General discussion


The Rating Meter should go live this week

By Jay Garmon Contributor ·
As discussed in an earlier thread, we're deploying a new signature system for discussions posts which will include a rating meter. The meter output is based on an evolving algorithm of site participation.

Now, many of you were recruited to the TRI ranks because you "set the bar" for community participation, so some of you might rightly expect to score pretty highly on this first incarnation of the meter. You will, at least relative to the rest of the group.

However, while our heavy hitters will certainly be the highest scorers, but you won't peg the meter. That's because the meter algorithm accounts for participation in features that aren't built yet. You can't score points in these areas because the areas don't exist, so don't be suprised when the best of the best still have ample opportunity to improve.

As we phase in features during 2005, I'm sure you guys will be able to race up the charts as fast or faster than anyone...especially since you'll know about the features first. So now you know.

That, and you guys are usually far better at fielding questions from the community masses than we are, so it's best to keep you informed as much as we can.


This conversation is currently closed to new comments.

Thread display: Collapse - | Expand +

All Comments

Collapse -

Reasoning behind 30 days

by sMoRTy71 In reply to Okay, I have a problem

This has been a topic of much friendly debate here at TR, so let me share our current thinking on the 30-day timeframe.

1. The selfish reason: we want to maintain a super-active community. We hope that, in some small way, the meter acts as a kind of reminder that you need to stay active to stay on top.

2. The better reason: we want new members to feel challenged by the meters of the top members; however, at the same time, we want them to know that if they, too, are very active then they have a chance to be recognized as one of our top members. Our current TechPoint system makes it impossible for a new member to ever reach the top.

Internally, we've been referring to the idea of a "persistent challenge," so things like the meter and other elements of the upcoming profile page design will have many different ways to show/reward your technical knowledge, skill and level of interaction with the site.

How do we avoid junk posts to boost meter?
Although I won't go into all of the details of the algorithm, I will say that the all of the variables are weighted to help minimize "meter abuse." So some community activities (like answering TQ&A questions for example) are weighted higher than others. And if a member wants to post a junk answer to attempt to get a meter boost, their meter will actually take a hit when the member who asked the question rejects it.

As I mentioned in another post, the meter and its algorithm are still in beta. We hope to add more variables so that no one activity will cause wild swings in status.

Please keep the feedback coming since we will be constantly looking for ways to make it better.


Collapse -

point re: rejections

by apotheon In reply to Reasoning behind 30 days

It occurs to me that there's a small problem with the relationship between point values for questioners and answerers. I'm not positive (I haven't read the point-system explanation recently), but I seem to recall reading something about people asking questions having to give up points when an answer is accepted. If that's the case, there's a huge incentive there to reject all answers, even when the answers provide the needed information.

In general, in any case, I think there needs to be at least three consideration stamps possible: Rejected (for answers that are just BAD), Accepted (for good answers, or answers that are in the right vein, but don't necessarily answer the question), and something else for answers that actually provide a solution or the information to get to a solution. That keeps good answers that just aren't perfect for that individual from being Rejected out-of-hand.

The reason the three options seem necessary largely arises from the fact that many people ask questions that are not specific enough to precisely explain what is needed. When someone tries to answer it, often the answer is the Right Answer for a given interpretation of the question, but not for the meaning the question-poster originally intended. As a result, I see a lot of good answers get rejected, which means that knowledgeable and helpful people are getting unfairly shot down.

Thanks again, guys, for all the hard work. It's good to see you taking constructive criticism so well.

Collapse -

Well from my memory

by HAL 9000 Moderator In reply to point re: rejections

"It's been a long time since I actually read the details" but you are able to chose several possible solutions or ratings for answers to a question. They are Rejected, Not Accepted and Accepted.

Naturally with the accepted choice the question should be closed and the points awarded but you also have to remember that TR gives you 1,000 points for asking the question and closing it as well no matter the outcome. I used to spend quite a lot of time in the Q & A section but after seeing far too many Rejected responses to what should be at the very least part way correct answers and the fact that The Chas always seems to be first in with what should be the correct answer I very rarely visit there any more.

While I do not expect my answers to be correct particularly when some very vague questions have been asked the outright Rejected statement without explanation tends to put me off a bit. Even on the few occasions when I've seen a question closed with some form of what has happened I've seen Rejected Responses on what should at the very least have put the person on the right track to a solution.


