> he was clearly falling short of Uber’s standard... I gave him five stars, of course.
Why? I wonder if the author would continually frequent a deli that regularly served him rotten meat out of some misplaced sense of guilt.
> He wouldn’t let us out of the cab until she promised to [give 5 stars]. Afterward, we agreed: That was a pretty annoying ride, but not so bad it was worth punishing someone financially over it.
Yeah, I always find it annoying when I get quasi-kidnapped too. But who's to blame the kidnappers? Petty corruption and extortion is a totally acceptable part of life in a civilized society! It's not as if there were some sort of completely anonymous, dead-simple mechanism to instantly eliminate this sort of behavior.
> Most regular Uber users have a story like this.
I don't, nor have I heard any such thing from anyone I know. Where's the data to back this statement?
> my sample wasn’t big enough to generate anything I’d call data
Ergo, you have no business writing authoritatively on this subject.
> Why would Uber invite riders to use a scale of one to five unless it wants them to make fine distinctions? But in practice, it means a rating intended by a user as a gesture of approval is in fact a vote to have a driver fired. That’s screwed up.
If it's so screwed up, then why would anyone drive for Uber? Surely, all of their drivers would be getting fired and no one would want to work for them if that were the case. The fact is, it's not, and the reason is clear if you have a basic understanding of statistics, particularly normal distributions.
> Uber likes this system because it enjoys being able to say all of its drivers have near-perfect ratings. But it’s a harsh one for drivers, and also for customers, who find themselves repeatedly forced to choose between guilt, spite and ignorance.
Again, if it's so harsh, why would either group voluntarily participate in the system? Nobody is forcing either side to participate, and yet they both do so in droves.
> Fundamentally, a five-star rating is far too simple.
And yet the article you cite advocates something radically simpler.
Wait, is this serious?
> he was clearly falling short of Uber’s standard... I gave him five stars, of course.
Why? I wonder if the author would continually frequent a deli that regularly served him rotten meat out of some misplaced sense of guilt.
> He wouldn’t let us out of the cab until she promised to [give 5 stars]. Afterward, we agreed: That was a pretty annoying ride, but not so bad it was worth punishing someone financially over it.
Yeah, I always find it annoying when I get quasi-kidnapped too. But who's to blame the kidnappers? Petty corruption and extortion is a totally acceptable part of life in a civilized society! It's not as if there were some sort of completely anonymous, dead-simple mechanism to instantly eliminate this sort of behavior.
> Most regular Uber users have a story like this.
I don't, nor have I heard any such thing from anyone I know. Where's the data to back this statement?
> my sample wasn’t big enough to generate anything I’d call data
Ergo, you have no business writing authoritatively on this subject.
> Why would Uber invite riders to use a scale of one to five unless it wants them to make fine distinctions? But in practice, it means a rating intended by a user as a gesture of approval is in fact a vote to have a driver fired. That’s screwed up.
If it's so screwed up, then why would anyone drive for Uber? Surely, all of their drivers would be getting fired and no one would want to work for them if that were the case. The fact is, it's not, and the reason is clear if you have a basic understanding of statistics, particularly normal distributions.
> Uber likes this system because it enjoys being able to say all of its drivers have near-perfect ratings. But it’s a harsh one for drivers, and also for customers, who find themselves repeatedly forced to choose between guilt, spite and ignorance.
Again, if it's so harsh, why would either group voluntarily participate in the system? Nobody is forcing either side to participate, and yet they both do so in droves.
> Fundamentally, a five-star rating is far too simple.
And yet the article you cite advocates something radically simpler.