themselves as out for air, etc). I would expect JtR spent a lot of time doing such things, and was out looking for victims much more often than on the nights he succeeded.

If he did live outside the area, which I do think is sufficiently likely that it needs considering, I don't think he lived very far outside. My first guess is that he entered the area in the vicinity of the Hanbury and Commercial hot spot area. So perhaps he lived North of there, and enters into the area via Brick Lane, or Wilkes Street, or from further North along Commercial. That, of course, is just a hypothesis. If the hot spot is primarily about the commonality of spatial information arising from the victims, then JtR could easily enter from any direction.
I don't see why Commuters could not be included in the hypothesis. Lechmere could have had a pub near to his work that he patronised after work which acted as his anchor. Druitt could have had a bolt hole where he underwent a transformation from the mild mannered Dr Jekyll to the working class Hr Hyde. Interesting that Richardson lived in what is currently called Wilkes St, and Sgt Badham lived in the barracks of Commercial St Police Station.

There are some interesting ideas that follow from some of the options to be considered. If one believes the apron piece was deposited as JtR fled the Eddowes scene (rather than the option where he "goes home comes back out"), then the implied direction of travel does have him heading back towards Commercial, which would be consistent with a desire to then head North on Commercial (if he was going to head south, then why turn "north" on Goulston to drop the apron?). Of course, if that assumption is wrong, and the apron was dropped there much later, then that would point to someone probably living fairly close to the GSG, and that works with the hotspot too, but probably places him internal to the region, rather than where I'm running to at the moment.
Such as Jacob Levy, whose home is also in a hot spot.

The other idea of interest follows from going with a later ToD for Chapman (I know you lean away from that, but bear with me), which would imply that JtR was willing to commit this murder as the sun was coming up. Obviously, that means more people around, increased risk of being spotted with blood on him, etc. If he was willing to take such a risk, that means something must be mitigating that risk assessment, and I would suggest that mitigating factor is that he lived close enough to #29 that he felt the short exposure time was acceptable to him. And his willingness to get quite bloody at the Kelly murder (he reaches up under her ribs to remove her heart after all), could also suggest he felt that he was close enough to home that the risk of extra "potentially visible evidence" was reduced by a short return travel time. And of course, for those who favour Kelly being murder the next morning, that gets magnified.
Keeping in mind that there was a horse slaughter yard just a few doors down the road.

As an initial idea, a search roughly in the red rectangle below might have been worth conducting, perhaps checking out the Pub circled in orange (the lower right corner is Commerical and Hanbury by the way). And to be clear, this is nothing more than running with an idea, and I myself would not be at all surprised if it is completely wrong. Fortunately, as we don't know who JtR is, nobody can prove I am! ha ha Sadly, I can do no better. There are, of course, many assumptions being made to get to the point where this suggestion is even worthy of mentioning, which I've tried to make clear above is what I'm doing - I'm choosing a set of the possibilities we have to consider, and seeing where that particular combination might lead us. The more "wrong guesses" I've made above, the less this idea need be considered. Even if by some stroke of luck I've guessed them all correctly, this is still just a case of offering a "well, let's consider this idea then" as one has to start somewhere.

- Jeff

O frabjous day! Callooh! Callay! We finally seem to be on the same side of the fence.

Running with a fresh hypothesis, even if it is brainstorming, is far more interesting than the constant re-examination, in frightening detail, of topics long since exhausted with their repetition.

Best regards, George

[GBinOz]

Hi Jeff,

RD just reminded my of his thread " Lodging Houses Link" where I posted this:

Here is an interesting interview from Lloyd's Weekly London, Feb 4, 1912. It is with retired Inspector Edmund Reid:

"I have been asked to tell the story of the "Ripper" series many times, but to do so would necessitate the devotion of weeks of labour to the matter.

But this I will say at once. I challenge anyone to produce a tittle of evidence of any kind against anyone.

The earth has been raked over, and the seas have been swept, to find this criminal "Jack the Ripper," always without success.

It still amuses me to read the writings of such men as Dr. Anderson, Dr. Forbes Winslow, Major Arthur Griffiths, and many others, all holding different theories, but all of them wrong.

