Hacker News Reader

Comment by chrissnell | Hacker News Reader

I live in small-town Kansas but commute to San Francisco occasionally for my job, which is based there. I just got back from a trip last week and noticed something startling about the Uber drivers that ferried me around: never before have I seen drivers this bad. Twice in one week, my drivers nearly ran over a pedestrian. On one of those incidents, a guy was walking across a crosswalk in North Beach in broad daylight. Everybody in this Uber Pool was screaming "stop!" as the driver ran right up on this poor guy but the driver seemed oblivious, like the guy was invisible. The only explanation, I think, is that our driver was extremely fatigued.

I predict that rideshare-pedestrian deaths are going to skyrocket in 2017 as more drivers take desperate measures to try to get by.

View on HN

This is why I use Lyft over Uber. It's anecdotal (though remarkably consistent, even accounting for confirmation bias), but when I get in an Uber — which isn't often; I don't even have the app installed on my device, so I'm only getting in one if I'm with someone else who summoned one — I can expect assholery in traffic. And I just don't see that with Lyft drivers, even when I'm riding with a driver who takes fares from both.

It's like there's some sort of un- (or under-) stated thing amongst the Uber drivers, such that fare turnover beats traffic safety.

I don't think that Uber is any better than Lyft. So many of these drivers work for both. I have no data to back this up but I believe the cause is shared rides such as Uber Pool and Lyft Line. The low-revenue/high-volume nature seems to encourage these drivers to drive as fast and as long as possible. I don't remember seeing bad driving so much, back before these shared services.

That's probably a relevant variable I don't have a lot of exposure to: I tend not to use the pool/line fares often.

Low-revenue/high-volume isn't a proper characterization of Lyft Line. The driver gets paid the rate for the number of miles driven / minutes elapsed. At least in LA the distance fare is comparable, the time fare is 2/3 the normal rate. But there is a balance, Line rides either end up being the minimum fare, so there's no difference for the driver, or they end up being longer, converting idle time into paying time. Also Lyft counts each pickup as a ride in your metrics, which counts toward driver bonuses.

Also I am very mindful that any report or flagging by riders concerning safety can result in removal from the platform. Other drivers might go balls out but they won't be on the platform for long.

There are some downsides for the rider. Be in the car within 105 seconds of when the driver taps arrive or pay $5. You can't ask the driver to divert or stop at the liquor store. And I've heard women say they won't take Line (to their home/work at least) because if there's a creepy guy in the car that gets dropped off after them, the guy knows where they live/work/hang out.

Lyft is more upfront about tipping too, which is why I prefer to use them now. I have issues with both, but Lyft seems to actually try to give the idea that they care about their "employees" whereas Uber seems to be on a literal crash course to making something work before self-driving cars can take over and they can fire all the humans.

In my experience, and I have driven for both, Lyft actually has been more humane to me in the role of driver. Uber's customer service (remember, drivers pay to use the network) is if you don't like it, you can leave. Lyft on the other hand has taken responsibility for every mistake they have made with me and has fixed it. They have also viewed my feedback about processes and offerings as a gift not a nuisance.

And Lyft's incentives are geared in a way where they give a lot, even up to where it looks to me like a loss on paper, to any driver who is dedicated and is bringing something good to the table.

The bottom line is, driver compensation sucks because people aren't willing to pay what the service is actually worth. They pay $10-$30 a day to Rideshare, and ignore the fact that a new car TCO is 2-3x that. And they get to spend their time doing something else instead of driving. Some people work, some socialize with their friends, some seek entertainment from their devices. And then there are those who avoid the $10,000 DUI.

Things are "worth" the price that the market will bear. If rides were worth more, at the current supply, then Lyft would charge more.

Why should a new car be valued by market principles but rides not?

To increase price, you must decrease supply.

That's hilarious. Uber and Lyft alike are running on a VC-subsidized business model. Leaving the comparisons of personal transport solutions aside, neither of them are turning a profit.

Doesn't matter. I mean, we can argue about the intrinsic value of a ride, but the only way to know the market value of something is to sell it. Or to buy it, if you're on the other side of the situation.

I'm not saying the value won't change in some years when the VC money dries up and the market shifts. Of course it will. To think otherwise would make me like the taxi medallion owners complaining about their mistaken investments.

Another way to look at it is that buying a car is a bad investment, because it's tough to recover the cost of ownership. Perhaps people buy them for non-business purposes and are driving the price up.

Uber drivers are commercial drivers. There's no way around it. They should have a chauffeurs license at minimum and be subject to DOT limits on hours per week and hours per shift and hours between shifts.

It's ride sharing, they are not commercial drivers, besides the robots will be driving for us shortly. This is more a story about people needing to resort to Uber to make money and in a city not near their home because even Uber isnt enough money in their home town.

Is it really ride sharing if you aren't sharing a destination with the driver? How is that ride share any different than a cab ride?

I've used uber in SF, NY, and London. It's painfully clear they are a taxi service going from 1 "customer" to another.

I'm sure there are ppl that legit share their ride to work, but I'd love to see the numbers of ride shares vs serial taxi-like behavior

Legally it's ride sharing and that's all that really matters from a legal, economic or employment standpoint. The fact these people cannot find other jobs is the real problem.

