| Páginas sobre el tema: [1 2] > | Proposal: KudoZ Won/Total Answers Percentage Autor de la hebra: Kirill Semenov
|
|---|
Kirill Semenov 
Ucrania
Local time: 09:03
Miembro 2004
inglés al ruso
+ ...
I think such a parameter would be highly useful. Here are the reasons I mean:
1) It shows the ratio of correct answers vs. total answers, so it may serve as a sign of a pro\'s `efficiency\' when helping askers.
It is very useful since some people are members for years and some joined proZ just a few months ago, so kudoZ points total sum is not really an indication of skill;
2) It may prevent `kudoZ hunters\' from giving random answers as it oft... See more I think such a parameter would be highly useful. Here are the reasons I mean:
1) It shows the ratio of correct answers vs. total answers, so it may serve as a sign of a pro\'s `efficiency\' when helping askers.
It is very useful since some people are members for years and some joined proZ just a few months ago, so kudoZ points total sum is not really an indication of skill;
2) It may prevent `kudoZ hunters\' from giving random answers as it often happens.
A person will think twice or more before trying to get kudoZ pts just for a chance, since losing the answer will decrease his
efficiency ratio.
3) It seems to be easy to implement, so it will not cause much load onto the server.
I guess it would be also reasonable to indicate the percentage for each category (and entirely for the pair). I mean, a pro may be good in one category/pair and not so good in another, so it is useful for potential clients and other pros to see his `efficiency\' in each field separately.
Another way (almost the same) is to show the ratio of kudoZ points won per answer in this or that category (and pair, of course).
Sorry, if I just repeat an old proposition which was casted back long ago. ▲ Collapse
| | | | gianfranco
Brasil
Local time: 03:03
Miembro 2001
inglés al italiano
+ ...
| We are considering doing it | Mar 27, 2003 |
If you look in the FAQ, there is a section containing a detailed description of something similar to your proposal
See: http://www.proz.com/?sp=faq#reliability_ratio
It has not yet introduced but it may be introduced at some point in the future.
The reasons behind this change, which is by the way quite controversial, is that above a ... See more If you look in the FAQ, there is a section containing a detailed description of something similar to your proposal
See: http://www.proz.com/?sp=faq#reliability_ratio
It has not yet introduced but it may be introduced at some point in the future.
The reasons behind this change, which is by the way quite controversial, is that above a certain threshold the site would like to emphasize the quality and accuracy, rather than the quantity of the contributions.
It is, of course, an imperfect system but even the simple accumulation of KudoZ points is far from perfect and exempt from criticism.
The members that provide help for a long time may be seen as too far ahead from the new members and to have reached some kind of inassailable position.
I remind that earning KudoZ points is not, by any mean, a perfect measure of competence but rather a sign of generosity and willingness to help colleagues with advice and terminology.
The new system could be fairer and reflect better the accuracy of the answers provided.
It is still an \'index\', not an absolute indication of skill or else, and just like the KudoZ raw total, has to be taken with a pinch of salt.
Gianfranco
[ This Message was edited by: gianfranco on 2003-03-27 17:17] ▲ Collapse
| | | | CHENOUMI (X)
inglés al francés
+ ...
Quote:
It is very useful since some people are members for years and some joined proZ just a few months ago, so kudoZ points total sum is not really an indication of skill;
I sometimes sympathize with newcomers who must feel at a terrible disadvantage, in comparison with long-time members. They must feel that earning massive KudoZ points is a daunting accomplishment and an extraordinary feat, if no... See more Quote:
It is very useful since some people are members for years and some joined proZ just a few months ago, so kudoZ points total sum is not really an indication of skill;
I sometimes sympathize with newcomers who must feel at a terrible disadvantage, in comparison with long-time members. They must feel that earning massive KudoZ points is a daunting accomplishment and an extraordinary feat, if not an utopia for them.
To avoid the huge gap, KudoZ points can be reflected on a yearly basis instead. This is the only way for the site not to intimidate future members.
For instance, in the year 2003, we would have a list of KudoZ leaders, and the following year, we would start anew with a fresh list. This would be fair, accurate and not established on a cumulative basis, which is, of course, very unfair to new ProZ.com members.
