Fix Bias in Pardons and Disputes

[Admires]

I feel that many times there is a sort of bias that occurs in pardons and disputes when one staff members tends to agree with other staff members to conform to the majority or previous denies/supports. Saying this doesn't happen simply isn't true because it's human nature to contain bias and conform. This could easily be fixed by somehow not showing staff members' decisions until all decisions have been made. Obviously, new evidence can be shared. I'm just talking about the actual support/no support. If there are other ways to get rid of this conforming bias, I'm open to ideas. I just believe this would make disputes and pardons a bit more fair, and it would rely on a staff member's actual beliefs with the evidence provided.

 

All staff supports/no supports are non-bias.

No support.

 

[Tyler]

Seeing as it would be very hard to do this, and there are alot of talks about pardon that goes on behind the scenes in the staff lounge I cant support this suggestion.

 

Cool idea! I'll pass this on to other people on staff.

 

[Admires]

Seeing as it would be very hard to do this, and there are alot of talks about pardon that goes on behind the scenes in the staff lounge I cant support this suggestion.

I was under the impression each staff member looks at the evidence and provides their support or no support. If there is a ton of behind the scenes group work, then I guess this suggestion wouldn't really apply. I just wanted to get rid of conforming bias that is inherent in many humans.

This would essentially get rid of things like an admin or global giving two no supports, and then a sectional would give a no support just because the admin and global gave a no support because of conforming bias. If each individual gave a response without knowing others' responses, the system would be better IMO.

 

I can tell you now this would never happen at the upper level. In terms of at a lower level, what system would you implement to let this occur? Would eliminating some of the conforming bias result in discerning the users that truly deserve another chance over those that don't better than before? It'd be a logistical nightmare and I don't see the benefit to the idea.

 

[Admires]

I can tell you now this would never happen at the upper level. In terms of at a lower level, what system would you implement to let this occur? Would eliminating some of the conforming bias result in discerning the users that truly deserve another chance over those that don't better than before? It'd be a logistical nightmare and I don't see the benefit to the idea.

I believe that it would result in a better system. Most people have conforming bias without even knowing it. People want to fit in, after all.

 

I support.

This is highly reasonable, why should someone's decision have an effect on another persons decision?

Maybe a secret voting poll for staff?

Put their two cents in on the comments of the pardon, but cant say their decision on the thread. Only way for the decision to be confirmed is by voting on the poll.

 

[Admires]

All staff supports/no supports are non-bias.

No support.

This is honestly a silly statement. How can you be sure there is no bias when bias exists in all of us?

 

Not a fan of this one. Sometimes I'll look through a few staff member's opinions of bounce ideas off one another to see their thoughts. It's helpful to see other's inputs and double think your own stance.

 

I believe that it would result in a better system. Most people have conforming bias without even knowing it. People want to fit in, after all.

Elaborate on how it would be a better system

 

[Yousuckv2]

I feel as if there would be certain aspects that would come to mind for different staff members that may not for others which could have a huge effect on whether or not the consensus is that the pardoning user should be un-banned. Seeing that another staff member has posted support or no support and given a valid reason why could change another staff member's vote for the better. I cannot recall any recent examples of this being a serious issue because in all honesty 49/50 of the users who pardon are scumbags that likely won't have a chance in USL anyways, and if they're that 1/50 person actually deserving of a pardon, there's generally no issue.

This is regarding pardons, of course. When it comes to disputes, again I don't notice this being that much of an issue, and I have seen multiple occasions where another staff member's perspective has been requested and it has been resolved as a result, so once more I don't really see a need for this.

 

[Apith]

I feel like you've chosen a certain pardon or 2 that you don't agree with, and don't like the overwhelming support from staff. If we all support, then there's a pretty good reason to, and majority of the users probably will if they were ever in our shoes. You are assuming there is bias and assuming there is a problem. Have a look at these instead of nitpicking a few.

http://www.sythe.org/request-pardon...ardon-scamming-70-5m-has-peen-refunded-2.html

http://www.sythe.org/request-pardon/1923084-my-pardon.html

There are 10 threads right now with 3 or more posts that you can see. 5 of those doesn't even have any staff supporting or not, it's just asking for explanations or more proof of something. That leaves 3 threads that staff have voted heavily in the favor of one side (I believe all 3 were no supports). So this is what is visible to the public right now, 2/5 of the pardons have mixed supports and no supports. Then there is also another process where the Globals+ vote on the pardons. Which also means if there's a mix of supports and no supports among sectionals, then there's already a pretty good chance you won't make it through the next and final stage. Not to mention not every heavily supported pardon gets accepted in the USL. There are normally discussions between us, hidden posts, information from other users, that leads us to vote heavily on no normally.