I have answered many of them in print, and would only add here that I was on the scene and ought to know."

"...Here are the only known facts. The whole of the murders were done after the public-houses were closed; the victims were all of the same class, the lowest of the low, and living within a quarter of a mile of each other; all were murdered within half a mile area; all were killed in the same manner.

That is all we know for certain.

My opinion is that the perpetrator of the crimes was a man who was in the habit of using a certain public-house, and of remaining there until closing time.

Leaving with the rest of the customers, with what soldiers call "a touch of delirium triangle," he would leave with one of the women.

My belief is that he would in some dark corner attack her with the knife and cut her up. Having satisfied his maniacal blood-lust he would go away home, and the next day know nothing about it.

One thing is to my mind quite certain, and that is that he lived in the district."

Best regards, George

[A P Tomlinson]

Geographical Profiling is an interesting intellectual exercise, but is not reliable enough to help actually catch the killer.

People citing Sutcliffe as an example need to remember the Folly of focusing resources on a specific geographical area, when better evidence points elsewhere. (Castletown in Wearside, anyone?)

Unless, like in Operation Lynx in the 1990's, hard evidence points to the area it's not accurate enough to rely on as a tool. A stolen credit card, and receipts from specific shops focused the area of the profile, which came up with two towns. (One where the kidnapper lived, and one where his mum lived.)
And when a competent detective has clues like that, do they reallly need a criminal profiler to tell them that the perpetrator spends time in those areas?

The irony of Sutcliffe is that had he been successful, instead of getting arrested, in Sheffield the geographical profile would have been thrown more towards the Wakefield/Dewsbury neck of the woods as a base. Neither of which had a prolific "Red Light District" such as Bradford's Mannigham Lane at the time, though the Wakefield area did have one of the highest "pubs per capita" ratios in the 70s and early 80s and that was where most of the local sex workers plied their trade. So that neck of West Yorkshire would have been a dangerous hunting ground for Sutclife.

They certainly spent a lot of time looking for him round the Dewsbury/Wakefield area.
It's where I live.
My dad worked in Huddersfield and drove the A642 to get home. He was regularly stopped and his car searched at Police stops along that road because of the car he drove. (White MkII Escort, that obviously looked enough like a Corsair)
Even if he didn't get pulled out of the line, he would often take two hours to drive the 12 miles home if they were running stops there. And he was pulled over at least two or three times a month.
On the map Lombro posted, Huddersfield is the unlabelled grey blob at the bottom and the A642 runs pretty much along that line that says "19km" and they were actively searching cars in that area, regularly, up to the point of Sutcliffes arrest.

[JeffHamm]

O frabjous day! Callooh! Callay! We finally seem to be on the same side of the fence.

Running with a fresh hypothesis, even if it is brainstorming, is far more interesting than the constant re-examination, in frightening detail, of topics long since exhausted with their repetition.

Best regards, George

Indeed, and I raise a glass to you!

Thanks for the lodging info. I've been busy with end of semester wrap up/admit duties, and haven't had time to properly try and find even simple things. Sigh. Anyway, as for Martha Tabram though, I think George Yard Buildings to the south is where Martha was killed. Curiously, it appears her last known address was 19 George Street, Spitalfields (known as Satchell's Lodging House), and while I have located the street on the map yet, Spitalfields would be to the North of Flower and Dean. I haven't found Gun Street either, but it doesn't appear to be in the immediate area, so I've not looked too far from there.

And I agree, other commuter type suspects could also fit the bill. If one could show an individual did, as you suggest, frequent one of the pubs, that would provide a definite tie in to the locality, which would be of interest. Without actual evidence, though, it's just another "well he could have ...", which is true, but what we need to know is if he did! Nothing wrong with speculating, provided one doesn't confuse speculating something that would support a theory with actually having further evidence in support of the theory. It's just a hypothesis the theory has to demonstrate is true.

Oh, I didn't realise we knew where Richardson lived! That's a new one on me. While I have no suspicions on him, it is interesting that he lived in one of the potential entrances to the area (but then, he was going to Hanbury Street, so perhaps it's not surprising he lives in a location that gives access to Hanbury Street, which in turn happens to be in the hot spot for some of the routines, not the Rossmo one, but it's still close enough to be of interest).

