Joe's Telex - entries for 2010


NIM and intelligence

NIM and statistics

Job descriptions


Dysfunction of NIM

The role of Principal Analysts

Legal Status of NIM

What then must be done?

Joe's Blog


Copyright notice: text from this website may be reproduced freely, so long as is quoted as the source

About this website
Click here for current Telex's

15 December 2010 - public bar wisdom on analysis

The following is a layman's view of analysis in the police. Read the post in its entirety at

Twenty five years ago every police station in Sussex was open 24 hours a day. Fast forward to now and Sussex Police now employs hundreds more people, both police officers and 'support staff'. Yet Sussex Police Authority have just approved plans to shut many police stations altogether, and reduce the opening hours of all the others ... Where then have all these extra resources gone? Well, a medium size police force like Sussex now employs over 70 analysts and researchers. These do a job which used to be undertaken by about 10 collators in 1985. They were police officers nearing the end of their careers, who put their detailed local knowledge into producing daily briefings for their colleagues. There was also the added benefit that if there was any kind of emergency the collator could turn out and help deal with it.

The 70 analysts are all well qualified, often holding degrees in subjects such as criminal psychology or behavioural profiling, imagining they were about to become a real life Cracker. Instead they devote themselves to churning out endless reports telling you that violent crime goes up on Friday evenings in town centres

... Why not get some of these researchers and analysts to do their work in the front office of the police station? I am sure the researchers will not like it, 'It is not in my job description to actually deal with public'. But if the claim to justify these closures is that no one calls there anymore then they will not get interrupted anyway?

The police service can manage without much of the analysis in any case. Most of it is only produced to feed a voracious beast called the National Intelligence Model which no one apart from its inventor understands anyway. As a well respected senior detective once said to me in passing - 'I have never stood on the steps of the Crown Court after a big trial and thought 'Thank God we had an analyst on this job''.

3 December 2010 - Forensic Engineer's opinion on Amanda Knox case

Provided previously (27 June 2010 entry, below) are links to a retired FBI agent's analysis of the Italian police investigation into American Amanda Knox and her Italian boyfriend Raffaele Sollecito for the 2007 murder of British student Meredith Kercher.

There is now available a series of articles by retired Forensic Engineer Ron Hendry on the crime scene, in which he points out many errors in the police investigation. Intelligence analysts do not often visit crime scenes, so I recommend these articles as a case study in crime reconstruction: Professional Opinion From Forensic Engineer Ron Hendry

26 November 2010 - Imperial Tobacco vacancy

Imperial Tobacco is advertising the above Intelligence Analyst vacancy in Police Review. According to the text, the role is to “protect against the illicit trading of our brands” and to “investigate, analyse and report on illegal and counterfeit activity”.

So far, so good. The advertisement then adds: “You’ll be a proven statistician” and will identify “themes, trends and patterns”.

I asked Imperial Tobacco the following question:

"I have a query regarding the vacancy for Intelligence Analyst, which I saw in this week’s Police Review magazine. The advertisement says that the role is investigative; however, it also says the role is statistical. Please can you clarify; is the vacancy concerned with tackling individual counterfeiters and illicit traders, and therefore investigation and intelligence, or is it a statistical/measurement/planning role?"
The following reply, received 22 December, indicates that the post is not that of Intelligence Analyst:

"I can confirm that the role is more around using data and manipulating it to understand trends etc. The intelligence aspect of the role is around applying this skill to the data in order to highlight and pass on any areas of concern to other members of the team who may look to investigate further".
1 November 2010 - response to a critic

The following message (in full) was sent to this website, by ‘umble Analyst:

So many things wrong with this, it's unbelievable. Hot spots - you claim most hot spot maps just show officers what they already know, yet this has been scientifically tested and proved not to be the case. I've sat in meetings and watched an area Inspector explain a crime issue was occurring at a specific time for a specific reason when the real times/reasons were in front of him on a presentation everyone could see! The two things didn't match! I've been involved in analysis that has led to criminals being caught within minutes of interventions being put in place thanks to smart analysis.
Your entire site focuses on a very singular offender-focused approach to analysis - in truth, a very narrow element of the analytical spectrum. A sad irony given your claims for analytical free-will!
The correspondent’s first paragraph refers to this statement (see in context here):
Crime statisticians use hotspots for two purposes: to allocate police resources, and to understand why crime is committed in a particular environment to aid crime prevention. The numerical and mapping skills used to define and identify hotspots do not assist investigating officers for two reasons: first, the officers work in the same organisational unit as the statistician and already know in broad terms what the statistician is going to tell them. Second, that a number of crimes occur in an area is no more a reason to assume that they are linked, than crimes in different areas can be assumed not to link. Until the crime report is read and researched with reference to the actions of who committed it, the investigating officer is not aided.
Note that my text specifies investigating officers, not “area inspectors” as referred to in ‘umble Analyst's message. The difference is that area inspectors are office-based and rely on reports and statistical analysis to understand what is going on; investigators work in the field, already know (generally) what is happening in their vicinity, and so seek information from outside that vicinity, linked specifically to their cases. This is why statistical hotspots assist people such as area inspectors, but not investigators, whose perspective is not area-based.

I think my original paragraph explains this plainly, so why does ‘umble Analyst think that I refer to area inspectors? The reason, I suggest, is that investigators simply do not register in ‘umble Analyst’s thinking. Why might this be? He goes on:
Your entire site focuses on a very singular offender-focused approach to analysis - in truth, a very narrow element of the analytical spectrum.
This tells us that ‘umble Analyst is not interested in investigation; in fact, he has a disdain for it. He even attempts to sideline it as a “very narrow element”. This suggests that his objection to this website is not that its message is incorrect - that intelligence analysis is under attack, and must be kept separate from statistical analysis - but that this website is entirely correct, and that he is one of the attackers, and does not wish to see intelligence analysis defended or take its place as a recognized profession.

I shall speculate on the correspondent’s motives below but, before that, we must address the statement that intelligence analysis is “a very narrow element of the analytical spectrum”. The claim is incorrect in two respects; first, there is no such thing as a “spectrum” of analysis; there are only two types of data analysis: one based on categories of people and events (statistics), and the other based on actual people and events (intelligence). All data analytical tasks fall under one of these two disciplines.

Analysis is not, therefore, accurately thought of as a spectrum of colours and shades that merge from one into the other, but as two blocks, black and white. People get bogged down in the “spectrum” idea because they dwell too much on what analysts are used for (which can be many things) rather than what analysts are employed as, which can be one of only two things: investigators or statisticians.

Nor is it true that intelligence is a narrow element. Let us run through some of the skills and duties that people employed as intelligence analysts are used for:

- Linking crimes (and incidents) and identifying offenders by researching modus operandi and suspect/vehicle description, etc.
- Investigation of telephone records and other communications (emails, etc), plus cell site analysis.
- Investigation of financial records, and transactions.
- Suggesting the existence of as yet undiscovered information (“going beyond the known facts”).
- Looking for evidential gaps (or flaws in evidence) and ensuring investigators are aware.
- Reconstructing crimes from statements and other documents (sequences of events).
- Identifying inconsistencies in statements and documents.
- Linking information from other sources, whether police sources, such as forensics and surveillance, or public sources, such as the internet and newspapers.
- Confirming or refuting informant information; assisting handlers in developing informants.
- Holding the reins on complex investigations, or enabling others to do so; telling investigators when they are barking up the wrong tree.
- Giving evidence in court, undergoing cross-examination.
- Attending court to hear defence cases and verify their points.