Actually legally it is whatever the law of the land says it is.

>besides the robots will be driving for us shortly.

Maybe in California. Up here in Canada, "the robots" are still paralyzed by the snowy roads that cover the huge majority of our roadways in the winter. I suspect that this is also true in large parts of Europe.

Even a large majority of the US (geographically, not population wise). Any place in the top half of the US which is not on the pacific coast is going to be facing snow and ice for a good portion of the winter.

Living in the Silicon Valley, where the days are always above 50 and mostly sunny, really skews your expectations of what the rest of the country faces for weather.

Disagree. Drivers are being paid. It's commercial. They are providing transportation to the public on public roads, therefore qualifications of the drivers and fatigue, etc. are matters of public safety.

From the standpoint of the law you are incorrect. Legally it's ride sharing, these people can make their own schedule, no one is forcing them to work 100 hour weeks behind the wheel.

Have their been test cases in the relevant jurisdictions or is this just Uber's interpretation?

The test case has been their business operations for the past 7 years (or long they have been in business). They require no certification, no medallion, nothing from the government to operate.

So you are just repeating Uber's opinion then.

I think the long term answer is to copy the rules for aviation, where a Private Pilot Licence is different from a Commercial Pilot Licence, and the rules about what the holders can do are clear and widely publicised. "Uber for planes" would be highly illegal, and everyone involved would get arrested.

Even without regulation, there must be a huge legal risk that Uber's lack of a serious fatigue management system will be found to be negligent. The ways this goes wrong in the trucking industry are very well known.

> "Uber for planes" would be highly illegal, and everyone involved would get arrested.

Unless, of course, it was actually Uber doing it.. then it would suddenly be made legal in 90% of the places it operated and have a blind eye turned to it in the other 10%. I saw a ton of gypsy cab services get shut down in my city before Uber came along.

> "Uber for planes" would be highly illegal, and everyone involved would get arrested.

Are you referring to the recent movie "I.T." https://en.wikipedia.org/wiki/I.T._(film) where Pierce Brosnan plays the role of a airline company CEO who roles out an app that does for planes what Uber does for cars? He gets wrongfully arrested in the movie, but not for his "Uber for planes", but for something else. The roles in that film didn't square with reality. Brosnan portrayed an ethical family man and company director ignorant about IT, whereas in real life he'd be fiddling his company books, cheating on his wife, and using his business's IT infrastructure to illegally monitor his employees. The antagonist could talk to girls and run a sophisticated surveillance system, but couldn't shrug off a snub by Brosnan's character. None of it was believable. I wouldn't use anything in that film as a reference to real life.

Probably to the real-life rules for non-commercial pilots, that strictly forbid to transport passengers for profit. They can only fairly share the operating costs, meaning if there are 3 "passengers" the pilot has to pay 25% of the cost and can not be paid any compensation for being the pilot, so you always lose money on it (in the US, but I assume rules are similar in many parts of the world).

I heard anecdotally that drivers with ratings below 4.6 have their Uber driver accounts shutdown.

In this instance, how would the rating work for drivers in an UberPool? Does each person's vote count for only 1/n, or is each one a full vote? If it's full votes for all riders in a single UberPool ride, all it probably takes are a couple of mediocre UberPool experiences and the driver is selected out of the driving pool.

I thought it was anything below 4, but who knows.

When I have unsafe drivers I aggressively rate them 1 or 2s depending on what it was. I know that it's "mean" to the driver, but I'm being kind to future passengers, pedestrians, and everyone who shares the road.

I wouldn't believe any assumptions around how Uber ratings affect drivers. In many cities Uber's demand outweighs their pool of available drivers. Its not in their best interest to remove drivers from their platform other than the most egregious cases.

> I heard anecdotally that drivers with ratings below 4.6 have their Uber driver accounts shutdown.

Uber and Lyft say they do this, but I've personally seen drivers with lower ratings kicking around. Anyone know what's up with that?

It's closer to stack ranking than a strict cutoff.

Uber doesn't disclose their exact process, of course, but it appears to be mostly stack ranking. Firing the bottom 10% or whatever ends up meaning around a 4.6 in most markets, but median driver ratings vary by market. The stack rank isn't immediate either, and may have factors such as number of trips driven. But that's basically what's going on. (I don't know specifically how this applies to UberPool ratings)

Being ready to fire workers quickly leads to additional recruiting pressure, workforce churn, and ultimately lowers the quality of service.

I think that's a natural effect of finding something that "works, but not at scale". If you try to have as many drivers at the same driver cost that cab companies have[1], you have to keep digging deeper into the applicant pool for less desirable and more frantic drivers (who realize they need to do a lot of fares to come out ahead).

Uber's use of technology has certainly delivered a better experience per dollar spent, but it still runs into these effects.

[1] cab companies don't pay drivers directly, but there's an "implicit driver cost" that underlies how much they're able to rent the cabs for relative to the revenue drivers pull in

There's a reason why Uber has taken almost $9B in funding. I say it's for a legal warchest: innovate without restriction and just deal with legal and ethical "messes" as they occur.

I remember when YouTube was bought by Google, they put aside $1B in a similar fashion to weather any legal troubles coming their way.

Reply on news.ycombinator.com