Insurance Rant

[SteveW]

Well its that time of year again that we all love.

 

Normally it doesn't bother me but this year I am steaming

 

Basically last year someone drove into the back of me and caused about £300 of damage.

She paid me the money to get the car fixed, job done no problem. Well so I thought..............

 

One of the conditions of my policy was you had to inform the insurance company within 24 hours of

any kind of incident or accident even if not at fault. No problem so I did.

 

Well now all my quotes are £400 - 500 pound higher if I declare this incident on this years policy

That is just an absolute joke!!! If I take the incident off I get a reasonible quote put the non-fault incident on, and boom jumps by £400-£500.

 

Wish I had kept my mouth shut and not informed the insurance company!

 

Bloody scadalous!

[Chap]

Simple solution - seek additional compensation from the party that had liability for the accident to the tune of additional insurance premiums over 3 years from the incident (think that is how long you need to declare them for).

[marzman]

You've got another car havent you bud? Signup to Admiral Multicar - its sooo much cheaper.

 

Even if your other cars' insurance doesnt run out for another 11 months, as long as you add it to the policy at some point you will get a 35% discount. My zed is insured for £735 with them (and it was around £1300 with everyone else).

[nixy]

Stinks doesn't it - they screw you whatever you do! We actually pay insurance so that costs are covered if we have an accident but then we do all we can to avoid making a claim and taking the money we are owed in case our premium takes a massive hike. Then like you, even when someone else is at fault and pays out YOUR premium goes up anyway! Might as well just claim for everything!

[HassanZ]

You've got another car havent you bud? Signup to Admiral Multicar - its sooo much cheaper.

 

Even if your other cars' insurance doesnt run out for another 11 months, as long as you add it to the policy at some point you will get a 35% discount. My zed is insured for £735 with them (and it was around £1300 with everyone else).

 

+1 on the Admiral multi-car.

 

When I was fiddling around with Admiral's quote system, I noticed the premium quoted didn't fluctuate when I added a non-fault claim on there. I tried it out of curiosity. Not sure why it's affecting you? Who are you with?

[nixy]

I tried Admiral multicar when i got a second car but it was no better for me.

[coldel]

I for one am struggling with this insurance hike - you drive safely, you don't hit anything, you don't claim on your insurance over the insured period then they put it up. That is irrational and downright ridiculous.

[nixy]

Sounds to me as if they said 'have you had any accidents in the last * years whether at fault or not' and the answer was yes - hence the hike. Maybe the answer should have been no! But if the other person claimed on THEIR insurance the details are probably all on a nice little database somewhere...........

[Ekona]

Here's the issue as I see it:

 

The reason premiums go up if you've had a non-fault claim is that, statistically speaking, you're more likely to go on and have a claim where it is your fault. As such, there's no point getting mad at the insurers themselves as they're only covering their arses and going on what the figures say, you really need to direct your anger at all the idiots who do go on and have that at-fault claim after a non-fault one as it's them that's causing the price hikes!

[nixy]

Not to mention all the people with £4k worth of imaginary whiplash!

[Sam Mcgoo]

Am I the only one who's insurance has gone down this year? More power, more mods and still gone down (only a little).

 

If your over 25 I would recommend trying Livingston Warman as advertised on here, they get a very big from me.

[synystermike]

yep its happened to me mate, some idiot on the phone wrote my last zed off in jan, now my insurance is 400 extra. did you claim against the person who hit you, cos you can claim it back. if you didnt move insurance company and dont tell them, or does it get regestered so they all know

[coldel]

Here's the issue as I see it:

 

The reason premiums go up if you've had a non-fault claim is that, statistically speaking, you're more likely to go on and have a claim where it is your fault. As such, there's no point getting mad at the insurers themselves as they're only covering their arses and going on what the figures say, you really need to direct your anger at all the idiots who do go on and have that at-fault claim after a non-fault one as it's them that's causing the price hikes!

 

Why would it be more likely? One act is surely independant of the other therfore not a valid driver (scuse the pun) for increasing insurance premium charges. If you are young, male and new to the road then these are clear drivers of the likelihood of an accident - being hit by someone else I don't think is, causality vs coincidence. I think it should be made law that consumers can see these all conquering databases, they can be made to say what you want them to say, I am sure there are attribute sets in there which are hidden which if were included would have an reducing impact (twice, scuse the pun).

[Ekona]

Why would it be more likely? One act is surely independant of the other therfore not a valid driver (scuse the pun) for increasing insurance premium charges. If you are young, male and new to the road then these are clear drivers of the likelihood of an accident - being hit by someone else I don't think is, causality vs coincidence. I think it should be made law that consumers can see these all conquering databases, they can be made to say what you want them to say, I am sure there are attribute sets in there which are hidden which if were included would have an reducing impact (twice, scuse the pun).

