Hacker News Reader

Comment by graeme | Hacker News Reader

Has anyone worked out what the base rate expectation of murder would be?

I think I read in the FT that Didi has booked a few billion rides in that period. How many people (and women) would normally be murdered in cabs?

The stories cited are obviously horrific and it sounds like Didi could have prevented them with a better system. But, I'd nonetheless be interested to know how safe this is to the alternative, and also how many murders we would expect in a few billion human interactions.

At that scale all our intuitions are off and we need to look at stats.

View on HN

I've personally had enough problems with taxi drivers (and I live in a developed country) that I will only take a taxi as a last resort. They overcharge, they demand cash because apparently their EFTPOS machine is broken, they drive you on the scenic route to increase the fare, or they'll try to set a fixed price when you know it's cheaper to run the meter, half the time the driver doesn't even know how to get there and I have to navigate them.

Just last week I got woken up at 3 AM by a distraught housemate being dropped home by the police because a cabbie sexually assaulted her. It's fucking disgusting what cabbies manage to get away with and then claim "Uber is bad because it's unregulated". Uber's self regulation has proven to be vastly superior to the taxi industry.

Meanwhile with Uber, my experience has been 90% positive. It's incredibly rare to have a driver try to rip me off and it's a lot safer in general. I'm absolutely sure that the cabbie who assaulted my friend has done it before to other people, and will do it again to some other poor drunk girl. I live in a Australia, this shouldn't be an issue we're facing, but it is.

I don't know what Didi is like compared to Uber, but I'm absolutely sure that it's better than cabs, especially in developing countries.

As I understand, people are outraged because Didi built the ability to "rate" passengers in terms of their looks, i.e. whereas riders would rate drivers in terms of "fast, professional, etc", drivers could share ratings of riders in terms of "young, sexy, cute, etc". This last murder also apparently happened a mere day after Didi's decision to re-enable drivers' abilities to see riders pictures (a feature which had been disabled as a response to the previous murder). The straw that broke the camel's back was the surfacing of another girl's report that she was harassed the day before by that same person and only got out safely because she dashed away when she got the chance (and her complaints were largely ignored by Didi).

Coupled with lackluster driver screening, and the non-responsiveness to complaints, this essentially created a platform for sexual predators to thrive. As my wife put it, the driver app was basically Tinder for perverts.

I think the argument that at a scale of billion of rides, assaults can fall between the cracks is downplaying the magnitude of the negligence. The police force can handle much higher rates of all kinds of emergency calls. Uber has literally tens of thousands of customer service reps and entires teams dedicated to harassment reports stationed across the globe. If there were 50 harrassment reports as the article claims, surely Didi could afford to have 3 people on rotation looking at a queue of reports, preemptively disabling accounts. There's absolutely no way Didi can make a believable excuse for keeping a criminal in the system after having received complaints about him being involved in a sexual assault.

Thanks for the clarification. I thought there might be more to the story than what I initially read.

That's pretty fucked up for Didi. Do you know if these "features" for the drivers were on versions of the app outside of China (i.e. in Australia)?

That’s an extremely freaky and unsettling design choice, to include such a feature in a system involving customer relations and client service.

It seems like the kind of idea you’d throw in, if you absolutely did not care about any amount of success.

Something like that is bound to blow up sooner or later, and destroy confidence in professionalism. It goes against an almost universal idea, cutting across all cultures, that customers shouldn’t be judged if you expect to do business with them.

Not to mention malfunctioning meters, especially around tourist spots when you don't speak the language.

With Uber, I at least know what I'm going to be charged

Yeah, that’s what companies do : look at the stats. And as long as stats say it’s rare enough, it’s not that bad.

People are boycotting not because there’s a higher chance of being murdered/kidnapped with Didi, but that the company was unresponsive, did not have proper procedure/policy for vetting drivers, and took way too long to get even basic info about the suspect driver.

What the stats say here, higher or lower, is moot. Because the issue is that Didi does not give a shit about passenger safety.

> Because the issue is that Didi does not give a shit about passenger safety.

Didi intentionally setup the system this way to promote its rideshare as a hook up service for its drivers. When describing the service, its top management made it clear that such a service in which drivers and passengers are locked up in a moving car provides a "sexy scenario" and thus Didi must take the opportunity and try its best to promote such hook up experience.

It is not like Didi doesn't care much about passenger safety, Didi actively puts passenger into danger for its own financial gains and knowingly denied police intervention when its passenger was in danger.

can you link a write-up about this?

if you google for the terms "滴滴 sexy" in which "滴滴" is the Chinese name of Didi, you get tons of reports.

That is insane. They deserve to go out of business.

I can't imagine that Lyft or Uber could have done something like this. But maybe I'm just too naive.

I didn't speak clearly enough perhaps. It sounds like Didi could have prevented these murders. So, they should have, both morally and from a self interest perspective.

But when we look from outside in at the significance of this, it's nonetheless worth thinking of it in the context of its scale. That's not to justify anything, but merely to understand.

I once trying to dig out the crime rate (not murder rate) of our city in South China, at least trying to have a grasp on what's going on. (TL;DR: and failed)

The only related data that I (as a citizen, not insider) can found is in the annual report published by the local court, which basically just a statement on how many criminals the court have prosecuted, only a number, no detailed categorization.