A “very narrow element”? Seems pretty wide to me. Yet, criticisms that seek to diminish or sideline intelligence analysis are not uncommon. Consider the three following quotations from the book, Integrated Intelligence and Crime Analysis; Enhanced Information Management for Law Enforcement Leaders (available here):
Jerry Ratcliffe (website): "Some definitions [of intelligence analysis] limit intelligence as a process that targets individuals or groups of offenders. Indeed, many in policing see criminal intelligence solely as a mechanism to examine the behavior of individual offenders or organized crime groups, separate from the main patrol functions and high-volume crime focus of most police officers. There is a growing realization, however, that at the opera­tional and strategic levels of law enforcement a deeper understanding of general criminality may be required, beyond simply trying to understand the behavior of individual offenders. The definition provided here places emphasis on intelligence that supports crime reduction and prevention as well as enforcement" (p8).
Christopher Bruce, President, International Association of Crime Analysts: "I’m not sure that most U.S. criminal intelligence analysts are trained to do much more than case support. Moreover, I’m not sure they see their own jobs as being more than case support. The problem lies largely with the way the profession has been sold to intelligence analysts in the United States, with its focus on criminal organizations, homeland security, and high-profile arrests and prosecutions. I think that many, if not most, criminal intelligence analysts in the U.S. prefer to think of their duties as providing analytical support to such endeavors. If criminal intelligence analysts thought of themselves more in a crime prevention role, or even a pattern analysis role, then certainly the intelligence analysis and crime analysis professions would have merged in the U.S. some time ago" (p9).
Deborah Osborne, blogger and author on crime analysis issues: "Analysis of investigative data in aggregates rather than case-by-case could help in discovering new patterns that might enhance both investigation and prevention of crime. What are the characteristics of burglary victims in a neighborhood? When are all the robberies occurring in a retail district? What are other factors that contribute to crime in a specific region? Weather? Events? How might the police be better deployed to address some of these issues? How might the community become involved by enhancing capable guardianship? Who are our vulnerable population and how might they be better protected?" (p10)
And the following is from the RISC training manual in intelligence analysis:
“In all cases, it is better to prevent than to detect crimes … the sad reality is that with too little resources, a reactive response to chasing the problem ensues (p33) … Why then does there seem to be a propensity for detecting crime compared to preventing it? (p34) … [linking crimes is defined as] comparing the information on similar criminal incidents with a view to discovering whether some of them may have been committed/organized by the same offender(s). You may wish to challenge this if you are aware of Problem Orientated Partnership [because] it may transpire that a comparative case analysis of locations or potential victims is also useful to design out the potential crime rather than seeking to chase the perpetrator after the event” (p69).
Pay close attention to what is being said in this fourth example: a training course in intelligence analysis is seeking to dissuade trainees from doing it.

Two drivers of anti-intelligence analysis sentiment can be identified in the above quotes; first, a lack of understanding of what it actually is. Mr Bruce defines intelligence analysis as “case support”; RISC say that it is chasing offenders “after the event”. Both these descriptions are inaccurate. It is without intelligence analysis that detectives may find themselves limited to investigating cases one at a time, after the event. Combining the detective with an intelligence analyst changes the dynamic so that the detective investigates one criminal linked to many crimes, and newly committed offences can be fitted into a known picture.

Second, the quotes are motivated by a preference for crime prevention over crime investigation: the first three feature the word “prevention”, and the fourth says “design out the potential crime”. The email from ‘umble Analyst refers to “crime issues” and “interventions”.

There is nothing wrong with people preferring crime prevention, or promoting the benefits of statistical analysis. But there is something very wrong in basing that promotion on subverting another profession. As illustrated elsewhere on this website - by job descriptions, training courses and software - the meaning of intelligence analysis has been corrupted to include statistical tasks: area profiles, hotspots, thematic reports, strategic assessments, seasonal studies of crime, etc. Hardly surprising, therefore, that new analysts think of intelligence as statistics. Recruitment of intelligence analysts can thus have the perverse effect of undermining the intelligence analysis profession, rather than reinforcing it; the email from ‘umble Analyst is an example of this.

The breakout comes when an analyst, perhaps by accident, links two crimes that cut across statistical categories. If the analyst has an investigative nature, the penny will drop that intelligence is not the same as “trends and themes”. From this point, the analyst’s perspective shifts; the analyst cannot be re-educated in the former way of thinking.

Instead of welcoming this natural division of labour, crime prevention progressives chide analysts for not thinking "correctly". What they may not grasp is that, by refusing to recognise analysts on the intelligence side of the fence, they are also reducing recognition of analysts on the statistical side; for example, analysts who consciously wish to become do crime prevention can find themselves being called upon to undertake investigations, which they are not positioned to do. This compromises investigators, who are not getting the correct type of analytical support and this, in turn, undermines intelligence analysts, who find that investigators do not trust them due to a belief that “analysts” do not have investigative skills. Consider this quote from an irate police officer, published in The Times in 2008:

“The experienced police officers are leaving in droves owing to management inefficiencies and incompetence and we are being left with a lot of very clever analysts and the like who wouldn’t know a Mr Big if he pulled out a gun and pointed it at their heads” (source).
Either the "very clever analysts" referred to had no investigative outlook or detectives were unwilling to trust them due to previous bad experience. Such sets of circumstances can force both intelligence and statistical analysts into a no man’s land, leading to departure of experienced staff. New analysts are recruited, and the cycle starts again.

