Feedback re Beta Forum Rep System

[Thanimal]

I'm fairly intrigued by this system, and hoping it can be useful. In it's current incarnation, I am very skeptical that it can be. I see that there are many barriers in place to prevent abuse, but I still see a couple of fundamental problems:

1) The current system hugely favors prolific posters. I'll explain more about why that is in a moment, but the first thing I want to highlight is that that is a very bad feature. Not only do we already have access to everyone's post count, but I regard the entire point of the reputation system to be to help figure out how likely it is that a single post is of value. Weighting by post count, even implicitly, is directly antithetical to this.

The reasons I believe the current system favors prolific posters are two-fold: A) The most prolific posters already have HUGE reputations -- even those that some folks seem to actively dislike. B) Somebody fishing with 100 nets should be expected to catch 100x as many fish as somebody fishing with 1. But that doesn't tell me they're a better fisherman.

Some sort of normalization by post rate should be added or else reputation will be too highly correlated to post count to be useful.

2) The current system reduces the weight of those who are most picky. (I'd argue those picky folks should be weighted MORE, but that may just be because I am one.) This occurs because of one of the safeguards, that requires you to spread favor around before re-favoring someone. I realize something like this is required to prevent intentional manipulation by a small group, but it also means that I, as someone who prefers to reserve favor for those that really impress me, end up getting fewer votes.

Perhaps the spread-around constraint can also have an expiration timer on it. Maybe after 48 hours you can hit the same person even if you haven't hit "enough" others yet? This seems likely to mute any systematic abuse, but allow folks to favor those they feel deserve it, without being required to favor those they don't feel deserve it first.

[Mudcnd]

What i don't get is the new mod is months away , and the majority of the player base is begging for new content , and people are quitting because of the lack of new content , but yet dev time is spent making this little "popularity" club thing.

I have been reading the forums for as long as the game was out and not once do i recall any one person posting that they want a little forum popularity contest.

To me it seems that turbine is trying to well squelch anyone that doesn't spew pro turbine stuff all the time , turbine is it really time to alienate some of your player base, cuz well thats how i feel. Before if you offended someone if it was warranted you got infraction points , but now in addition to that you get negative rep if the forum addicts disagree with your post.

If you want to do this system be fair and give all paying customers the ability to give negative rep .

Heh i just posted that the new found power of the forum addicts should be given too all players , im sure i will see red boxes soon.

[Borror0]

The current system hugely favors prolific posters.
I agree that is a problem.

Some sort of normalization by post rate should be added or else reputation will be too highly correlated to post count to be useful.
That can't happen without letting everyone vote on every post, positively and negatively.

Even then, it has a lot of flaws:
1. Rating them becomes a choir
2. If making neutral value posts, you lowers your average
3. Moderating ratings becomes insane

[Fenrisulven6]

yet dev time is spent making this little "popularity" club thing.

I doubt it took much time to swipe the code from the political forums.

Never liked Rep Systems. Always abused and, at best, only encouraged peeps to conform to mob rule.

Stupid idea, but whatever.

[Gornn]

What i don't get is the new mod is months away , and the majority of the player base is begging for new content , and people are quitting because of the lack of new content , but yet dev time is spent making this little "popularity" club thing.


The people developing this rep system are not the same ones developing content for the game. These are web guys and they don't have the same job as the game coders.

Thats all...

[Laith]

The people developing this rep system are not the same ones developing content for the game. These are web guys and they don't have the same job as the game coders.turbine's web team just installed, setup titles & images, and switched forum reputation on.

The code for reputation was created for vBulletin (http://www.vbulletin.com/docs/html/reputation) by Jelsoft. Scroll all the way down. turbine didn't code these forums: they bought them.

[Gornn]

turbine's web team just installed, setup titles & images, and switched forum reputation on.

The code for reputation was created by vBulletin (http://www.vbulletin.com/docs/html/reputation).

My point was less to do with who specifically did the work and more to point out that a lack of recent content had nothing to do with the forum system being implemented.

Though, given how easy it is, you actually enhanced that point.

[Laith]

Though, given how easy it is, you actually enhanced that point.glad to see you caught on :p

[Belwaar]

My point was less to do with who specifically did the work and more to point out that a lack of recent content had nothing to do with the forum system being implemented.

Though, given how easy it is, you actually enhanced that point.

Gornn's spot on. The "devs" are really the online web team...not the game devs, fwiw.

...I can't believe I just sided with Gornn...:eek::rolleyes::eek:

[Thanimal]

I agree that is a problem.

Yep, I would have placed a large bet that you'd agree that favoring post-count was a problem, even though that obviously favors you. Your objectivity is impressive, but no longer surprising. Thanks.


