They should call the system Sherlock...
...they've already got a Watson.
England and Wales residents will soon be able to vote in their own local cop chiefs. Police officers in 41 forces across the UK are set to come under the direction of the new officials - elected police and crime commissioners (PCCs) - from next week. And it seems IBM wants to have a word on how the new brooms plan to do their …
...they've already got a Watson.
since the Microsoft Kin.
We have another tier of a politically elected official that will cost the tax payer more money to administer. No one knows who they are or what they stand for except that they belong to a political party.
They are neither use nor ornament to anyone.
The worst thing about it is that IF they are elected they will consider they have a mandate, but the chances are there will be less than 10% of the population that will bother voting.
No one but their political masters actually want them. I say spoil your ballot paper with the words 'sod off'.
Expect a voter turnout of less than 3% and that is optimistic.
In fact, spoiling your vote paper will be a "win" for the politicians as they will be able to claim the whole thing is a success by the number of votes. Not voting is actually the best way to show that you don't want any part of introducing political party commissioners.
They've already had HOLMES (Home Office Large Major Enquiry System)
What on earth are they doing wasting all that money on SPSS when they can get GNU PSPP free???
Would you put yourself up for the PCC election for the salary of £85,000 a year?
Because that's what it will cost.
So the UK is following the USA and getting the populace to vote for police hierarchy.
a Are they going to go the whole way and have elections for the local dog-catcher, too and
b What does Joe/Jill average know about what makes a good commish?
A third question:
Does Joe/Jill Average actually know who they're supposed to be voting *for*? If this area is in any way representative, they'll have heard precisely bugger-all about who their prospective candidates are, much less if they're actually qualified to do the job in hand (OK, so chances are they're failed councillors who couldn't hack it at the big kid's table, but you get the idea)
The whole exercise strikes me as an example of devolution of responsibility.
For the first time in my long life I won't be voting on Thursday. Managed to track down the candidates' info - and they are all big political party hacks. Members of this council, that council, this committee, that committee. Nowhere could I find anything other than party line sound bites about their "mission". As there are no extremist candidates who might be let in by accident - then my non-vote is a protest about the whole process.
Committees are notorious for compromises - especially if there has been a reasonable attempt to recruit members from various areas of society. However in this case compromises are the way maintain some balance.
It was worrying that some apparently independent PCC candidates' £100k campaigns were being quietly resourced by right-wing USA organisations.
I'm another oldie who won't be taking up a vote for the very first time ever. I have not the remotest idea what qualities a Commissioner needs. Nor do have any idea whether any of the candidates have those qualities.
As a long suffering taxpayer (who hasn't had a pay rise for three years) I strongly resent the total waste of my money.
There are some dubious outfits trying to cash in:
One outfit called 'One Team Policing' offers consultancy for prospective PCCs. One Team appears to be run by "Mike Glanville, who recently retired as assistant chief constable of the Dorset Police ... You can get a “Standard Web Presence” for £359, or a “Premium Web Presence” for £1,799. Among the advice they offer are tips on “strategic leadership”, “cost reduction” and “working with the media”.
Another outfit is Crest - "Crest was established last year by Gavin Lockhart-Mirams ... The man who developed the policy on PCCs when the Tories were in opposition, then firmed it up for the Conservative Party manifesto, and saw it implemented in Downing Street, is now hoping to profit from it commercially."
Here is the DT article on the USA resourcing of at least one PCC "independent".
No! The UK comprises Scotland and Northern Ireland in addition to England and Wales, but it's only the latter two countries that are sleep-walking into this mad waste of money.
Have a read over at Inspector Gadget's blog for a thought-provoking analysis of it all... including the fact that the poor saps are being set up to take the flak when the whole system collapses and the finger-pointing starts.
20% cuts are bringing policing to its knees, and Theresa May's famous statement (29 June 2010, a few weeks after becoming Home Secretary) that the mission of the Police Service is "... to cut crime, no more and no less" rather seems to indicate that anyone seeking assistance and help with missing kids and elderly people, road traffic collisions, sudden deaths, the mentally-ill, crowd control at sporting events, etc., etc., etc., are going to have to get on with it on their own or (more likely) pay a private sector company a sizeable fee to come and assist instead.
Police Chiefs are not elected in the USA. Typically, they are appointed by the City Mayor as they are in Memphis, Tennessee, USA. The County Sheriff's are elected officials. Whoever wrote this story really screwed up a lot of details on Blue CRUSH and SPSS. Blue CRUSH is a methodology that was created by the University of Memphis and the Memphis Police Department.
It still does Holmes II and last Holmes III is sort of bobbing around procurement.
IBM has a tool called Initiate which has been trialled in a number of Police forces and national systems, but wider adoption of any tool is hindered by a confusing approach to IT across the forces and home office, and indeed they way it's all funded. Big systems like the PND have huge scope for development, but it's more likely that new systems will be developed at twice the cost than to invest in someone else's legacy project, as was noted for GMP. PND expanded to Police funding is now very tight and big sales are unlikely in the short term unless they can be justified as Spend to Save, but IT is to most CC's a low priority compared to officer numbers.