Regards,
 Sandra
[ This Message was edited by: CHENOUMI on 2003-03-27 18:01] ▲ Collapse
| | | | Anila Mayhew (X)
Estados Unidos
Local time: 23:03
inglés al albanés
+ ...
| Couldn't agree more | Mar 27, 2003 |
To avoid the huge gap, KudoZ points can be reflected on a yearly basis instead. This is the only way for the site not to intimidate future members.
For instance, in the year 2003, we would have a list of KudoZ leaders, and the following year, we would start anew with a fresh list. This would be fair, accurate and not established on a cumulative basis, which is, of course, very unfair to new ProZ.com members.
Regards,<... See more
To avoid the huge gap, KudoZ points can be reflected on a yearly basis instead. This is the only way for the site not to intimidate future members.
For instance, in the year 2003, we would have a list of KudoZ leaders, and the following year, we would start anew with a fresh list. This would be fair, accurate and not established on a cumulative basis, which is, of course, very unfair to new ProZ.com members.
Regards,
Sandra
[ This Message was edited by: CHENOUMI on 2003-03-27 18:01]
[/quote]
Sandra,
I couldn\'t agree with you more. Let\'s hope it can be implemented.
Anila
[ This Message was edited by: Mayhew on 2003-03-27 18:50] ▲ Collapse
| | |
|
|
|
| shame that... | Mar 27, 2003 |
the answers are often chosen by complete beginners, who can - and often do - opt for the wrong ones. So, maybe you give 100 correct answers, but because of the ineptitude and little experience of the askers the incorrect ones are selected... all of a sudden, you are a donkey.
Regards,
Giovanni
| | | | #41698 (LSF)
Malasia
Local time: 14:03
japonés al inglés
+ ...
| Pinch of Salt (II) | Mar 27, 2003 |
Yes,
- it does give some indication.
But beware of
- wrong answers chosen (Giovanni said)
- discourages people from giving answers if he/she doubts that it would be the best answer.
| | | | Andrea Ali
Argentina
Local time: 03:03
Miembro 2003
inglés al español
+ ...
| Not always... | Mar 27, 2003 |
[quote]
On 2003-03-27 16:08, Kirill wrote:
I think such a parameter would be highly useful. Here are the reasons I mean:
1) It shows the ratio of correct answers vs. total answers, so it may serve as a sign of a pro\'s `efficiency\' when helping askers.
It is very useful since some people are members for years and some joined proZ just a few months ago, so kudoZ points total sum is not really an indication of skill;
2) It may prevent `kudoZ hunters\' from giving random answers as it often happens.
A person will think twice or more before trying to get kudoZ pts just for a chance, since losing the answer will decrease his
efficiency ratio.
--------------------------------------------
1)The fact that your answer is chosen or not does not necessarily mean it is correct or wrong:
- Maybe it was in second or third place
- Maybe it was \"correct\" but the asker found another answer more appropriate
- It may happen that you have 3 chosen answers out of 4 given in a particular pair/field. This DOES NOT mean you are an expert in this field but simply you came across a question for which you know the answer. In the same way, you may give 50 perfectly correct answers (within your field of expertise) and the second \"Maybe\" is applied...
2) As per \"KudoZ hunters\", let them be!
Just my personal opinion, this is by no means a disagree to your ideas!!!
Regards,
Andrea
| | | | Paul Stevens
Local time: 07:03
Miembro 2003
español al inglés
+ ...
| Chenoumi and Giovanni are spot on! | Mar 27, 2003 |
Whilst I sympathise with Kirill\'s comments about newcomers, it is a drawback that we have all had to live with since becoming a member of the site.
However, after a while, if you answer questions on a regular basis, you will quite quickly accumulate points if you are a good translator.
As for Kirill\'s suggestion, I strongly agree with Chenoumi that, quite often, because of the asker\'s inexperience, the wrong answer is chosen. This would therefore ma... See more Whilst I sympathise with Kirill\'s comments about newcomers, it is a drawback that we have all had to live with since becoming a member of the site.
However, after a while, if you answer questions on a regular basis, you will quite quickly accumulate points if you are a good translator.
As for Kirill\'s suggestion, I strongly agree with Chenoumi that, quite often, because of the asker\'s inexperience, the wrong answer is chosen. This would therefore make the proposed system somewhat ridiculous.
