I think it's a good idea, percentage is a statistic I think the best researchers will strive for. Sure as Darcy said you won't always win when deserved and vice/versa but I think it would even out. I do however think it wouldn't be a bad thing if people did strive to offer the best.

If it was implemented the list should only include people who have answered at least 50 questions. Since a new member could walk in give 1 best answer and walk away from the site with a 100% best answers.

It's a good idea, obviously not a top priority idea but certainly a great feature if the devs could get 'er done.

We will have a sort based on number of best answers, but I've never thought about the percentage of best answers. It can't hurt to have more filters of the data... it would be nice to see who are the snipers in the system: the people who take few shots about hit on them.

I think we'll see different types of folks emerge in the system and certainly there these three:

1. Snipers: only answer question they think they will get best answer for.
2. Shotguns: those who answer as many questions as possible to help as many folks as possible--but as a result have a lower % (but might have a much higher total number of best answers).
   
3. interrogators: Those who just ask questions, but don't answer them. 

More power to each type of user.

Yes, it should be.

I will note that I am a biased source. At this moment, I'm at 40% most helpful. My time on here is limited - not all that much, but I don't have enough time to answer every single question on here that I could answer usefully. If I only have a bit to add, or can't add more than someone has already put forth, then I either don't respond, or I add my little bit of knowledge in a comment - I don't think a little tidbit is worthy of an answer, and may in fact serve the asker better, as you can comment on the answer that your tip may be building off of.

I am not saying that points or belts should be eliminated, but if there are going to be multiple ways to sort people, but all of those ways are accumulation based, which can all be achieved with quantity. Do we judge Baseball players by how many hits they had, or their batting average? (Or, better still, on base percentage). An analogy here would be that the accumulation stats would be like "hits", "walks", whereas "best answers" is more like batting average. It's not a perfect measure. It's not on base percentage, but it is a better indicator of quality than any of the accumulation stats.

As someone that pushes for more "best answers", I choose the questions I answer more carefully than someone else might - but if I answer something else, I probably won't be spending the time answering the question I can handle better - so I will choose the question I'm best suited to answer OR the question in most need of a quality answer (IE, no responses, or only a couple poor quality responses). I'd be more inclined to go for more quantity if I had more time to spend here, but, as it stands, there are more questions than what I can answer reasonably, so I will answer the ones I can answer best, to be the most helpful to the people asking the questions.

Picture this site, a year down the road. Frequent contributors will average 50-150 points/day, but a quality, infrequent contributor might only average 10-20 points/day. It won't take too long before all of the infrequent contributors are completely knocked off of all of the top charts, whereas being able to look at best answers % will give some of them a little bit of recognition. The best answers % could even go by belt, so you can see quality and quantity: top 25 best brown/purple/green/yellow belts. This way, people of like quantity can be compared - or there could be some other baseline, like number of best answers.

In Darcy's case against this, she notes some specific instances where what she feels is the best answer did not "win", or that an answer she made that she didn't feel was "best" won anyway... I have been on the same two sides, but a significant majority of the time, what I feel is the best answer is chosen. If the asker chooses, frankly, it's just a game like "Apples to Apples" - some people will try to tell the asker what they want to hear. But most will be more prone towards supporting their opinions/facts. So, I'd say that very, very, rarely is a "best answer" chosen by the asker going to be incorrect...

That human subjectivity is also what makes Mahalo Mahalo. No, everything is not always going to work out perfectly, but sometimes, it's more helpful to have human help, even if imperfect, then it is to look at raw information yourself and decide. So, sometimes you'll have a great answer that isn't selected, others, you'll wonder why yours was chosen. Given that for each of those cases, there is someone with a better/worse answer, there's another half to that equation - it all balances out pretty closely.

Answers chosen by vote? There are few better pieces of wisdom than those from the crowd. If the asker doesn't know what the best advice is, letting the crowd speak its will should act as a solid backing to whomever crafts the best possible argument, regardless of possible disagreements. A truly well crafted argument can, and will attract people that disagree with the point they are arguing from if their research is thorough, and points clear. One can put forth an argument that goes against the grain here, and still win over the crowd. My biggest knock against this is that there could and should be more diverse opinions for best results - there are plenty of topics where I see a very strong consensus on here, to the point where someone making a strong argument against the consensus is stifled somewhat, even if that argument is justified. But I have seen the one going against the grain come out with a "best answer" and positive comments: this community should be as inclusive as possible. More people = always better.

Lastly, I will close in saying that I've not once been mad, or angry, or remotely emotional about not getting a "best answer", even if I felt I was the most informative and helpful. I go "huh" for .1 seconds, and carry on with my day... though in the back of my mind I look at it, and see why the other answer was chosen so I have a better chance the next time the chance to answer a similar question comes around. This is the internet. It is for fun, learning, and discussing things with cool people like you guys. I might disagree with you on a given topic, but that doesn't mean I don't welcome discussion. I'm always happy to see any of your constructive responses to any question I've answered, and, by all means, feel free to refute/disagree with me. We can banter back and forth a bit and the asking person or crowd can decide who helped them the most.

I would say no. While there are some members who are focused on their percentage of best answers, focusing too much on this number can lead to people getting way too upset when they are not chosen. People also might not answer questions unless they are assured they will be chosen--and I think sometimes a group can provide more information that an individual.

