Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We Found Insurance Fraud in Our Crash Data

levs.fyi

Post reply on HN
Print view View original

41–50 of 77 posts

Re: We Found Insurance Fraud in Our Crash Data

#41

Post by chickenzzzzu »

chias wrote: post #39
chickenzzzzu wrote: post #35

Earlier quoted context omitted.

Now it's time for the devil's advocate-- if the title was "some people get in a lot of accidents", would the top comment be "actually it's more likely that insurance fraud is the answer"

For what it's worth, the headline here does NOT say "we find it more likely than not that there is evidence of insurance fraud in our crash data".

i have absolutely no problem with that, i suspect that no matter the title, whoever has a contrarian opinion to it will be the one to speak up while the rest remain silent

Re: We Found Insurance Fraud in Our Crash Data

#42

Post by macintux »

gnfargbl wrote: post #25

I don't see nearly enough complexity in this analysis to justify the claim of having found any insurance fraud. Firstly, there's no account for correlation between the features identified. The article mentions VINs which have several single-vehicle accidents, for example, but someone who has one single-vehicle accident is probably more likely to have another. Switching coverage is another of those potentially-correla…

> The article mentions VINs which have several single-vehicle accidents, for example, but someone who has one single-vehicle accident is probably more likely to have another.

I haven't followed this story to know whether she's still driving, but one driver was involved in 7 crashes in 4 years, including two fatalities.

https://www.indystar.com/story/news/crime/2022/06/17/car-cra...

Re: We Found Insurance Fraud in Our Crash Data

#44

Post by hn_throwaway_99 »

chickenzzzzu wrote: post #35
gnfargbl wrote: post #25

I don't see nearly enough complexity in this analysis to justify the claim of having found any insurance fraud. Firstly, there's no account for correlation between the features identified. The article mentions VINs which have several single-vehicle accidents, for example, but someone who has one single-vehicle accident is probably more likely to have another. Switching coverage is another of those potentially-correla…

Now it's time for the devil's advocate-- if the title was "some people get in a lot of accidents", would the top comment be "actually it's more likely that insurance fraud is the answer"

That doesn't really matter. I mean, sure, I could look at this data and say "hmm, looks like insurance fraud to me."

But the point being, I really don't have evidence one way or the other. Implying insurance fraud is saying you think a criminal act occurred, and when you say something like that the burden should be on you to provide some evidence to that effect.

Random internet commenters bullshit all the time. But this blog post provides specific data and an unambiguous conclusion that fraud occurred (the title does not hedge, it is "We Found Insurance Fraud in Our Crash Data"), so if you do something like that you need to provide evidence for it. I appreciate the parent commenter for pointing out that, despite all of their data, they didn't really provide much evidence for their conclusion.

Re: We Found Insurance Fraud in Our Crash Data

#45

Post by KingMob »

So many alternative explanations unmentioned:

1. Drink/drugs impair your judgment, making you more likely to commit fraud if you were on the fence, or the impulsive sort.

2. Drunk drivers get into multiple accidents, but some are able to keep evidence of impairment off the record (buddy on the force or the insurance co, etc.)

3. Drivers doing drugs that aren't detectable on the spot, and clear the body quickly, reducing evidence of drug use in crashes.

4. More people are likely to do drink/drugs at night, so if 1-3 are true, we should expect to see more crashes at night. Likewise for a lack of witnesses (fewer passengers or bystanders).

Most importantly, statistics are completely absent from the article. Does OP have any evidence these numbers wouldn't fit the correct random distribution? I know the point of fraud is to appear natural, but if done well, that means evidence has to be collected by other means, and this kind of analysis is only a weak signal.

Re: We Found Insurance Fraud in Our Crash Data

#47

Post by gnfargbl »

hn_throwaway_99 wrote: post #44
chickenzzzzu wrote: post #35

Earlier quoted context omitted.

Now it's time for the devil's advocate-- if the title was "some people get in a lot of accidents", would the top comment be "actually it's more likely that insurance fraud is the answer"

That doesn't really matter. I mean, sure, I could look at this data and say "hmm, looks like insurance fraud to me." But the point being, I really don't have evidence one way or the other. Implying insurance fraud is saying you think a criminal act occurred, and when you say something like that the burden should be on you to provide some evidence to that effect. Random internet commenters bullshit all the time. But t…

Exactly this. When I open a post like this I get excited because I think I'm either about to learn something interesting about data analysis, or I'm about to see a neat application of a technique that I already know about and it's going to make me happy.

When the post actually boils down to "hey if you look at a large dataset then you'll find unusual events in it, now visit our AI startup!" then that irritates me enough to whine about it in the comments.

Re: We Found Insurance Fraud in Our Crash Data

#48

Post by kstrauser »

camtarn wrote: post #17

It's an interesting article, but the title ('we found fraud!') doesn't quite match the article, which continually points out that there are suspicious patterns in the data but that they aren't ironclad evidence of fraud. That said ... yeah those particular VINs do look seriously dodgy to me! If nothing else, then the drivers reporting 5 crashes in a 12 month window probably shouldn't be driving.

I got into 2 crashes in a couple of weeks once:

I was in the left lane, stopped, signaling to make a left turn, when someone rear ended me because they didn't notice me.

I was stopped at a red light. The octogenarian behind me revved his engine, dropped into drive, and peeled out to rear end me a second time, because he “had places to be”. When I asked for his insurance, he drove off and I had to file a hit-and-run report with the police.

Just saying, sometimes you can be minding your own business and get really unlucky.

Re: We Found Insurance Fraud in Our Crash Data

#49

Post by KingMob »

chickenzzzzu wrote: post #35
gnfargbl wrote: post #25

I don't see nearly enough complexity in this analysis to justify the claim of having found any insurance fraud. Firstly, there's no account for correlation between the features identified. The article mentions VINs which have several single-vehicle accidents, for example, but someone who has one single-vehicle accident is probably more likely to have another. Switching coverage is another of those potentially-correla…

Now it's time for the devil's advocate-- if the title was "some people get in a lot of accidents", would the top comment be "actually it's more likely that insurance fraud is the answer"

No, because insurance fraud is vastly less likely.

It's the same reason I don't assume memory bits flipped by solar radiation are the most likely cause of my bugs.

These are not symmetrical claims.

Post reply on HN