 

[Admires]

Elaborate on how it would be a better system

It would eliminate the bias I'm talking about. Evidence could still be added by staff members, just the no support or support should be separate. Dunworry stated above he looks at others' decisions to get ideas. I feel that if staff members made their own decision first and stated why, the bias would be completely eliminated, and over a large sample size, pardons and disputes would be more accurate overall in terms of accepting them or rejecting them.

 

[Admires]

I feel like you've chosen a certain pardon or 2 that you don't agree with, and don't like the overwhelming support from staff. If we all support, then there's a pretty good reason to, and majority of the users probably will if they were ever in our shoes. You are assuming there is bias and assuming there is a problem. Have a look at these instead of nitpicking a few.

http://www.sythe.org/request-pardon...ardon-scamming-70-5m-has-peen-refunded-2.html

http://www.sythe.org/request-pardon/1923084-my-pardon.html

There are 10 threads right now with 3 or more posts that you can see. 5 of those doesn't even have any staff supporting or not, it's just asking for explanations or more proof of something. That leaves 3 threads that staff have voted heavily in the favor of one side (I believe all 3 were no supports). So this is what is visible to the public right now, 2/5 of the pardons have mixed supports and no supports. Then there is also another process where the Globals+ vote on the pardons. Which also means if there's a mix of supports and no supports among sectionals, then there's already a pretty good chance you won't make it through the next and final stage. Not to mention not every heavily supported pardon gets accepted in the USL. There are normally discussions between us, hidden posts, information from other users, that leads us to vote heavily on no normally.

Interesting, thanks for the information. I actually didn't have any pardons/disputes in mind. I just noticed over the past couple years how the system is prone to this bias, and wanted to throw out an idea on how it could be better.

 

[Apith]

Interesting, thanks for the information. I actually didn't have any pardons/disputes in mind. I just noticed over the past couple years how the system is prone to this bias, and wanted to throw out an idea on how it could be better.

I can't speak for every staff, but I was once a member and I do understand where you are coming from. However more than just what you see on the thread goes on.

 

It would eliminate the bias I'm talking about. Evidence could still be added by staff members, just the no support or support should be separate. Dunworry stated above he looks at others' decisions to get ideas. I feel that if staff members made their own decision first and stated why, the bias would be completely eliminated, and over a large sample size, pardons and disputes would be more accurate overall in terms of accepting them or rejecting them.

I'm still not convinced that eliminating conforming bias through this elaborate system will increase our ability to discern users that truly deserve another chance over those that don't better than before better than before.

In other words, the bias exists, bias will always exist with humans. I'm not convinced it matters too much however, certainly not enough to create this elaborate system that we'd need to dance around. Yousuck & Apith nailed the reasoning.

 

[Tyler]

It would eliminate the bias I'm talking about. Evidence could still be added by staff members, just the no support or support should be separate. Dunworry stated above he looks at others' decisions to get ideas. I feel that if staff members made their own decision first and stated why, the bias would be completely eliminated, and over a large sample size, pardons and disputes would be more accurate overall in terms of accepting them or rejecting them.

Alot of staff dont personally know the person who is pardoning, and a lot of the times pardons provide very little information. For example on my pardon Brendan probably did not know a lot about me because he recently came back to sythe, however because I listed what I have done for the community he based his support off of that; however not all situations are going to be like this. If Brendan knew nothing about me except for the Oakdice incident, do you think he would have voted to support me?

 

This is honestly a silly statement. How can you be sure there is no bias when bias exists in all of us?

Yes, but to reflect your opinion by another individual's opinion is good, too. If you think one way, a support from a different member can persuade you otherwise. This isn't bias, this is just how it is.

 

[Bert]

Support this idea. I beleive each mod should make their own choice and should not talk to eachother about a pardon. They should all make their own decision based on each case.