That can't happen without letting everyone vote on every post, positively and negatively.

That's too broad of a claim. A simple system that divides your rating by your post count (since rating began) would eliminate the advantage of frequent posters. However, that is far from a perfect solution. From your list, the "flaw" that remains is:


2. If making neutral value posts, you lower your average

BUT, this is directly compensated by the "100 nets" phenomenon -- someone with a lot more posts also has a significantly higher chance of someone reading one and deciding to favor it (assuming at least some of them are good posts, of course!). The two effects cancel each other out, which is exactly what I am aiming for.

In fact, I'm going to go a step further, and argue that making neutral value posts *should* lower your average. One of the core problems of a forum is getting past the noise to the information. If people carefully consider the value of each post before posting it, then I consider that to be an important criteria for giving them favor.

My "simple solution" is not complete -- many issues remain. However, it does address a part of the problem, and therefore something like it could be part of the solution.

[Borror0]

Your objectivity is impressive, but no longer surprising.
Well, I try to be as much as I can be but many posters on these forums seem to think otherwise.

A simple system that divides your rating by your post count (since rating began) would eliminate the advantage of frequent posters.
The problem is that users have a limit of uses of reputation per day.

Thus, if someone makes fifty awesome posts, he will have a lower ranking than someone who makes one awesome post per day and then logs out. It also penalizes users for posting neutral comments like "QFT", "LOL" or simply a more neutral post that is good but not one that most will bother giving reputation for.

BUT, this is directly compensated by the "100 nets" phenomenon
Not if it's someone who posts very little.

Making neutral posts does not have a strong correlation with posting a lot.

In fact, I'm going to go a step further, and argue that making neutral value posts *should* lower your average.
I see your point but you are assuming neutral posts are not adding to the community. That's false.

If the process of posting a post changes from "is this a good post?" to "will it get me enough reputation?", there is something wrong with the system.

[Thanimal]

If the process of posting a post changes from "is this a good post?" to "will it get me enough reputation?", there is something wrong with the system.

QFT

(Arg -- did I just lower my reputation! :))

Well, let's continue to explore. I feel like this rep system *could* have value, but I'm pretty sure its value so far is very small. I think a lot of folks (including Borror and I) agree on some of the flaws, and perhaps better solutions will come to mind if we keep thinking about it!

But if not, I may eventually find myself in the "just scrap it" camp.

[Gornn]

Gornn's spot on. The "devs" are really the online web team...not the game devs, fwiw.

...I can't believe I just sided with Gornn...:eek::rolleyes::eek:

/hug.

I approve this message. :D

[Issip]

I agree with the OP, and unfortunately I don't see any easy fix. The forum rep may as well be a straight tally of how many posts someone has - as an engineer I can garantee I can directly correlate, probably linearly, the number of posts someone has made with their rep, hence it adds no data value beyond an indicator of how often someone posts.

I tried to garner negative rep, and even that didn't work, I got positive rep for my negative posts - go figure. Ican keep my rep down by not posting, but I can't actually get bad rep because I'd have to cross the line that the cube guards to even try.

[cdbd3rd]

I agree with the OP, and unfortunately I don't see any easy fix. The forum rep may as well be a straight tally of how many posts someone has - as an engineer I can garantee I can directly correlate, probably linearly, the number of posts someone has made with their rep, hence it adds no data value beyond an indicator of how often someone posts.

I tried to garner negative rep, and even that didn't work, I got positive rep for my negative posts - go figure. Ican keep my rep down by not posting, but I can't actually get bad rep because I'd have to cross the line that the cube guards to even try.

Welll...http://home.att.net/~cdbd3rd/wellumm.gif
Adding to your problem with that, are the folks who "approve" your attempt, and thereby inflict further positive rep onto your project.

:rolleyes:

[Thanimal]

I wonder if perhaps I haven't given this system enough time, at least with respect to my first point (implicitly weighting by post-count).

At the moment, it is very hard to get negative rep, because so few people have crossed the threshold where they can give it.

However, long term I would expect negative rep to be a lot more common. This matters because those with high post counts are *also* much more likely to get negative rep.

It's possible that, over time, this will cancel out the advantage currently held by high post count.

Anyone buyin' that?

[Borror0]

It's possible that, over time, this will cancel out the advantage currently held by high post count.

Anyone buyin' that?
That's why I have not starting ranting yet. :p Oh, and help for your collection of boxes.

[Thanimal]

That's why I have not starting ranting yet. :p Oh, and help for your collection of boxes.

Thanks! Only about 1400 to go and then I can "return the favor." :)

Mwaaaa haaa haaa haaa!

[Baranor]

Why do I feel like a dog that goes to park with his master and when my master pretends to throw the ball I just run around looking for the ball like an idiot.