The bottom line is that, despite the large number of official bodies, institutes and organisations that espouse “integration” of intelligence and statistical analysis, not one has explained the actual means by which an employee can combine the two; not the Jill Dando Institute, the National Police Improvement Agency, ACPO, Her Majesty’s Inspectorate of Constabulary, the Research and Intelligence Support Centre, Lancashire Police, i2, or the former National Criminal Intelligence Service, despite their relentless insistence over ten years that intelligence and statistics are the same profession. This insanity must stop.

11 October 2010 - Jill Dando Institute gets it wrong, again

This week, the Jill Dando Institute (JDI) launches a new training course, Advanced Crime Intelligence Analysis. Here are some observations on the course outline:

This course is aimed at those that want to contribute to the development of crime intelligence analysis – and wish to engage with like-minded professionals in debate and future thinking.
Normally, training courses are about the development of trainees; the course reverses this and aims itself at people who want to develop intelligence analysis itself. In other words, JDI is not regarding intelligence analysis as a profession to be taught, but as a topic that is subjective and to be debated. The result is that intelligence analysis is denied; it becomes a hobby horse.
Entry requirements: Analysts should have completed the JDI ‘How to Become an Effective Problem Solving Crime Analyst Course’ or an ‘Initial’ or ‘Introduction to Analysis’ Course or have experience in managing / supervising analysts.
The course in "How to Become an Effective Problem Solving Crime Analyst" has nothing, repeat, nothing to do with intelligence analysis. It is a statistical analysis course aimed at crime prevention. The book that accompanies the course explicitly disowns intelligence work: “The first instinct of police, even when they have been involved in a detailed analysis of a problem, is to try to solve it by beefing up enforcement. You [trainee analysts] should expect this and not oppose it, even if the impact is usually short-lived” (Chapter 6). This tells us that the people running JDI courses have no identification with, or interest in, intelligence analysis.

But let’s get to the meat:

Advanced Crime Intelligence Analysis aims to integrate crime analysis (the systematic study of crime and disorder problems) with criminal intelligence analysis (a primary focus on offenders) and critical thinking skills.
JDI is wrong on two counts. First, it is not necessary to integrate “study of crime and disorder” into intelligence analysis, because intelligence analysis is a study of crime and disorder. Where intelligence analysis differs from statistical crime analysis is that crime and disorder is studied in relation to individuals that cause it. JDI has instead assumed, wrongly, that “focusing on an offender” means a lack of attention to crime and disorder. The correct understanding is that crime is the activity of the offender, which means that focusing on the offender, and focusing on the crime, is the same thing.

The JDI’s second error is to state that integrating statistical crime analysis into intelligence analysis is even possible. It is not, and here’s why: in order to identify linked crimes, you have to separate crimes by offender. You cannot separate crime by offender if, at the same time, you are grouping crime by some other frame of reference. Get it?

JDI’s problem arises from its intruding in areas that are outside its competence. Its declared remit is statistics and crime prevention, not intelligence and investigation. Consequently, JDI can only conceive of study of crime in the statistical sense; it knows nothing else. So, when JDI says it needs to “solve” intelligence by “integrating crime”, it causes a problem, because it is importing statistical studies of crime and disorder into a non-statistical profession.

21 September 2010 - Analysts' reporting etiquette #2

Sometimes, intelligence analysts are instructed to report findings to weekly management meetings, responsible for the direction of personnel and resources. This advice is incorrect, or at least incomplete.

The person to whom the analyst reports is the officer-in-charge of the case that is the subject of the analyst’s report. Always. The reason for this is that the said officer has been authorised by the law enforcement agency to investigate that particular offence. In this context, a constable is equal to the chief constable; it is a matter of hats, not rank. To address findings to someone other than the allocated officer is a breach of protocol, to say nothing of manners.