It's more likely because that's what the statistics say. I've no idea of the real logic or reason behind it, but them's the facts and that's life. Maybe it's because people who have a non-fault then drive their car before it gets fixed in a slightly more care-free manner, leading to them causing an accident? I dunno, just a guess.

[DannyBoy]

Here's the issue as I see it:

 

The reason premiums go up if you've had a non-fault claim is that, statistically speaking, you're more likely to go on and have a claim where it is your fault. As such, there's no point getting mad at the insurers themselves as they're only covering their arses and going on what the figures say, you really need to direct your anger at all the idiots who do go on and have that at-fault claim after a non-fault one as it's them that's causing the price hikes!

 

Why would it be more likely? One act is surely independant of the other therfore not a valid driver (scuse the pun) for increasing insurance premium charges. If you are young, male and new to the road then these are clear drivers of the likelihood of an accident - being hit by someone else I don't think is, causality vs coincidence. I think it should be made law that consumers can see these all conquering databases, they can be made to say what you want them to say, I am sure there are attribute sets in there which are hidden which if were included would have an reducing impact (twice, scuse the pun).

 

I know that's the reason they give Ekona, but i'd bet that the statistics only show what they show because certain areas are very accident prone and people who drive a lot of miles in these areas are likely to be involved in more incidents. So while this will show that people who get crashed into will also crash into other people, the causeof that is actually the area and annual mileage, which are already accounted for in the premium. So basically they're double weighting the policy for the same thing.

 

I will lick the inside of my exhaust pipe if anyone can come up with stats which show that given two identical drivers, in the same area, car, profession, and driving the same annual mileage the one who got crashed into is more likely to then have a crash based solely on that fact.

 

DB

[coldel]

Why would it be more likely? One act is surely independant of the other therfore not a valid driver (scuse the pun) for increasing insurance premium charges. If you are young, male and new to the road then these are clear drivers of the likelihood of an accident - being hit by someone else I don't think is, causality vs coincidence. I think it should be made law that consumers can see these all conquering databases, they can be made to say what you want them to say, I am sure there are attribute sets in there which are hidden which if were included would have an reducing impact (twice, scuse the pun).

It's more likely because that's what the statistics say. I've no idea of the real logic or reason behind it, but them's the facts and that's life. Maybe it's because people who have a non-fault then drive their car before it gets fixed in a slightly more care-free manner, leading to them causing an accident? I dunno, just a guess.

 

My point I was trying to make is that any statistic will not indicate causality. Having done a stats degree and spent a decade working with databases in my job I am continuely shocked at the level, or lack of it, of proper analysis of it. I wonder how many of these insurance companies have done the numbers properly, do they for instance use multidimensinal driver models maybe even neural networks to properly evaulate 'what makes things happen' or do they get lists of numbers and literally get some 'stats' from them by looking at averages/correlations/standard deviations which for the purposes of creating an algorithm/model for individuals insurance premiums is literally useless? I would wager a full british pound its the latter.

 

...as for stats - be careful to quote so easily.

[HassanZ]

Am I the only one who's insurance has gone down this year? More power, more mods and still gone down (only a little).

 

If your over 25 I would recommend trying Livingston Warman as advertised on here, they get a very big from me.

 

Hopefully mine will drop too at renewal time as I'll be out of my first year of Zed ownership. The first years tends to be higher due to the 'lack of experience' driving your new car.

 

On my BMW I was paying around £700 (owned it for 2 years) and it jumped to around £900 when I bought the Zed. If the ownership time period was the same on both cars, the insurance quotes wouldn't be much different.

 

I'm 23 at the moment and turning 24 this month, as soon as I hit 25 let's hope it drops further

[coldel]

Here's the issue as I see it:

 

The reason premiums go up if you've had a non-fault claim is that, statistically speaking, you're more likely to go on and have a claim where it is your fault. As such, there's no point getting mad at the insurers themselves as they're only covering their arses and going on what the figures say, you really need to direct your anger at all the idiots who do go on and have that at-fault claim after a non-fault one as it's them that's causing the price hikes!

 

Why would it be more likely? One act is surely independant of the other therfore not a valid driver (scuse the pun) for increasing insurance premium charges. If you are young, male and new to the road then these are clear drivers of the likelihood of an accident - being hit by someone else I don't think is, causality vs coincidence. I think it should be made law that consumers can see these all conquering databases, they can be made to say what you want them to say, I am sure there are attribute sets in there which are hidden which if were included would have an reducing impact (twice, scuse the pun).

 

I know that's the reason they give Ekona, but i'd bet that the statistics only show what they show because certain areas are very accident prone and people who drive a lot of miles in these areas are likely to be involved in more incidents. So while this will show that people who get crashed into will also crash into other people, the causeof that is actually the area and annual mileage, which are already accounted for in the premium. So basically they're double weighting the policy for the same thing.

