Geographic Profiling

[clark2710]

Has this been tried to see where the killer likely lived versus where the killings were? If so, has anyone cross referenced where some of the suspects lived and/or worked? Who lived in that area?

[DJA]

geoprofiling jack the ripper - Google Search

[JeffHamm]

Has this been tried to see where the killer likely lived versus where the killings were? If so, has anyone cross referenced where some of the suspects lived and/or worked? Who lived in that area?

Hi clark2710,

The short version is yes, Rossmo's "Rigel" routines have been used (they tend to focus on Flower & Dean), and Canter's "Dragnet" has also been used. I've done some research in this area and have been developing my own routines as well. In testing, though, all 3 do pretty much equally well in general, though there can be some differences on individual cases.

There's a lot of misinformation about what this sort of analysis does. It does not "solve" a case, rather, it's a way of prioritizing physical search space so that over a number of cases one should do better. Think of it this way, our family, friends, co-workers, people at the pub, etc, are all part of our "inter-personal space". Most murders are between people that know each other, with partners being very high on that list, and other family members, and so on down the line. With no other information to go on, one can organize that inter-personal space into high priority people to interview down to low priority people. If you did that, you would locate the offender far sooner than if you simply searched all the people in the city at random. Over the long run, that's an efficient use of resources, and helps reduce the time to a result. But, and this is the important part, not every murder is committed by a person's partner, or a family member, or a friend, or co-worker, etc. And some are stranger murders. So even that prioritized list isn't "the solution", it's just an order to the search that will reach the solution faster.

You can think of geographical profiling as doing the same thing. The higher priority zones are just the spatial locations that, in general, are more likely to contain one of the offender's anchor points. It's often thought this is the offender's residence, and it often is, but it may also be their place of work, or maybe it's the pub they go to, or their church, or even their partner's place of work. Basically, the anchor points are locations that the offender has some connection to in their every day life.

And also, the analysis routines are based upon offenders who are "marauders", which just means they are committing crimes in the areas they spend their normal daily routines (live, work, entertainment, etc). If they specifically travel to a location out side their normal daily routine, and then search for victims, they are a commuter. In this case, the profile may simply locate an area where the offender tends to enter the region, perhaps an off ramp from the major highway, etc. It might still indicate a good spot to look for them, but if they have no other connection to that location, how would you find them since off ramps (in this example) are used by hundreds, or thousands, of people?

Here's the Rigel (Rossmo's) routines profile for the C5. The pink area is the highest priority zone, followed by yellow, then orange, then light red, dark red, green, cyan, and blue. 50% of offenders get located by the time you search the orange area, and 75% by the time you reach the end of the light red area (sorry, hard to see the difference here I'm afraid). After that, things tend to trail off quite a bit, so they're just for completeness. Oh, those numbers are based upon a set of serial offenses where the number of commuters is grossly under-estimated as well, but it's hard to get those because they're trying to optimize the search when the offender is a marauder (it's a problem in this area that really needs to be addressed).




And for comparison, here's the output from my own routines (I've left out the light and dark blue here, as they tend to contain almost no cases). Interestingly, JtR is one of the situations where the different set of algorithms produce fairly different looking solutions, but it's a matter of degree (Flower and Dean is still in the orange in mine, and Dragnets, and the area between Kelly and Chapman fall in the orange and light red regions for Rigel). Mine and Dragnet focuses more around the locations of Kelly's and Chapman's murders (with Kelly's right in the highest probability region), while Rigel is shifted to the east around Flower and Dean, with interest around Middlesex as well (which is interesting given it's between Mitre Square and Goulston Street; if JtR were there, and went home for a bit before heading back out to get rid of the apron piece and write the graffitti, it makes some sense - he's not travelled far before dropping it, and he headed away from Mitre Square, which could be seen as an attempt to mislead the police by suggesting he had gone further along that line, rather than extending the line further, if that makes sense). Rigel also points a bit to the North West, which is curious. Anyway, for the most part, all three tend to result in similar areas ranking high priority, though there can be some cases, like this one, where they can look more variable.