How do analysts reconcile their obligation to the case officer with a requirement that they report to weekly management meetings? The solution is to address the report to the case officer and cc it to all other interested parties, be it the weekly management meeting, detectives in linked offences, or specialist squads (if, say, an analyst identifies a suspect for a credit-card deception, the report should be copied to the fraud squad).

In this way, the intelligence analyst can accommodate the needs of all parties, while acknowledging the authority of the case officer.

1 September 2010 - Intelligence or Investigative Analyst?

A visitor to this site recently made the following comment: "I think the problem partly comes from the ambiguous job title 'Intelligence analyst' - why not have 'Investigative' analysts and 'Strategic' (in the stats sense) analysts, in true reflection of job expectations?"

In many UK police forces, the title intelligence analyst was preceded in the 1990s by 'crime analyst'. This was understood in the intelligence/investigative sense, in contrast to the USA, where crime analyst was (and still is) taken to mean crime statistician. 'Crime' analyst was changed in the UK to intelligence analyst to emphasise that the role concerned individuals and to avoid any possible confusion with the American definition. Some forces modified 'crime' analyst to 'criminal intelligence' analyst or 'crime intelligence' analyst, for the same reason.

As noted on this website, the term 'intelligence analyst' has since been corrupted to mean statistical analysis. So, both the terms crime analysis and intelligence analysis are now compromised.

Should the profession's title change once more, to investigative analysis?

The first question to consider is whether there is any difference between intelligence analysis and investigative analysis. No, there isn't. Let us suppose that we did make a distinction; an intelligence analyst might be someone who links in the following order: person to person, person to crime, crime to crime.

An 'investigative' analyst might proceed in the reverse order: crime to crime, crime to person, person to person.

The standpoint of this website is that there is no difference in principle between the two approaches; it does not matter at which end the analyst starts: the person, the crime, or a piece of information in-between. Nor does it matter whether the analyst is starting within the formal framework of a case, or as a result of informal research. So long as the analyst has a 'starting point', he or she will research in as many different directions as possible, and will end up with the same picture as had they started somewhere else, at least in theory. So, where or when the analyst starts is a matter of circumstance and preference; it does not change the job function. For this reason, the terms intelligence analysis and investigative analysis are interchangeable.

The advantage of the title investigative analyst is that it is much more resistant to misinterpretation or misrepresentation. An analyst who explains, "I am an investigative analyst" is certain to be understood by just about everyone. People might not be too clear on the "analyst" part, but that is solved by education.

The disadvantage is that some might interpret 'investigative' too narrowly, such as that the analyst is assigned to existing cases, rather than also prompting or revisiting cases via exploratory intelligence development. This is a serious drawback, and would again have to be dealt with by education.

On balance, I think that the advantage outweighs the disadvantage, so I have added the correspondent's suggestion to the page, "What then must be done". I will, however, continue to use the term intelligence analyst on this site.

30 July 2010 - Analysts' reporting etiquette #1

When an intelligence analyst researches crime, the aim, preferably, is not to produce a string of crimes that are merely similar, but crimes that are believed to be linked. But even having selected such crimes, the analyst must never declare them linked; the analyst should use phrases such as, "These crimes are believed to link, because …"

The highlighted words mark the dividing line between the role of the intelligence analyst, which is to produce intelligence, and that of the detective, which is to produce evidence. The analyst can believe that the crimes are linked, and be right, but the detective wants to know why the analyst believes this, and that is where the strength of the analyst's report must lie.

Where a presumed link is based on something other than similar modus operandi or suspect description, such as a matching car number plate or forensic link, the analyst should use the phrase "according to"; for example, "… according to witness statements, both crimes featured a car with the same number plate". Again, the analyst is leaving one step between themselves and the evidential stage.

This rule applies even in the areas of telecommunication, where the analyst can perform a forensic-like role, linking phones and computers. While calls and emails can make links with absolute certainty, the analyst has no first-hand knowledge of who is using the phone or computer. So, even here, the analyst must use words like 'believe' and 'because' when referring to the likely user.