 

I will lick the inside of my exhaust pipe if anyone can come up with stats which show that given two identical drivers, in the same area, car, profession, and driving the same annual mileage the one who got crashed into is more likely to then have a crash based solely on that fact.

 

DB

 

+1

 

It would be good to hear a view on the level of sophistication from the Insurers on here. Although I suspect they all come from a central database and they are literally drip fed the numbers and have little insight into the analysis?

 

I do have this awful feeling they are just using out of date 'stats' that thirty years ago Mr Dave Admiral sat down with a pen and back of fag packet and put together...

[martinmac]

I was in a car as a passenger a couple of weeks ago. We were hit from behind abd thankfully the driver saw the other car charging up behind him and booted it onto the grass verge minimising the impact.

 

Very light damage to the back of his 5 day old Audi. He called his insurance company and when he mentioned rear impact the lady said " oh right, so you will be claiming for personal injury then. This is why insurance is going up. Until the law changes its going to continue.

[coldel]

I thought that I read somewhere there are already moves in place to cut back on personal injury claims and in particular whiplash claims.

[Ekona]

I will lick the inside of my exhaust pipe if anyone can come up with stats which show that given two identical drivers, in the same area, car, profession, and driving the same annual mileage the one who got crashed into is more likely to then have a crash based solely on that fact.

But by the same token you could do exactly the same with any two people and end up with the same result. There is no reason that completely separate people should have the same experiences as each other, but when you start including millions of people in that often a pattern will emerge. I'm no expert on statistics by any means and I can't get my head round the reason that any two people out of fifty share the same birthday or whatever silly low figure that particular one is, but companies don't tend to raise their figures for no reason at all. There has to be a business case for it, and in this situation they will have the stats to back up their need to increase their premiums to cover their own risks.

[nixy]

Guess that is what an actuary is for - what an interesting job that must be!

[Chris`I]

I wonder how many of these insurance companies have done the numbers properly, do they for instance use multidimensinal driver models maybe even neural networks to properly evaulate 'what makes things happen' or do they get lists of numbers and literally get some 'stats' from them by looking at averages/correlations/standard deviations which for the purposes of creating an algorithm/model for individuals insurance premiums is literally useless? I would wager a full british pound its the latter.

I would bet on the latter too. As you know working in stats, its costs a hell of a lot of money to come up with working logic for this stuff. To build and train a nerual network would cost far more than they would get back from having accurate risk models. And they would need a lot of good clean data to drive it and allow it to learn and adapt. Its not even just the pure number of claims they'd need to run the model, its the breadth of information about each claim too. It makes far more sense to have a crude and rudamentry system that keeps their quote about the same as the competition based on a few basic facts and have a bit of wiggle room that should someone want to be price matched, they can easily do it.

[coldel]

I wonder how many of these insurance companies have done the numbers properly, do they for instance use multidimensinal driver models maybe even neural networks to properly evaulate 'what makes things happen' or do they get lists of numbers and literally get some 'stats' from them by looking at averages/correlations/standard deviations which for the purposes of creating an algorithm/model for individuals insurance premiums is literally useless? I would wager a full british pound its the latter.

I would bet on the latter too. As you know working in stats, its costs a hell of a lot of money to come up with working logic for this stuff. To build and train a nerual network would cost far more than they would get back from having accurate risk models. And they would need a lot of good clean data to drive it and allow it to learn and adapt. Its not even just the pure number of claims they'd need to run the model, its the breadth of information about each claim too. It makes far more sense to have a crude and rudamentry system that keeps their quote about the same as the competition based on a few basic facts and have a bit of wiggle room that should someone want to be price matched, they can easily do it.

 

In other words, laziness £20k on a Baysia licence, two weeks time with a decent marketing scientist with access to their database and they would have a much better model...although one which probably means debunking a lot of the nonsense we see on 'important information' sections of insurance premiums nowadays.

[Chris`I]

In other words, laziness £20k on a Baysia licence, two weeks time with a decent marketing scientist with access to their database and they would have a much better model...although one which probably means debunking a lot of the nonsense we see on 'important information' sections of insurance premiums nowadays.

Just Googled Baysia and cant find anything - what is it? I'm assuming its some kind of nerual net program? Having worked in business intelligence for a few years at very large data producers (telco's, news providers, payment processors) that seems far too easy. If that can handle the millions of claims a year that an insurer sees, then you need to get on and sell it to them. The ROI to them would be massive. Dont underestimate the poor state that their database will be in. You'd need to do a hell of a lot of data cleansing, conformation and aggregation before you can even think of putting it into a neural net or even a plain old data miner.

 

I do agree they should put the effort in though