Again, all 3 do equally well in the long run, and all are far better than having nothing and just searching randomly. But, they are only about "where to search first", they are not evidence. One still has to find evidence, and real evidence that points to a real individual always trumps these things. They are just suggestions to help maximize the use of limited resources, they do not solve crimes, and the offender is not always in the hotspot (those are just the ones you hear about in the press).




And here's Dragnet's (Canter) output, while different again in some ways, the main focus is again around Kelly and Chapman's area.




So, while these are interesting to some, they should be viewed as "solutions" by none.

- Jeff

[Darryl Kenyon]

As anyone done a Geo profile with not just the C5 ? It would be interesting to see the results
Regards Darryl

[JeffHamm]

As anyone done a Geo profile with not just the C5 ? It would be interesting to see the results
Regards Darryl

Hi Darryl,

I've done a bunch of variations, such as dropping Stride, or including Tabram and Millwood, or including MacKenzie, etc. They're posted somewhere, I think the in Escape from Mitre Square thread. It's a long thread though, so it may take some time to find them; but they should occur over a few pages, so you could check only even pages numbers to get through it quicker. The version of the routines I was working on then were an older iteration of the one posted here, but they will be close enough for present purposes. There's not a lot that changes, with the most dramatic change being when Stride is removed. Her location rounds out the locations to form a sort of circle, when she's removed things shift much more crescent shaped, but roughly the main area of interest remains to the west, and generally up towards Kelly and Chapman. If JtR is a commuter, which is less common these days and probably would be in 1888 as well, it might be indicating that he enters into Whitechappel near Commercial and Hanbury, so from the north. But basically, up around the area the C5 produces, which is where the police were focusing their interest at the time as well, tends to get highlighted.

For example, here's the solution if we add Millwood, Tabram, and McKenzie (so the C5+3 additional). And the main area of interest stays focused around Kelly's location, but shifts a bit south, including areas around Goulston Street (the yellow square is the graffiti location; red squares the crime locations, and the teal squares are various suspect locations). However, the area between Kelly and Chapman remains in the orange area, which for this set of routines comprises the 2nd through 4th priority zone (this shows 40 zones; the first one I posted above shows 80; for analysis purposes I need to know these things, for pragmatic purposes, probably only the first 10 zones would be helpful, and more likely just the first 5. If that didn't produce anything of interest, the specific case is probably "not like the others." Sometimes these are more useful when there's a list of potential suspects, as then one can look at how their known anchor points fall on the profile. If you've got someone with lots of anchor points all falling in fairly high zones, they might be worth checking out before someone who doesn't. You never discard someone just because they don't fall in the "right place", this is just a way to help provide an order, as that may result in finding something of interest sooner, when there's nothing that really indicates which person on the list to pick first).



- Jeff

[Darryl Kenyon]

Thanks Jeff excellent work, very much appreciated.
Regards Darryl

[JeffHamm]

Thanks Jeff excellent work, very much appreciated.
Regards Darryl

No problem Darryl.

- Jeff

[erobitha]

Dr Chainey's profile...

[JeffHamm]

Dr Chainey's profile...

Thanks. There's a few issues most of us would take exception too (probably the biggest two being the time of Polly's murder, and the idea that Deimshutz saw someone flee when he entered Dutsfield's yard). Also, the inclusion of Emma Smith's murder gets discussed, but that's a bit of a minor point as it's just part of a side line.

Anyway, from the profile he shows it's a Rigel profile, like the first one I posted above in post 3, only he's limited it to the top area (roughly corresponding to the pink, yellow, and half of the orange areas in the one I plotted; based upon my own analysis of Rigel, that's roughly where 50% of cases end up. That reduces the seach space to about 7.5%, so you get a 50% return after searching only 7.5% of space, which is a good result in the long run, but also means there's an equal chance that this case isn't in the indicated zone). I tend to continue the profile out to less practical areas because it's necessary for proper analysis, but that's for theoretical work not practical application. The needs are different.