I also searched for the "population" of the local prison, the only data I can found is inside a news article about prison location transfer published on the Ministry of Justice's website few years ago, which stated "All XXX prisoners are successfully transferred".

None of the data is reliable enough to calculate the real crime rate of our city, as the local and prison may also handle cases from other cities, vice versa.

Because of all that, the value of my little "research" is far less than having a small chat with a friend of mine who work for the local police department.

Maybe this also the reason of why rumours in China got it's charm.

This, and before that taxi driver murderers were also heard of often in China (though I’m not sure if they were more likely to be murderers than average).

Also, Chinese police are overworked (China is extremely under policed ironically enough) and so tend to focus on seemingly easy but impractical solutions to fighting crime.

There’s often political pressures on police departments to fix unpopular statistics, and not the actual issue.

Most famously this leads Japanese departments to label difficult murder cases as suicides.

Excellent point. This is a prime example of how government intervention mucks things up by setting up relatively benign goals, but in practice leads to an outcome that is unchanged, or perhaps even worse. What if all Japanese citizens were allowed to carry guns out in the open? An armed society is a polite society.

> An armed society is a polite society.

Perhaps suggesting that the Japanese need more guns to be polite wasn't the best example. Japan is already one of the most (the most?) polite society in the world - at least talking about how strangers interact.

It’s less about governments, and more about how the law of unintended consequences applies in all human matters.

Example: anthropologists believe that on the whole, more polite societies typically murder each other at a higher rate than less polite societies.

That's an amazing stat - do you have a source for that?

Here’s a forum thread with an overview and some links to primary sources: https://www.debatepolitics.com/philosophical-discussions/221...

amazing, thank you.

> An armed society is a polite society.

*An armed society of rational people is a polite society. And people tend to slip in and out of a rational state depending on their emotions, pride, sobriety, etc.

An armed society is a polite society

"Polite" just means "in keeping with cultural norms". If it's polite to deal with disagreements, injuries and insults by pulling a gun, that's still polite but I certainly wouldn't feel safe.

You can legally own an artillery piece in Pakistan, but I can't vouch for manners of locals

> An armed society is a polite society.

If you believe that, you've never lived in Atlanta (or Japan for that matter).

Do you have data supporting data for your claim? I'm a Chinese and this kind of murders are really rare. The last one I heard in my hometown (a small costal city) is over 10 years ago.

And with taxi-hailing apps' commence, this is becoming really rare since the driver can easily pin-pointed (by cross-referencing ubiquitous street cams). Actually in this case the murderer got arrested soon afterwards.

While I agree China is under policed, they are generally more effective in body offenses and thanks to street cams (again). These years even home break-ins are rare. As who lived in SF for a year, it's definitely better than SFPD.

I don't think the point was that cab murders were high, just that both were low, so the whole freak out is probably not rational.

The "heard of often" just means it gets plastered on the news / social media when it does happen.

> not rational

If it could have been prevented, it is the right reaction.

Besides, this is how China works most of the time. Pressure (boycot) on companies leads to results... or put them out of (nearly) business.

I believe that was my original point: Chinese officials will rely on didi to police it’s drivers, since having the police instead do that isn’t practical. But it winds up being worse than that, since if didi wasn’t around we would all be taking anonymous black cabs again.

What is wrong with the usual cabs? Sure, slightly more expensive but reliable.

No, it’s the opposite actually. Cheaper, less reliable, especially the black cabs.

In Beijing it wasn’t uncommon to hear of a PKU student every few years being murdered by a taxi driver (especially black taxis), once they went full in on satellite campuses causing lots of transit headaches. And like I said originally, I don’t know if it was higher than average or just well publicized when it happens.

Ride sharing and taxi hailing apps have definitely made taking a taxi mug safer. Conserving how didi has replace black cabs, that is definitely a good thing. I’m not sure why the government wants didi to shutdown its ride sharing business rather than rely on it as a sure fire way of at least solving murders after the fact. Even full background checks won’t solve murders 100%, and knowing you’ll get caught is a great deterrent.

Property crimes are still a huge problem in China, there is a reason all 2nd tier apartments are still barred like crazy. Chinese police don’t really care, they are not using the CCTVs to actually fight property crime, rather for other things.

> I’m not sure why the government wants didi to shutdown

You finally mention it. Without the government wanting it, this story wouldn't have broken. It seems like the c. gov could easily hide hundreds of murders.

China has one of the lowest intentional homicide rates in the world.

I’m actually writing this from Didi at the moment. It’s quite irrational to overhype these didi-murders since China is a country of about 1.4 billion people and statistically these incidents are pretty much meaningless. I guess it’s same kind of irrationality like in the west, when people try to come up with problems about Uber that have always existed in other contextes.

well done on defending the company with zero social responsibility in mind. the problem is never about the murder itself, it is all about the fact that didi worked extremely hard to intentionally prevent any help that can be offered to save the victim -

1. another customer reported the murder a day before for harassment, Didi refused to look into the case

2. when the victim's family & friends reach out to Didi after receiving messages from the victim asking for help, Didi refused to do anything other than asking them to wait.

3. when the victim's family & friends reported the incident to the police, the police immediately contacted Didi but again Didi refused to provide any help.

Reply on news.ycombinator.com