What's your Uber score?

Started by Xer0, March 15, 2018, 09:04:44 AM

Raza

Quis custodiet ipsos custodes?
If you can read this, you're too close


2006 BMW Z4 3.0i
http://accelerationtherapy.squarespace.com/   @accelerationdoc
Quote from: the Teuton on October 05, 2009, 03:53:18 PM
It's impossible to argue with Raza. He wins. Period. End of discussion.

Laconian

Star ratings should be computed from a simple thumbs up/down. We shouldn't entrust people to make the right call for a 1-5 rating because of millenial entitlement grade inflation get off my lawnwharrgarbl
Kia EV6 GT-Line / MX-5 RF 6MT

dazzleman

Quote from: Laconian on March 20, 2018, 08:04:04 PM
Star ratings should be computed from a simple thumbs up/down. We shouldn't entrust people to make the right call for a 1-5 rating because of millenial entitlement grade inflation get off my lawnwharrgarbl

:lol:

Those rating systems mean very little because everybody has a different idea of what each score on the spectrum means.

We have the same issue at work with employee rankings.  It used to be 1-5, but word got around that whatever the official name of the score (and 3 was "good"), anything under a 4 was bad, and the average better be around 4.5 or else the person would be looked down upon.  So everything was clustered in the 4-5 range and it was impossible to differentiate anybody unless the person was really bad, if then.  They changed the system to 1-9 to try to combat that, but now the "rule" is that anything below 7 is bad, even though 6 is supposed to be the top score for standard performance (and 7-9 are supposed to represent various degrees of exceptional performance).  I have been forced to inflate my scores so that my people don't look bad in comparison to others.  The whole thing is ridiculous.

Even without an overall accounting, it seems a similar thing is happening with Uber scores.  I wonder if the scores the drivers give a somehow analyzed and then "normalized" to try to establish a single standard of rating.
A good friend will come bail you out of jail...BUT, a true friend will be sitting next to you saying, DAMN...that was fun!

TBR

Quote from: Lebowski on March 20, 2018, 07:21:22 PM

With averages in the high 4s all it takes is the occasional 4 or (gasp!) 3 to pull it down.  Maybe the driver expected a tip (before tipping was added to the app), maybe he didn't like something you said, maybe you had him drive out of his way etc. Who knows.

For the customer rating it makes more sense to be binary IMO.  Either it's an acceptable fare, or it isn't (belligerent, barfed in car etc).

Yep, "Would you give this person a ride again? Yes or No"

Soup DeVille

Quote from: dazzleman on March 20, 2018, 07:14:57 PM
If the range is 1-5, all these scores seem pretty high.

I've never used Uber or Lyft, so I don't have a score.

The range is really more like 3.5-5.   
Maybe we need to start off small. I mean, they don't let you fuck the glumpers at Glumpees without a level 4 FuckPass, do they?

1975 Honda CB750, 1986 Rebel Rascal (sailing dinghy), 2015 Mini Cooper, 2020 Winnebago 31H (E450), 2021 Toyota 4Runner, 2022 Lincoln Aviator

Lebowski

Quote from: dazzleman on March 20, 2018, 08:09:13 PM
:lol:

Those rating systems mean very little because everybody has a different idea of what each score on the spectrum means.

We have the same issue at work with employee rankings.  It used to be 1-5, but word got around that whatever the official name of the score (and 3 was "good"), anything under a 4 was bad, and the average better be around 4.5 or else the person would be looked down upon.  So everything was clustered in the 4-5 range and it was impossible to differentiate anybody unless the person was really bad, if then.  They changed the system to 1-9 to try to combat that, but now the "rule" is that anything below 7 is bad, even though 6 is supposed to be the top score for standard performance (and 7-9 are supposed to represent various degrees of exceptional performance).  I have been forced to inflate my scores so that my people don't look bad in comparison to others.  The whole thing is ridiculous.

Even without an overall accounting, it seems a similar thing is happening with Uber scores.  I wonder if the scores the drivers give a somehow analyzed and then "normalized" to try to establish a single standard of rating.


It's not really analogous. An Uber score shouldn't be like a GPA, the vast majority of people on both sides should be very close to a 5. If a driver's car is clean, they drive safely and get you to your destination in one piece that's a 5. If a rider is ready to go and not a complete belligerent ass, that's a 5.

dazzleman

Quote from: Lebowski on March 21, 2018, 07:06:34 AM
It's not really analogous. An Uber score shouldn't be like a GPA, the vast majority of people on both sides should be very close to a 5. If a driver's car is clean, they drive safely and get you to your destination in one piece that's a 5. If a rider is ready to go and not a complete belligerent ass, that's a 5.

I agree with that.  I know how much employees get their balls busted over all this rating stuff, so I rate very highly unless there's a real problem.  If the driver didn't crash the car or get nasty with me, I would give him a 5.

I did a rating for the cruise I just took, and I gave the people who took care of me in the dining room and the room steward a 10 (out of 10).  They were very nice and tried very hard to please, and I am very fortunate to be able to take such a nice vacation.  The least I can do is be nice to the people who helped make it a good time, and not cause any problems for them.
A good friend will come bail you out of jail...BUT, a true friend will be sitting next to you saying, DAMN...that was fun!

Tave

I think it's a decent system. Notice Kevin's been teasing us for "low" ratings but everyone's number is well within his personal tolerance.

Most consumer surveys are fairly crappy because they have to be optional and it creates a huge self-selection bias among the respondents. Uber dialed it in pretty well by making it a mandatory but extremely easy part of using the app.

A binary system for passengers would be bad for the customer, clearly. Drivers would still give low scores unfairly but each one would be worth the equivalent of 2.5 stars in the current system. At least now if you get a crabby driver the worst he'll probably cost you is a star.
As I write, highly civilized human beings are flying overhead, trying to kill me.

Quote from: thecarnut on March 16, 2008, 10:33:43 AM
Depending on price, that could be a good deal.