|
Hey Chris,
Right now, when you vote an answer up, it gives 5, 10, 20, or 40 points to the member who made the post based on your status. That part is good.
But correspondingly, the person doing the voting also needs to be similarly rewarded.
Right now it's always 1 point for voting up a post. This needs to be 1,2,4, and 8 based on the member's status.
This will encourage more folks to vote on answers. And that will only improve the QA forum because regular visitors will be able to pick out the best answers more visibly (this is one thing where StackOverflow truly excels at).
|
|
|
|
|
Nishant Sivakumar wrote: This needs to be 1,2,4, and 8 based on the member's status.
Which member? the upvoter, or the receiver?
I think I disagree either way.
1. the upvoter? where will it all end, if, once one is silver/gold/platinum, every action is rewarded 2/4/8 times as much. This is just creating an exponential curve.
2. the receiver? I don't want to get extra points when I agree, admire, whatever something said by one of the heavy-lifters. My effort, and its value to the community, remains the same.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Luc Pattyn wrote: Which member? the upvoter, or the receiver?
Uhm, the receiver already gets weighted points (5, 10, 20 or 40). The voter always gets non-weighted points (1 point).
Luc Pattyn wrote: 1. the upvoter? where will it all end, if, once one is silver/gold/platinum,
every action is rewarded 2/4/8 times as much. This is just creating an
exponential curve.
No exponential curve here. The receiver already gets more points when a higher status member votes his answer up. So correspondigly the voter should also get higher points.
Luc Pattyn wrote: 2. the receiver? I don't want to get extra points when I agree, admire, whatever
something said by one of the heavy-lifters. My effort, and its value to the
community, remains the same.
Huh? It is already this way. If a Platinum guy votes an answer up you get 8 times the score that a Bronze guy will get you.
|
|
|
|
|
I think the receiver should get the weighted points and the voter should receive a constant value (e.g., 1). This is how it is currently done.
My reasoning is that you cannot control who votes on your post, so the weighting will not cause exponential growth as a member gains reputation. However, were a member to get more points for voting, this would cause exponential growth (up to a point) as their reputation grows. Essentially, they'd be getting a lot more points for doing the same stuff.
I'm thinking reputation should be linear, so you can easily see that a member with 1000 points has contributed 100x as much as a member with 10 points (rather than, say, 10x as much).
Fixign now. | But who's fixing the fixign? |
|
|
|
|
|
In that case, then the points for voting an answer up should be increased to 5 per vote. 1 is too low!
|
|
|
|
|
I'll agree to raising the number of points given for casting a vote; but I'm strongly against scaling the number of a voter receives based on his/her status. It would just be a case of the rich getting richer. Gold/platinum members reached their status by a large level of activity on the site. We don't need an additional multiplier to our reputation accumulation.
3x12=36
2x12=24
1x12=12
0x12=18
|
|
|
|
|
Ok, that makes sense.
The whole point in my asking the points to be raised is to encourage voting (which very few people do now).
|
|
|
|
|
It seems that way sometimes, I wonder how many gold/platinum organizers there are vs other rep types.
3x12=36
2x12=24
1x12=12
0x12=18
|
|
|
|
|
Sorry for the late reaction:
Dan Neely wrote: I wonder how many gold/platinum organizers there are vs other rep types.
CP Vanity provides some insight. When looking over the top 5 article count, message count, and total rep point pages in Who's Who, it finds:
Debator; 152 platinum + 12 gold
Author: 93 platinum + 47 gold
Participant: 83 platinum + 67 gold
Authority: 58 platinum + 18 gold
Organizer: 38 platinum + 15 gold
Editor: 11 platinum + 4 gold
Enquirer: 1 platinum + 1 gold
So Organizer looks comparable to Authority, and I think that is good.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
At 1/3rd smaller I think calling it comparable is a bit of a stretch; but it's better than I thought it would be.
Enquirer OTOH definitely needs adjusted to be easier to earn. Maybe editor too, although IIRC until fairly recently you could only edit your own articles/messages/etc; with QA it should be much easier to accumulate points there than it used to be, so maybe we should just give it more time.
3x12=36
2x12=24
1x12=12
0x12=18
|
|
|
|
|
Keep in mind CP does not really offer a way to get the top N performers in each and every specific rep category. By looking at article count, message count, and total rep there are bound to be a lot of the experts and few of the enquiring members in my list. If someone asks 10,000 questions without ever publishing an article or posting many messages, he would not be visible to CP Vanity and yet be the #1 enquirer, with platinum, and not part of the counts I gave earlier.
The color thresholds are different for different categories, and I think they are quite good.
I see two major problems in the rep system:
1. inflation: right now reputation is cumulative without degradation, so the number of platinums and golds will increase without bounds over the years. IMO one should do something about that, maybe one of these:
- drop old actions (calculate rep based on the most recent time span, say 3 years);
- apply a decaying factor to every rep item (say a CP-activity half-life time);
- probably easiest: raise the color thresholds regularly.
2. total rep is being abused: e.g. it is shown in the Q&A system (and in the new forum pop-ups) as the only indication of rep, so answer from a 100K driveler may seem a lot more trustworthy than one by a 10K authority person, where the former may be a long-time soapbox figure, the latter an expert who just joined CP.
Both issues could be addressed, maybe not completely fixed but certainly improved.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
modified on Friday, April 22, 2011 9:27 AM
|
|
|
|
|
If you raise the rep thresholds regularly then you create an old-boys network that allows no newcomers. I also don't think authors are going to appreciate having their articles, one day, suddenly no longer count. The most popular article on the site, dan.g's Todo list, would be an early candidate for having it's points wiped, which I don't think anyone would consider fair.