- Jeff

[clark2710]

Hi clark2710,

The short version is yes, Rossmo's "Rigel" routines have been used (they tend to focus on Flower & Dean), and Canter's "Dragnet" has also been used. I've done some research in this area and have been developing my own routines as well. In testing, though, all 3 do pretty much equally well in general, though there can be some differences on individual cases.

There's a lot of misinformation about what this sort of analysis does. It does not "solve" a case, rather, it's a way of prioritizing physical search space so that over a number of cases one should do better. Think of it this way, our family, friends, co-workers, people at the pub, etc, are all part of our "inter-personal space". Most murders are between people that know each other, with partners being very high on that list, and other family members, and so on down the line. With no other information to go on, one can organize that inter-personal space into high priority people to interview down to low priority people. If you did that, you would locate the offender far sooner than if you simply searched all the people in the city at random. Over the long run, that's an efficient use of resources, and helps reduce the time to a result. But, and this is the important part, not every murder is committed by a person's partner, or a family member, or a friend, or co-worker, etc. And some are stranger murders. So even that prioritized list isn't "the solution", it's just an order to the search that will reach the solution faster.

You can think of geographical profiling as doing the same thing. The higher priority zones are just the spatial locations that, in general, are more likely to contain one of the offender's anchor points. It's often thought this is the offender's residence, and it often is, but it may also be their place of work, or maybe it's the pub they go to, or their church, or even their partner's place of work. Basically, the anchor points are locations that the offender has some connection to in their every day life.

And also, the analysis routines are based upon offenders who are "marauders", which just means they are committing crimes in the areas they spend their normal daily routines (live, work, entertainment, etc). If they specifically travel to a location out side their normal daily routine, and then search for victims, they are a commuter. In this case, the profile may simply locate an area where the offender tends to enter the region, perhaps an off ramp from the major highway, etc. It might still indicate a good spot to look for them, but if they have no other connection to that location, how would you find them since off ramps (in this example) are used by hundreds, or thousands, of people?

Here's the Rigel (Rossmo's) routines profile for the C5. The pink area is the highest priority zone, followed by yellow, then orange, then light red, dark red, green, cyan, and blue. 50% of offenders get located by the time you search the orange area, and 75% by the time you reach the end of the light red area (sorry, hard to see the difference here I'm afraid). After that, things tend to trail off quite a bit, so they're just for completeness. Oh, those numbers are based upon a set of serial offenses where the number of commuters is grossly under-estimated as well, but it's hard to get those because they're trying to optimize the search when the offender is a marauder (it's a problem in this area that really needs to be addressed).




And for comparison, here's the output from my own routines (I've left out the light and dark blue here, as they tend to contain almost no cases). Interestingly, JtR is one of the situations where the different set of algorithms produce fairly different looking solutions, but it's a matter of degree (Flower and Dean is still in the orange in mine, and Dragnets, and the area between Kelly and Chapman fall in the orange and light red regions for Rigel). Mine and Dragnet focuses more around the locations of Kelly's and Chapman's murders (with Kelly's right in the highest probability region), while Rigel is shifted to the east around Flower and Dean, with interest around Middlesex as well (which is interesting given it's between Mitre Square and Goulston Street; if JtR were there, and went home for a bit before heading back out to get rid of the apron piece and write the graffitti, it makes some sense - he's not travelled far before dropping it, and he headed away from Mitre Square, which could be seen as an attempt to mislead the police by suggesting he had gone further along that line, rather than extending the line further, if that makes sense). Rigel also points a bit to the North West, which is curious. Anyway, for the most part, all three tend to result in similar areas ranking high priority, though there can be some cases, like this one, where they can look more variable.

Again, all 3 do equally well in the long run, and all are far better than having nothing and just searching randomly. But, they are only about "where to search first", they are not evidence. One still has to find evidence, and real evidence that points to a real individual always trumps these things. They are just suggestions to help maximize the use of limited resources, they do not solve crimes, and the offender is not always in the hotspot (those are just the ones you hear about in the press).