I2 is used by practically every force in the UK, mainly for social network analysis and court presentation.
Data-savvy analysts on the Force need only write some decent query templates to display predictions, and set up some reasonably-easy-to-reverse-engineer documentation. Make that part of the job description of a core team of 20 at the national level, with distributed counterparts at various city or county levels, so that no one in those billets or roles gets promoted laterally, vertically, or allowed to transfer unless the tool works efficiently without the users having to ask IT for help. And, without the remaining team members being at the mercy of vendors constantly upselling unncesssary customizations or fiendishly withholding no-brainer features.
Lotus Approach, as a WYSIWYG tool can be a decent prototype, but any decent database that does not demand anyone to be a programmer can be a front end. Cops could get 5 days of initial exposure, then every two months get refresher training. No need to leave ALL the bits and bobs of data mining to just the annointed, just deprive unwarranted access to real intel or sealed records. Then, allow cops to run reports in their patrol vehicles, as many in the USA seem to be able to do, on laptops.
If Lotus Approach had some serious updates to bring it into year 2013, I myself, as a non programmer could do spatial and temporal analysis templates. But, any real devs should be able to do PSPP and save their taxpayers boatloads of money.
"Fellows said it had reduced crime by 30 per cent by predicting where a crime would happen."
This is, of course, impossible. What they actually did was put more plods on the street in areas where there was a high level of crime. They also did the 'community liaison and outreach ' thing in a sensible manner. Apparently, the marvellous computerised system was able to predict that areas with a historically high level of crime would have lots of crimes committed in the near future. It sounds like old fashioned and sensible policing to me. It's a pity they'd lost those skills and needed computers and consultants to teach them about it.
As long as the predictive systems can accurately predict where the crime will move to when the police numbers are increased in a crime hot spot. Also, it seems IBM assumes there are sufficient numbers of the right kind of criminologists to effectively deal with the data they provide.
None of this is sounding terribly joined up.
Close but no cigar.
Stop. Think for a while. Stop again. Think again for a while. Then consider the following:
Do you really want to give someone who does not understand "Correlation does not equal Causation" as well as the basic difference between cause and effect a toolkit which is being advertised as "Crime prediction". The mere thought of what the most likely result is going to be makes me shudder.
I have worked in this particular environment for the last 30 years, police IT (or any other product) procurement has always been based on 'what's the cheapest' not 'what's the best'. I regularly have to use my own computer to get anything done (and yes I am well aware of the security issues, so no sensitive stuff).
Jeremy Clarkson keeps having a pop at our police vehicles which are procured in the same way. I hang my head in shame that we actually bought the final production run of Austin Allegro's because of this cheapskate mentality.
Having witnessed up close and personal how not to run the main Olympics operations room this summer, we are all doomed. After all, there was only 7 years to prepare for this. IBM can whinge all they like but they, or any other company hoping to drag 'Plod' into the 21st Century with a proper IT procurement policy has two hopes, and one of them is Bob!
AC (a current 'plod')
Forget buying software - the best investment would be in people who can ask the right analytical questions. After that go out and use R, or one of the other open source data-mining tools, or go to one of the excellent specialised UK companies that produce stuff that is fit for purpose.
Yes, you are right, investing in people who know how to ask the right questions. However you might be under the mistaken impression that Police forces can do this, some government Agencies do this, but mostly they prefer feeling collars. Using open source is however more difficult, it requires a level of expertise and commitment that just does not exist in this market. Some of the background papers issued by the home office indicate that forces should use specialist companies to do this, rather than waste warranted officer time.
This is where the problem lies, budget cuts mean that to spend money on IT and data analysis, then you have to invest in Analysts, which means conversion of warranted officers into analysts, you can't make them redundant and employ cheaper "civilian" staff. Trouble is you have to have the right type of person to do this, the IT industry is packed with them, which makes you wonder why they think they would be cheap. Recruit a bunch of graduates at 24K a year, train them in evidential data management, no mean feat, and then they become just as expensive as they become more experience. The social analysis skills required to do this exist already in Experian and alike, and they are sniffing around the Police as well to take on this market. Nothing open source about Experian.
that you are cynical about election of police commissars. Your nation's existence has, within living memory, been threatened by a society whose law-enforcement functionnaries were blatant political appointments from amongst elected officials. These officials proceeded to so influence their police forces as to render nugatory real commitment to the common good in favour of the dictated dominant political paradigm. Fuck Godwin.
I presume the software is not very good if they are having to approach people that so far up the food chain that they would have no experience in its day to day running. If there is something I have learnt as a developer, push back when the directors decide on a technology without consulting the people that use it, and run a mile if that technology is made by IBM. I still shudder to this day when someone mentions they use websphere.
That way, they could have know in advance what was the chance of recovering the billion spent on it.
I bet that the predictions software predicted success. Now, if only the prediction software could predict the election outcome they would know which politicians to convince instead of wasting their efforts on everyone.