Forums Forums National News Rising crime?

Just to say these were stolen from Ben Nevis Road Tranmere so anybody nearby check gardens or bushes, just might have been chucked..

Before 1995 in Merseyside:-

How much gun crime do you remember?
How many times did you see motorbikes doing wheelies?
How many cars did you see drifting?
How many parked cars were damaged other than keying?
How many things were stolen from your garden?
How many times were tyres slashed?
How many times were you threatened by someone you didn't know?
How many women did you see getting beaten in the street?
How many fights of more than two people did you see in the streets?
How many people walked past you obviously smoking cannabis in public?
How many times did the police not respond to incidents?
How many pubs had bouncers?

Apart from witnessing gun crime first hand (though I know people that have been on the receiving end), all the others I have seen in the last few years, many more than once. These were extremely rare in the 70's and 80's.

I have CCTV and a dashcam not because of my perception but of actual incidents that left me little choice. I've contacted the police more times in the last 10 years than in the previous 40.

I worked in Merseyside (Liverpool and Wirral) clubs in 70's and 80's, not once did I feel unsafe nor the need to be acutely aware of what was going on around me, odd fights broke out and they were easy to break up without anybody having much more than their pride damaged. These days just driving past the clubs worries me from the amount and type of violence I see outside them, seeing A&E on club nights is like the aftermath of mass Roman gladiator fights.

Its not necessarily just the quantity of crime, its the way the crime is being carried out.

School holidays are coming up, Kids will be getting to bored.

More vandalism and theft.

I bet if it happened to them it would be a different story Cools. ... like that does not like the taste of there own medicine.

Spot on DD.

It depends how what we consider ' being safer now ' means. Maybe for a particular individual going for a walk at night, (although if someone had been mugged, it's not something to be chanced again. That usually stays with one for life) but what about the youngsters being groomed on line, and all the new types of crime ? People trafficking, sex slavery, child abuse, these crimes are the rising, but they are hidden. I very much doubt kids are safer, they can't even go to the corner shop unaccompanied today.

I definitely think crimes on the increase, nobody seems scared of the consequences if cought. Yes and more violent, that case last night upset me, that vile brute of a man beating a five year old little boy to death because he lost a shoe!! Big Big man! I hope he gets all that he deserves in prison.

Electric chair would be better Cools.
He wont be so tough if it came to the death sentence.

Be quite interesting offering perpetrators the choice of the death penalty or an eye for an eye punishment and see them squirm as they choose . (Feeling sadistic towards monsters tonight ! )

"If you have a group of 10 people and only 2 witnessed an event(20%), if you interviewed 3 people(30%) out of the 10, you would be unlikely to interview someone that witnessed the event.

If you multiply all those numbers by a million or even a billion, the same applies despite an unusually large sample size of 30%..."

This is absolutely not the case.

The actual figures with a population of 10 with a sample size of 3 is that you have a 0.47 probability of missing all three witnesses and and a 0.53 probability of interviewing at least one. A huge margin of error. You might as well flip a coin!

Now consider a population of a hundred. With a sample size of 10 (a third of they taken for the population of 10) the probability of of missing all 20 witnesses is under 1%. The probability of interviewing at least one is 99%.

I hope you now appreciate that population size DOES matter, and you need a much smaller sample size to get an accurate picture from a large number than from a small one.

You can find confirmation from any school statistics textbook.

38,000 is a HUGE sample size for a simple question like "Have you suffered a crime in the past year" - far bigger than is needed. The reason it IS as large as that is because multiple questions are asked and the probability of suffering some crimes is tiny.

Rest assured. the CSEW have clearly understood their mathematics. Their result is credible. It would be good to say the same about the recorded crime figures, but as no data on their methodology has been published the best I (or you or anyone else) can say is that we simply don't know how accurate they are.

I suspect the lack of detail in the police figures is an advantage to mathematically naive journalists. The CSEW report requires a certain knowledge of statistics to understand and I doubt the average journo has it.

A great pity, because the message "Crime is increasing" frightens people to the extent that some are afraid to leave their houses at all, let alone after dark.

Rubbish. The hospital figures have fallen as the CSEW figures have. Here is the evidence:

https://www.aol.co.uk/news/2017/04/...eated-at-aande-for-violence-related-inj/

Admin Note : Personal Insults have been reported and removed

I have also seen massive queues during the daytime with ambulance staff and their stretchers filling the corridors.

Your own link supports what I have just said:- "but casualties peaked at weekends amid spikes in alcohol-fuelled incidents"

Sorry, yes, I got them back to front 0.47 vs 0.53 but it still shows that with a large sample size of 30% its still roughly 50/50 whether you detect any events.

I will concede to your refresher on sample/population size.

But, just to show its not always clear cut, using my example and multiplying by ten.

If you have a group of 100 people and only 20 witnessed an event(20%), if you interviewed 30 people(30%) out of the 100.

You end up with an decreased probability of detecting an event (0.492) ... ie an increase in not detecting an event

I assume that anomaly is something to do with the sample proportion being greater than the event probability???

Or have I got it back to front again?

Don't forget all the crime that goes on that never get reported,from simple things like dropping litter, drinking booze on the street, not picking up your dog poo,scamber bikes up and down the road, at least three or four a day by me, but i put the blame on the parents, most of the kids that do crime are from broken families, most don't even know who their father is or when they last saw them, most if not all are out of work and that includes the parent [ s]

Wrong.

The way to do it is to calculate the probability of missing a witness after 30 attempts and subtract this from one. In other words, you multiply 80/100 x 79/99 x 78/98 x.... to thirty terms.

I make this about 0.0003, so the probability of NOT missing it completely is 1 - 0.0003 or 0.9997.

(I just noticed I slipped a decimal point earlier. If you interview only 10, the probability of missing any witness is a bit less than 0.1 and the probability of interviewing a witness is a bit more than 0.9. Apologies.)

With very large populations - like a million of which 200,000 witnessed the event - you can save yourself a lot of work by not bothering to reduce numerator and denominator by one for each term, and simply raise 0.8 to the power of your sample size. Thus interviewing 30 from a population would give you a 99.9% chance of interviewing a witness.

Doh, I think I left the sample size at 3 instead of 30. I had been using a high precision calculator because I had also been playing with 800!, unfortunately it won't got to 38,000! let alone UK population.

(80!/49!)/(100!/69!) = 0.00022 (2sd) so pretty good with your estimate.