27 June 2010 - FBI Agent's opinion on Amanda Knox case

Retired FBI Agent Steve Moore has written a series of articles on the investigation by Italian police into the 2007 murder of British student Meredith Kercher, resulting in the conviction of American Amanda Knox and her Italian boyfriend Raffaele Sollecito. Mr Moore works his way through the investigation methodically, demonstrating either a miscarriage of justice or, more likely in his view (and mine), a perversion of justice. Among many other things, he quotes investigator Edgardo Giobbi: “We knew she was guilty of murder without physical evidence”.
Investigation of Violent Crimes is My Life; Not a Hobby
The Mountain of Missing Evidence
Evidence Collection
The Luminol Lies
The Interrogation That Never Was
20 June 2010 - job description from East Mids Special Ops Unit

The East Midlands Special Operations Unit is advertising a new intelligence analyst vacancy. Is the job description any good? Here are the short-listing requirements that candidates must match:

The "essential" column includes qualifications in mathematics and statistics. In the "desirable" column, a degree in social science is suggested. This is not a vacancy for an intelligence analyst, but a statistical analyst. So, why doesn't the East Midlands Special Operations Unit call it what it is?

Five duties are listed for the post-holder:

It seems to me that the second and fourth duties are different ways of saying the same thing.

The role description falls apart irreparably in the third duty, which says the postholder will compile criminal profiles on "individuals and groups". This conflicts with the short-listing for mathematicians and statisticians which, in turn, conflicts with the job title, "intelligence analyst".

14 June 2010 - why Human Resources staff cannot be trusted

I saw an Essex Police advertisement for "Senior Intelligence Analyst", so I dropped the Human Resources manager a line to ask: is this an intelligence role, or is it statistical? Here is the reply:
"The job description supplied for the position is a generic Senior Analyst which covers all aspects of policing. This particular position is primarily an intelligence and investigation role. However an awareness and knowledge of utilising statistics for predictions is desirable particularly for the long term strategic planning if required".
The HR manager says that the job description is "generic". This is an incorrect use of the term generic. The purpose of a job description is to describe a job. Note the use of the singular; job description; not jobs description. The term "generic job description" is supposed to apply when specific details are omitted; an example is a typist, whose skills are transferrable between departments. The generic version of the job description will not specify any one department; this does not affect the job role.

The HR manager is instead using the term "generic" to justify adding functions to the job advertised. She says that the position is "primarily" intelligence and investigation, but that the post-holder must also use statistics for predictions and planning "if required".

The result is not a generic job description, but a licence to harass, which she is helping to draw up.

A competent HR officer would realize in an eye-blink that any such practice is invalid. Job profiles provide the basis of contracts between employees and employers; they are necessary for employees to identify jobs they understand and wish to do, and for employers to employ the right people in the right positions. The role of HR is to uphold the integrity of the employment process. Yet, here we have an HR manager who is undermining that process.

The grim reality is that many (most? all?) HR staff in the police are like the one above. They are not HR staff in the proper sense, but administrative staff who see their function primarily as implementing whatever instructions are given to them. If a police manager says: "put two jobs into the same job description", they will do it. If police managers say, "apply the Chief Constable's annual plan", HR staff can't draw action plans on whiteboards fast enough. Absent are the standards of scrutiny and objectivity that one would expect from Human Resources departments. This is why intelligence analyst job descriptions have reached the state they are in.

6 June 2010 - Return of the Body Snatchers

I have just finished reading the second edition of Integrated Intelligence and Crime Analysis by Jerry Ratcliffe, which I will Telex on at a future point [update: see Telex entry of 1 November 2010, above]. I draw attention to the quote on page 9 by Christopher Bruce, President of the statistics-orientated International Association of Crime Analysts. The extract is representative of sentiments in the rest of the book:

I'm not sure that most U.S. criminal intelligence analysts are trained to do much more than case support. Moreover, I'm not sure they see their own jobs as being more than case support. The problem lies largely with the way the profession has been sold to intelligence analysts in the United States, with its focus on criminal organizations, homeland security, and high-profile arrests and prosecutions. I think that many, if not most, criminal intelligence analysts in the U.S. prefer to think of their duties as providing analytical support to such endeavors. If criminal intelligence analysts thought of themselves more in a crime prevention role, or even a pattern analysis role, then certainly the intelligence analysis and crime analysis professions would have merged in the U.S. some time ago.
The quote is an example of what I mean by my reference to Invasion of the Body Snatchers (see here). To demonstrate the point, I will reverse Mr Bruce's comments:
I'm not sure that most U.S. crime statistical analysts are trained to do much more than management support and planning. Moreover, I'm not sure they see their own jobs as being more than management support and planning. The problem lies largely with the way the profession has been sold to them, with its focus on area policing, crime prevention, and high-profile patrol deployment. I think that many, if not most, crime statisticians prefer to think of their duties as providing analytical support to such endeavors. If crime statisticians thought of themselves more in a crime investigative role, or even an intelligence analysis role, then certainly the crime analysis and intelligence analysis professions would have merged some time ago.
I'm sure Mr Bruce does not appreciate people talking about his profession in this manner, so why does he do it to the intelligence analysis community? There's a solution to this; it's called the jobs market. If someone does not think intelligence analysis is worth doing, they don't have to do it. But leave alone those that do.

16 May 2010 - relationship between tactical and strategic analysis

Read this article: Intelligence Analysis Needs Priority (local cache)

The linked article concerns U.S. "fusion centres", set up in the wake of 9/11 to centralise information for more effective analysis. My interpretation is that the writer uses 'tactical' and 'strategic' to mean intelligence analysis at different levels, as opposed to strategic to mean statistical analysis. I draw attention to this excerpt:
"Several managers at the centers I have visited have told me that … personnel focus on a tactical response, and often strategic analysis never occurs. When the tactical mission overtakes the strategic mission, it becomes impossible for the fusion center to paint the "big picture" of major plots or threats".
The gist is that tactical level intelligence is limited and can get in the way of strategic analysis. I have not visited a U.S. fusion centre, but I suggest that something is awry with their understanding of the relationship between intelligence analysis at the tactical and strategic levels.

A practical example: I once did investigative research into thefts of spectacle frames from opticians. Criticism was conveyed from above that it was inappropriate for an intelligence analyst to look at such minor matters as shoplifting. Quite properly, I ignored this view and carried on. The thieves were oriental (Vietnamese, as it later turned out) which made it straightforward to find linked offences. I did not confine my research to easy-to-spot thefts of frames by oriental offenders, but widened it to include thefts of frames (on the same dates) where no offender description was available; and shop thefts (again, on the same dates) where the offender was said to be oriental, but where the stolen goods were other than spectacle frames. In this way, the group's modus operandi was found to include theft of a range of items, including household goods from home furnishing stores. For some reason, they liked to drive Honda Accords, which expanded the search perimeters further, and led to the discovery of more linked crimes.

I put together thirty or so offences and sent a report to neighbouring forces which undertook similar research, and found dozens of matching offences. Because of the scale of the problem, investigating officers went to London for a national meeting and found themselves talking to representatives from the FBI and Interpol. It transpired that the Vietnamese shoplifters were the financing base of an international drugs and kidnapping operation - and all this was going on right under the noses of my police force which would have known nothing had the errant analyst not done some very basic research.

The point that I make here is that there is no meaningful division between intelligence analysis at the tactical and strategic levels. If the analyst starts at the tactical periphery, he will arrive, eventually, at the centre, thereby completing the picture. If the analyst starts at the centre, he will ultimately reach the edges; again, the picture will be complete.

The writer in the above article offers this solution:
"Strategic analysts develop after years of training and work experiences, and they aren't hired right out of college. There must be a division of analytical labor. For example, it doesn't make sense to have highly trained analysts working on tactical daily activity like 'turning and burning' routine Suspicious Activity Reports".
I don't agree. The only division of labour that needs to be made is between analysts doing crime/criminal specific work (intelligence/investigation) and crime/criminal non-specific (statistics). To divide the former so that the expertise is confined to the "strategic" levels will not increase the intelligence but reduce it, since it eliminates half the potential opportunities for discovery (those at the local end). Moreover, without an overall context, how will the analyst know the strategic centres when he comes across them; strategic, in relation to what?

The correct solution is the opposite to that suggested in the article: to raise the investigative standard at the tactical level, and to let analysts, wherever they are, pursue leads as they present themselves, transitioning from one level to another (a bureaucratic distinction, in any event) until they are satisfied they have exhausted all enquiries, and the picture is complete.