And yes, Levy is in the area of interest. Hutchinson is too, or at least an address for him is; I seem to recall that he may have only moved into the residence on Wentworth after the Kelly murder? Again, that's a vague memory I have, so I'm not positive on that. Druitt had a cousin, I believe, who had an office of some sort south of Whitechapel Road, but I don't believe there is any evidence that Druitt himself ever visited there.

I had a map where I had tried plotting out all the various suspects' addresses, but I sort of botched a few of them, and I really should redo it when I have the time to do it properly (and with care). I can then run the spatial analysis, and plot all the suspects on it, so people can see where they resided.

Anyway, I agree, it is fun to run with some ideas. I don't actually "believe" that JtR lived to the North West, because there's no actual evidence of that. However, I think it is an interesting hypothesis that can be drawn, given the spatial analysis output, and what we know about the victims (and how that could be influencing what the spatial analysis is telling us), the nature of the crimes, and the fact that after the most bloody murders (Chapman, Eddowes, and Kelly), one would expect he would have to get "out of sight" pretty quickly to reduce the chance of being spotted. (Perhaps that is why he "balked" with Stride? He realised he was "too far from home"? Again, just tossing around possibilities, not suggesting these are factual statements).

Also, I'm not sure my "NW corner" idea depends upon Chapman being murdered late either. It would still be a crime where his hands would be quite bloody, and even if he wiped with a cloth, the desire to get out of sight would have been fairly high (unless he was psychotic and completely away with the fairies, but if so, it is amazing he wasn't spotted at some point over the entire series if after his crimes he just went for a merry stroll).

I don't think any of the named suspects, though, live up in that area, so if that is where JtR resided, well, he may remain a mystery for all time.

- Jeff

[A P Tomlinson]

Jeff,
this really seems to be your wheelhouse, so I wonder if you would have any idea how much the YR geo-profile would have changed had the Sheffield incident gone the other way?

[JeffHamm]

Geographical Profiling is an interesting intellectual exercise, but is not reliable enough to help actually catch the killer.

Spatial analysis no more catches a killer than does comprising a list of people close to a victim. Catching an offender requires going out and doing actual police work, to find evidence, follow leads, and so forth.

Spatial analysis is similar to making that list, you order people in terms of their personal closeness to the victim (spouse/ex-spouse, family members, friends, work mates, etc) and you start looking at them, eliminating those you can, and following up on those you can't. But rather than order people, spatial analysis orders locations, that's it. It tells you where to prioritize your limited resources, and work your way down. It's information that, over cases, does reduce the search space, so in that way can be helpful.

It's not helpful if it is viewed as anything more than an informed way of prioritizing locations (just like making a list of people isn't useful if the police decide from the start that the spouse did it!).

People citing Sutcliffe as an example need to remember the Folly of focusing resources on a specific geographical area, when better evidence points elsewhere. (Castletown in Wearside, anyone?)

Unless, like in Operation Lynx in the 1990's, hard evidence points to the area it's not accurate enough to rely on as a tool. A stolen credit card, and receipts from specific shops focused the area of the profile, which came up with two towns. (One where the kidnapper lived, and one where his mum lived.)
And when a competent detective has clues like that, do they reallly need a criminal profiler to tell them that the perpetrator spends time in those areas?

The irony of Sutcliffe is that had he been successful, instead of getting arrested, in Sheffield the geographical profile would have been thrown more towards the Wakefield/Dewsbury neck of the woods as a base. Neither of which had a prolific "Red Light District" such as Bradford's Mannigham Lane at the time, though the Wakefield area did have one of the highest "pubs per capita" ratios in the 70s and early 80s and that was where most of the local sex workers plied their trade. So that neck of West Yorkshire would have been a dangerous hunting ground for Sutclife.

They certainly spent a lot of time looking for him round the Dewsbury/Wakefield area.
It's where I live.
My dad worked in Huddersfield and drove the A642 to get home. He was regularly stopped and his car searched at Police stops along that road because of the car he drove. (White MkII Escort, that obviously looked enough like a Corsair)
Even if he didn't get pulled out of the line, he would often take two hours to drive the 12 miles home if they were running stops there. And he was pulled over at least two or three times a month.
On the map Lombro posted, Huddersfield is the unlabelled grey blob at the bottom and the A642 runs pretty much along that line that says "19km" and they were actively searching cars in that area, regularly, up to the point of Sutcliffes arrest.