I favour Giovanni\'s suggestion about yearly totals, which is a very good one, but would also like to see the accumulated totals kept. ▲ Collapse
| | |
|
|
|
Kirill Semenov
Ucrania
Local time: 09:03
Miembro 2004
inglés al ruso
+ ...
PERSONA QUE INICIÓ LA HEBRA | Askers' influence | Mar 27, 2003 |
Quote:
On 2003-03-27 19:50, Verbum Ltd. wrote:
the answers are often chosen by complete beginners, who can - and often do - opt for the wrong ones. So, maybe you give 100 correct answers, but because of the ineptitude and little experience of the askers the incorrect ones are selected... all of a sudden, you are a donkey.
I agree it happens. But in the long run, it seems unlikely al... See more Quote:
On 2003-03-27 19:50, Verbum Ltd. wrote:
the answers are often chosen by complete beginners, who can - and often do - opt for the wrong ones. So, maybe you give 100 correct answers, but because of the ineptitude and little experience of the askers the incorrect ones are selected... all of a sudden, you are a donkey.
I agree it happens. But in the long run, it seems unlikely all 100 answers are neglected by askers\' fault. The big numbers law cares about balance.
BTW, I think the percentage has to be calculated after a registered pro has answered 10, 20, 50 or 100 or more questions -- to prevent 100% ratio for just one good answer to the only one simple question. ▲ Collapse
| | | | | Backing up the Kirill's proposal | Mar 27, 2003 |
Though Giovanni\'s concern is very reasonable.
What should be done with beginners opting for a wrong (or half-wrong, or incomplete, or ambiguous, or misleading, or, or, or...)answers???
It\'s so simple:
The Reliability Ratio could be calculated as follows:
1. Weighed Point Number (WPR) = (0,7 * Number of Points for Chosen Pro Answers + 0,3 * Number of Points for Chosen Easy Answers)... See more Though Giovanni\'s concern is very reasonable.
What should be done with beginners opting for a wrong (or half-wrong, or incomplete, or ambiguous, or misleading, or, or, or...)answers???
It\'s so simple:
The Reliability Ratio could be calculated as follows:
1. Weighed Point Number (WPR) = (0,7 * Number of Points for Chosen Pro Answers + 0,3 * Number of Points for Chosen Easy Answers) : Total Number of Points
So, different weight of Easy and Pro Answers can compensate for the occasional \"incompetence\" of some Easy-Askers
2. Adjusted Answer Grade (AAG)
One should take the peer notes into account:
e.g.
For each point of the chosen answer
each \"agree\" adds +0,1
each \"neutral\" substracts -0,1
each \"disagree\" substracts -0,2
So, a 4-point-answer with 2 \"agrees\" and 1 \"neutral\" will weigh 4,4 points. This grade should be used for subsequent calculation of WPR and Reliability Ratio
3. Finally, Reliability Ratio of Answerer
RR = (WPR)*(Total Number of Answers Given) : (0,7*Number of Pro Answers Given + 0,3*Number of Easy Answers Given)
An example to illustrate this approach:
Answerer A: Suggested 10 answers (8 pro, 2 easy). Gained 2 pro-questions á 4 points and 1 easy-question á 2 points.
Peer notes: pro questions: 3 agree, 1 neutral
easy questions: 2 agree, 1 disagree
(Today it is: 10 points)
Answerer B: Suggested 10 answers (2 pro, 8 easy). Gained 1 pro-question á 2 points and 2 easy-question á 4 points.
Peer notes: pro questions: 1 agree, 3 neutral
easy questions: 5 agree, 2 disagree
(Today it is: 10 points as well)
Step 1: Calculation od AAG:
Answerer A:
AAG pro = (1+3*0,1-1*0,1)*4*2= 9,6 points
AAG easy = (1+2*0,1-2*0,2)*2*1 = 2 points
Answerer B:
AAG pro = (1+1*0,1-3*0,1)*2*1= 1,6 points
AAG easy = (1+5*0,1-2*0,2)*4*2 = 8,8 points
Step 2: Calculation of WPR:
Answerer A: (0,7*9,6 + 0,3*2):(9,6+2)= 0,65
Answerer B: (0,7*1,6 + 0,3*8,8):(1,6+8,8)=0,36
Step 3: Caluclation of Reliability Ration of Answerer (RRA)
Answerer A: [0,65*10] : (0,7*8 + 0,3*2)= 1,04
Answerer B: [0,36*10] : (0,7*2 + 0,3*8)= 0,94
So then the Answerer who deals with pro questions more, gains a higher ranking,
q.e.d.