The bigger point, though: why does it matter if we have more and more top-level members? In fact, I want more high level members. I want more members being able to police the site, to report inappropriate material, to have edit rights, to be able to moderate. This is a Good Thing.
As to the total rep being abused, I can think of only a few specific cases of someone whom I could imagine you labelling a "100K driveler". Does the regularity of this occurrence outweigh the respect the current system gives to the top authors (for instance)? I'm guessing, though, that what might suit you better is simply to show points for technical, not social, skills. We don't have a label for a combined value of Author and Authority, and my fear is showing a number that we aren't presenting in the member profiles would be confusing.
Is the issue truly an issue?
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
|
Are you one of those guys who thinks the economic problems of a country can be solved by printing more money? 
Seriously, if the act of voting an answer up gets you 5 points, then what activity is worth 5 times less (1 point)? I suppose one could see voting a general forum message (e.g., Lounge post) as worth significantly less value, but I'd say they are about the same. 1 point seems fine to me. Maybe 2 points just to show that it has greater value than general votes.
Fixign now. | But who's fixing the fixign? |
|
|
|
|
|
My idea is to encourage voting in the QA forum. It's such a thankless job that very few people do it.
People like to post answers as it gets them 10 points. But the number of members who take the pain to vote answers up is very small. This devalues the forum responses, compared to sites like SO where it's very rare for good answers to go unvoted.
|
|
|
|
|
I'll repeat my question: if you vote my message up, I get points that depend on your color, OK. And you want to get points that depend on whose color, yours or mine?
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Luc Pattyn wrote: And you want to get points that depend on whose color, yours or mine?
Mine.
|
|
|
|
|
As the system currently functions, there is little reward for voting, so I think your idea definitely has some merit. What if simply voting continued to be rewarded by 1 point, and adding a reason, even if it only says "My 5", for your vote gained you an additional point? Personally I find it odd that you can vote and remain anonymous ...
Espen Harlinn
Senior Architect, Software - Goodtech Projects & Services
|
|
|
|
|
Espen Harlinn wrote: As the system currently functions, there is little reward for voting, so I think
your idea definitely has some merit. What if simply voting continued to be
rewarded by 1 point, and adding a reason, even if it only says "My 5", for your
vote gained you an additional point? Personally I find it odd that you can vote
and remain anonymous ...
Well the comment gets you an extra point now. So if you vote and comment, that's 2 points. I think the vote should be at least 5 and as for the comment, I reckon 1 is fine.
BTW I specifically had you in mind when I made this thread. You contribute so much with your answer reviews and appropriate votes, but I don't think you are rewarded fairly for that. It's a thankless job and yet you seem to do it every day. I truly respect that! 
|
|
|
|
|
Copying three words nearly at random from the question, searching google, and pasting in a link to the first answer, without reading it, awards me at least 10 points, more if I was lucky and hit an actual answer.
Working out a reasoned answer is often less awarding ... this seems to justify some change in how we reward people who actually makes an effort when it comes to evaluating answers, as it should serve as a mechanism to improve the average quality of the answers.
The real problem would be to come up with a solution that rewarded me for making at least a little effort checking the correctness of a solution.
Blind voting currently awards me 1 point - That I actually try to put a little effort into the activity gains me nothing as far as the system is concerned, on the other hand it cannot be denied that it encourages some members to review my own answers, and I guess purely blind voting would have the opposite effect
Espen Harlinn
Senior Architect, Software - Goodtech Projects & Services
|
|
|
|
|
Yeah I agree. In fact after seeing your efforts, I've started doing it too. I try and go through the answers and then I mark the best ones 5 and I choose one answer specifically that I think is the most correct approach (in my opinion) and I propose it as answer. Sadly the OP sometimes ignores this and marks a lesser quality response as the accepted answer. Not a big deal though since it's evantually the OP's call there!
|
|
|
|
|
I'm not sure that's such a good idea. I think 1 point for down-voting (because down-voting is a particpatory act, too), and 2 or maybe 3 for upvoting, and that should be as far as it goes. The ability to vote isn't based on level, so the points awarded to the voter shouldn't be either.
".45 ACP - because shooting twice is just silly" - JSOP, 2010
-----
You can never have too much ammo - unless you're swimming, or on fire. - JSOP, 2010
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass." - Dale Earnhardt, 1997
|
|
|
|
|
Members don't need encouragement to downvote, especially malicious ones, and encouraging members to upvote is more valuable. Further, if someone wants to game the system then the pain caused by spurious upvotes is not as bas as that from spurious downvotes.
There will not be points for downvoting. We do reward reporting of items because the reporting system will display the names of those who reported the item should the item get closed. It's transparent to all (especially site admins since we keep track of all reports).
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris Maunder wrote: then the pain caused by spurious upvotes
I would say that's extremely unlikely to happen. Unless someone makes sock puppet accounts to vote their own posts up. That's one reason I suggested the weighted system there!
|
|
|
|
|
I think the voting points shouldn't change. I like points to be somewhat proportional to the combination of effort spent plus value added.
Posting an answer in a programming forum takes some thinking and some doing, and gives you 10 Authority (plus the possibility of getting upvotes and bookmarks, and the risk of getting downvotes). Upvoting (and downvoting and bookmarking) the same message should result in only a fraction of such points to the person voting/bookmarking as it is easy, mostly invisible, and open for abuse.
FYI: bookmarking has been abused for a while now; I haven't reported it before, I just saw another one of my "you're welcome" messages (which I post when my reply gets upvoted and acknowledged with a "thank you" message) being bookmarked. That does not make any sense whatsoever.
Suggestion: there should be a lower limit to a post's length to make it eligible for up-voting and bookmarking. Say if there are less than 50 non-whitespace characters, just don't offer voting and bookmarking tools.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