Collapse -

I feel your pain

by Oz_Media In reply to Well from my memory

Now if we could only get that Chas guy....hmmmmm. JUST JOKING CHAS!! Man, that guy is pretty damn clever with hardware!

I think there is an option in Q&A to reject an answer and still leave the question open for response and then accept it and award the points too.

Now, if someone DOESN'T get the right answer they can reject all answers of course, but there MUST be a forced reason or comment window for them to explain either what DID fix the issue so we all learn from it, or explain that they ended up reformatting or something. It is just plain RUDE to not at least say "Thanks for the tries everyone but I found it to be due to ..."
I mean it is classless and disrespectful, bottom line.

You ask for someone's help and they take time to find links, type out responses, sometimes work for several posts to help someone through an issue and then the answer is rejected and no points awarded. Have a little respect people, what the heck worked????

Yes it is frustrating and after a while it most certainly pushes you away from helping people, now I am even one who has said that we need to put the a**holes aside and continue in Tech Q&A because not everyone is a jerk, but to be proven wrong time and time again sure is frustrating.

Sorry to go on, it's a bit of a peeve for me.

And now that I have ruffled up my feathers a bit, WHERE'S THE DARN 'VIEW ALL OPEN QUESTIONS' LINK!!!! AAAAAAARRRRGGGGG!

Why I oughta.....!

Collapse -

required comments and open questions

by apotheon In reply to I feel your pain

all open questions:
At least, I think that's what that is.

To put it simply, it seems to me that there needs to be a requirement to post a comment, minimum of something like thirty characters perhaps, every time an answer is rejected or a question closed.

Collapse -

What the.....?

by Oz_Media In reply to required comments and ope ...

How the **** did you find THAT page????

Collapse -


by house In reply to required comments and ope ...

That link was posted a little while back. Where were you? You're fired.

Max gave it to us with thanks to TheChas, and I posted it elsewhere under one of apotheon's comments here at TRI when he showed up.

Collapse -


by apotheon In reply to required comments and ope ...

By the way, that's a "secret" link that we're not supposed to pass on to the hoi polloi, I think. It seems that they use a display-time search script to generate that result, which puts a bit of a load on the servers. As such, the general public should not be made aware of its existence, considering the way the servers already labor pretty heavily during peak hours.

Collapse -

Let's go Adam Smith on this

by Jay Garmon Contributor In reply to required comments and ope ...

While sMoRty continues our healthy debate about whether Tech Q&A and Discussions need to be unified, let's proceed right now from the assumption that a distinct Tech Q&A will survive.

We continue to discuss "regulations," if you will, that will force participants in the Tech Q&A "economy" to behave appropriately. As with any regulated economy, excessive oversight risks strangling the market. If we place too many rules on Tech Q&A, no one will want to participate, especially the new users we need to grow the site. This is the "administrative burden" that rightfully concerns sMoRTy.

So, let's conjure up a little Adam Smith, and make a case for more market transparency, not more market regulation. What if we started tracking stats on how often unique users close questions, how often they reject vs. accept answers, how often they post closing comments, how generous they are with techpoints, and how long they leave questions open and idle. Condense these stats, and place key indicators on their Tech Q&A signatures, like we do with the meter in the discussion area.

If you are a persistent offender in Tech Q&A, these indicators will call you out, and the community can self-police by simply refusing to answer an "unfriendly" user's question. Conversely, if you act positively--awarding points and posting back--there will an incentive for others to answer your questions. By becoming more transparent, the market will be able to sort itself out.

Just an alternate idea.


Collapse -

to: Jay, re: market transparency

by apotheon In reply to required comments and ope ...

I love you! Holy cow. Yes. A libertarian approach to T Q & A!

Hmm. Is "Jay" short for something feminine? I could definitely go for a wife that thinks like you do.

Heck, things are changing in this country. Maybe it doesn't matter WHAT your gender is.


Seriously, though, I quite like your idea for how to approach it (even discounting the fact that some of it echoes other comments I've made). Even my suggestion for requiring comments with rejections was an attempt at increasing process transparency, though in retrospect I see that it was a little wrong-headed. Your approach is by far superior.

Let's go with that. I declare it The Best Option (so far). Since my main superpower is Being Right, this means that all should heed your words.

Related Discussions

Related Forums