This is the output for my routines using the Peter Sutcliff case, showing the 75% region (meaning, 75% of offenders are located inside this region). Sutcliff moved during the murders (the two "blue" boxes are his two residences). This map comes from Rossmo's publicly available thesis. Anyway, the yellow region is Zone 1, and both of Sutcliff's residences fall in Zone 1. (reducing the entire search space to 2.5%). The White spots inside Zone 1 just highlight areas that would be 1% (I call Zone 1 a bulls eye, and inside the 1% area a double bull). Now, because Sutcliff's murder range was very large, the potential search area is also huge. Even reducing it down to 2.5% still leaves a lot of leg work to do. But given they interviewed him multiple times, he looked like the sketches, he worked at a business that distributed the 5 pound note found at a crime scene, I'm pretty sure that if the senior police had seen that Sutcliff lived in the top priority zone, he would have ignored that too in favour of his Geordie Accent man.

Rossmo's routines puts the Dark Blue (southern) residence in Zone 2, and the Teal Blue (northern) remains in Zone 1. Canter's routines puts them both in Zone 2 (although the Teal one is pretty much on the border of zone 1 & 2).

[Fleetwood Mac]

Five to six murders aren't a sufficient number of 'data points' for any meaningful profiling analysis.

Rossmo was brought over to England to hunt down a serial rapist. He conceded that the number of attacks, which was five, was going to be a problem for geographic profiling. He did track down the rapist to a certain part of Leeds, which turned out to be correct, and Leeds is a much wider area than Whitechapel.

Having said that, Rossmo had other information to go on, other data points, such as credit card purchases.

Aye, Rossmo seems to agree with you.

On the other hand, not everyone believes we have knowledge of only five of six victims, and so for some of us there are more than five or six data points.

And, Kind used methods which may still be applicable to the WM, even when considering only five or six victims:

1) He focused on later TODs (in terms of time of day). I don't think you're going to get much later than half three in the morning, no matter how many data points you have.

2) He focused on the earlier attacks. In the event you believe there were five or six victims, then you still have the earlier attacks.

3) He focused on the centre of gravity and permeated all of the possible options in terms of which victims were at the hands of Sutcliffe and which weren't. Each time, the centre of gravity was the same place. In the event you have three more data points for the WM, will it change the centre of gravity? Quite possibly not. You could argue that's because Whitechapel was geographically a small area.

Either way, whether you believe the number of victims was five or six, or you believe all of the attacks in that period were attributable to the WM, or somewhere in between; Kind's methods would not lead to Flower and Dean Street and the surrounding streets.

You would end up somewhere just off the Whitechapel Road in an area encompassing the London Hospital and the eastern end of Hanbury Street.

[JeffHamm]

Jeff,
this really seems to be your wheelhouse, so I wonder if you would have any idea how much the YR geo-profile would have changed had the Sheffield incident gone the other way?

Sorry, I know a bit about the Sutcliff case but I'm not fully versed in all the details so I don't know what you mean when you say the "Sheffield incident"?

- Jeff

[Fleetwood Mac]

I agree that Kind could have come up with south of Leeds.

He could have in the event he was taking a wild stab in the dark.

But, he wasn't. His method told him that it wasn't South Leeds or most other parts of West Yorkshire, it was North Bradford.

The point was that he deduced broadly where the murderer lived, in an area of 780 square miles; by applying a method that he has left to us in a report.

[jason_c]

I don't think the perp was of the commuter variety. I would say a returning sailor is not a commuting serial killer. Such a sailor will undoubtedly have regular digs. Even if he's not living in the exact same London digs every time he will likely be drawn to live at a local base. However, I think the perp was likely to be local to Whitechapel(I include a sailor as local). At times the area was filled with police but he was always forced to kill in Whitechapel. I don't care what anyone says, street prostitutes were available outside Whitechapel, yet at no time does he seem willing to travel any great outside Whitechapel distance to murder.