And here's Dragnet's (Canter) output, while different again in some ways, the main focus is again around Kelly and Chapman's area.




So, while these are interesting to some, they should be viewed as "solutions" by none.

- Jeff

may i ask do you have a cross reference on the suspects that lived or worked in that area?

[JeffHamm]

Hi clark2710,

here's the map with no overlay, with a key at the bottom. I've left the Chapman marker in place, but be aware he didn't move there until after the murders. Also, the upper left Peabody House doesn't have a suspect, but I found information on it and it sounded like it provided low rent housing, etc, and so interested me as a "that could be where he lived"? type moment.

Also, the Druitt location is a relative's (cousin I think) office, and to my knowledge it is only theoretical that Druitt every visited it.

- Jeff

[JeffHamm]

Oh, I've never included Smith, as I don't think it likely she's part of the series. And, when I've included Millwood, I always include Tabram, as their attacks could be viewed as related (it's not a for sure thing though), but if Tabram isn't part of the JtR series, then Millwood isn't either as Tabram would act as the link in that chain.

In the end, though, their inclusions just sort of jiggle things, rather than change anything dramatically. Same with McKenzie, in or our, not a lot really changes.

There is justification for including Stride but not Eddowes, simply because if they are both JtR victims, once Stride was killed, Eddowes is no longer an independent event, rather her location becomes influenced by Stride's location. As I recall, at least with my own routines, the high priority zones end up in roughly the same location (around Dorset Street).

Curiously, I think there was a pub at the end of Dorset Street that had some story about a man with blood on his hands? I could be completely making that up, but it just popped into my head. I know there's some tale of a man going into a pub with blood on his hands, or cuffs, but not sure where the pub was though I have this memory it was at the west end of Dorset Street? But I may be lying, so don't quote me on that.

anyway, here's the "no Eddowes" version, and as you can see, the hot spot area is in the same place.



- Jeff

[caz]

There is justification for including Stride but not Eddowes, simply because if they are both JtR victims, once Stride was killed, Eddowes is no longer an independent event, rather her location becomes influenced by Stride's location.

That's an interesting observation, Jeff.

For those who believe Eddowes was the next independent event, because they see no possible connection with the earlier murder in Berner Street, it becomes slightly less easy for them to explain the one-off move to the City police area, before returning to 'home turf' for MJK.

Love,

Caz
X

[JeffHamm]

That's an interesting observation, Jeff.

For those who believe Eddowes was the next independent event, because they see no possible connection with the earlier murder in Berner Street, it becomes slightly less easy for them to explain the one-off move to the City police area, before returning to 'home turf' for MJK.

Love,

Caz
X

Hi Caz,

If JtR were aware of where the police boundaries were, and the implications of that, then yes, I can see an argument as to why he might have chosen that location for Eddowes following the Stride murder. Whether or not he was so aware is something we may never know, but it is an idea that has some interesting implications with regards to Eddowes' crime location.

The problem, though, is for us to accept that line of reasoning, we have to suggest that JtR was thinking pretty quickly on his feet to both decide he was going to commit another murder that night, and also to specifically target this impromptu murder in area within City Police territory. And to be confident he would find a victim in time. I would think doing that would require more specific planning, so could see an argument that it would be more likely if, in fact, Stride was not a JtR victim. (I can see it working either way, basically).

Regardless, it's an interesting idea to ponder, but what was actually going on inside JtR's mind at the time, even if we were to solve the case, is something that I think is lost forever.

- Jeff

[rjpalmer]

Curiously, I think there was a pub at the end of Dorset Street that had some story about a man with blood on his hands? I could be completely making that up, but it just popped into my head.

It sounds like you are thinking of Mrs. Fiddymont's suspect in the Prince Albert, 21 Brushfield Street, near the corner of Steward Street. Looking in the mirror, she noticed he had caked blood on his hands.

When the man was followed, he walked west towards Bishopsgate, and then turned south.