I know in my lengthy history of best answers there are some that I lost and some that I won that made me say "WTF?"

One of my favorite examples (that makes me chuckle) is one that occurred while the system was in beta: http://www.mahalo.com/answers/movies/does-anyone-know-what-the-wilhelm-scream-is

I posted my answer, searched for a great clip, and included the link to wikipedia BEFORE Jason posted his answer. He came along, copied from the link I provided (I am sure unknowingly) and got best answer. (I have no hard feelings towards Carrie or Jason about this. MA was still new back then. I can see why she chose Jason's answer.)

To contrast, I have no CLUE how I won best answer for this question:
http://www.mahalo.com/answers/teaching/does-anyone-know-of-an-opening-for-a-grade-812-history-teacher-in-or-within-driving-distance-of-boston-ma

I answered with no intention of getting the best answer. When I saw the question, there were several other answers there, however, I didn't see my favorite site listed so I thought I would add it. If I didn't know better, I would swear I rigged the voting on that one--but I didn't. I'm not even sure I could without it being caught eventually.

My point is, we will all win and lose questions that we don't agree with. In the end, the best answer is a subjective decision. Having that be one of things highlighted could lead to some hard feelings.

I also think percentage of best answers is a very useful measure.

It's esp useful for people that don't answer a lot of questions, but answer really well when they do, and for newbies who are doing a great job but haven't racked up a lot of points or tips yet.

You'll probably need to limit people who are included to those who've answered at least 8 questions though. Otherwise there will be people at the top of the table that have only answered one and got 100% best answers, or only two and got 50%, which isn't very meaningful.

Since it would be only one way of sorting the results out of many, I don't see why it would be a problem, any more than seeing the top 100 by tips earned is a problem.

I take Darcy's point that it might deter people from answering questions when they think they won't get best answer. But two points about that. a) If people haven't got a lot to add, sometimes it's best all round if they don't answer. b) The current points system actually encourages people to answer, even if they haven't got anything much to say.

Since we'll have lots of measures, I don't think we need to worry too much about things like that though.

I do though agree with what's in the background of Darcy's thinking, that we don't want this place to get to be all about competing and keeping score and forget about having fun or helping people.

I'm not sure *any* way of keeping score can be free of the danger of people caring more about the score than being helpful.

And part of the fun of this place is the special mix of game, community, and service.

One thing that is nice here is that we have several ways of thanking and rewarding people. Tips, best answers, marking answers helpful, voting for best answers, and of course, the good old fashoned way, by telling them.

Come to think of it percentage of helpful answers is also an excellent measure, and it would be interesting to see the top 100 for that too.

Or maybe not because a lot of us will be packed in the 95%-100% range!

Last point about percentage of best answers... there is actually a hitch in the way that is calculated. Because what that percentage shows is number of answers selected as a proportion of total answers given. But there are lots of answers that haven't had a best answer selected, either because it's too soon, the asker forgot, or not many people have voted on it. So those cannot count towards your possible number of best answers, and your true best answer ratio is always understated. It's especially understated for newbies, who have the highest proportion of questions answered that haven't yet had any best answer selected.

In summary, I think best answer ratio is a very useful measure, and people with good ratios deserve recognition. We're not going to be relying solely on any one measure of people's contribution or participation, so I can't see any real downside with having that as one of several available indicators.

I agree that there should be another measure of "top" contributors rather than just belt color. When you think about it, purple belt is pretty much a demonstration of volume, not quality. Most (not all) of the folks who achieve that quantity of interaction with the site are employed by Mahalo.

The problem with best answer alone as a guide is that it becomes too much a personality game, where you're more likely to recognize and reward top members who you personally like. Which would be fine for recognition, except it will prevent others from ever breaking into that inner circle.
(Example? Of the top purple belts, Jason Calacanis has the highest percentage of best answers. Is that because his answers are indeed the best, or because people recognize the boss and subconsciously give his answers an upgrade because of it? No offense, Jason :))

Personally, I'm not here for volume, I'm just here to contribute quality where I can. It'd be nice to get recognized for it, sure, but I'll take what I can get. Perhaps a measure that combines several factors - say, percentage of best answers as well as percentage of *not helpful* answers working against you? I spotted a few purple belts with not helpful percentages over 10%, for instance. That would prevent driving up your score by simply responding to every question under the sun. Throw in a bonus for your user rating, and perhaps then you end up with a score of someone's overall quality contribution to the site, since all those measures are voted on by the users.

Yes, Yes, Yes. Mahalo needs to add anything and everything that makes this beta more navigable and interesting. Like why the can't I sort by dollars? And why is the topic navigation at the bottom of the column on the page? Why are their scripting errors slowing down my browser? Why doesn't the default page go to my topic of expertises? Why? Why? Why? Make the pain stop.

I don't think the measure would be particularly useful. It mostly tells you what sorts of questions the person is willing to answer (the survey and opinion type questions are always a long shot).

On the other hand, offering the sort on the menu costs nothing and would be trivial to program so why not?

I suppose that having this ranking should be somehow rewarding for those who put much efforts into answers and it could drive people to do better for the sake of the "competition".
It's a nice idea