All quotients are chosen arbitrarily. I\'d like only to suggest a possibile way out: not to expose the Answerer to the risk to be undervalued, but to weigh this risk in some proper and objective way.
I believe that the computer will calculate this new KudoZ faster than I did.
Criticism of any kind as regards this model is welcome and would be appreciated )
Yours,
Sergey
[ This Message was edited by: Sergey Strakhov on 2003-03-27 22:31] ▲ Collapse
| | | | | Oh, Sergey ! | Mar 27, 2003 |
Are you a mathematician or an engineer?
I hope computers won\'t make mistakes because I wouldn\'t be able to see or to correct them!
You give the impression that you can give anything an equation form.
| | | | CHENOUMI (X)
inglés al francés
+ ...
| A Post-scriptum to my suggestion | Mar 28, 2003 |
Quote:
I favour Giovanni\'s suggestion about yearly totals, which is a very good one, but would also like to see the accumulated totals kept.
Actually, this was my suggestion! , but I think you got caught up between the \'i\'-sounding names,
CHENOUMI
GIOVANNI...
In the event the KudoZ calculating system is... See more Quote:
I favour Giovanni\'s suggestion about yearly totals, which is a very good one, but would also like to see the accumulated totals kept.
Actually, this was my suggestion! , but I think you got caught up between the \'i\'-sounding names,
CHENOUMI
GIOVANNI...
In the event the KudoZ calculating system is implemented on a yearly basis, that would in, no way, obscure members\' past accomplishments! We could still come up with an icon or link pointing to the previous total of KudoZ they have accumulated!!
CHEERS for the KudoZ and ProZ.com! ▲ Collapse
| | |
|
|
|
Kirill Semenov
Ucrania
Local time: 09:03
Miembro 2004
inglés al ruso
+ ...
PERSONA QUE INICIÓ LA HEBRA | Has anyone said "to re-rank" ? | Mar 28, 2003 |
Quote:
On 2003-03-28 07:26, 1964 wrote:
I think some new comers try to re-rank all proz.com \"as to what they have in hand\" possibly because they answer less and higher \"hit\" in their answers.
Since I am the only new member in this topic, the remark is probably aimed at me. 
Please, do not attribute to me an idea of \"re-r... See more Quote:
On 2003-03-28 07:26, 1964 wrote:
I think some new comers try to re-rank all proz.com \"as to what they have in hand\" possibly because they answer less and higher \"hit\" in their answers.
Since I am the only new member in this topic, the remark is probably aimed at me.
Please, do not attribute to me an idea of \"re-ranking\", since I had not said anything like that. Moreover, I do not think the reliability ratio or something has to substitute current total points ranking system.
A reliability indicator, calculated on this or that algorithm, should probably be just an additional feature. ▲ Collapse
| | | | gianfranco
Brasil
Local time: 03:03
Miembro 2001
inglés al italiano
+ ...
| Pinch of Salt (III) | Mar 28, 2003 |
I just would like to say that all systems are imperfect, including the ratio proposed in the FAQ, Sergey\'s system and any other variation we can imagine.
The current system is also very imperfect, and who knows it cannot disagree.
The programmers of this site and the moderators have been discussing a lot these issues and the conclusion is that there is no simple solution.
In any case, if the system is changed, the new recipe will try to strik... See more I just would like to say that all systems are imperfect, including the ratio proposed in the FAQ, Sergey\'s system and any other variation we can imagine.
The current system is also very imperfect, and who knows it cannot disagree.
The programmers of this site and the moderators have been discussing a lot these issues and the conclusion is that there is no simple solution.
In any case, if the system is changed, the new recipe will try to strike a balance between all elements:
- the quantity of answers provided
- the peer-grades
- the points assigned by askers
- the difficulty level of the questions
- the duration of the membership
and then add a pinch of salt to all these ingredients...
The general idea is to take off the edge from the current \"simple accumulation\" of points but to maintain into the equation the generosity shown trying to help other colleagues.
Thank you for all contribution.
Gianfranco
[ This Message was edited by: gianfranco on 2003-03-28 08:53] ▲ Collapse
| | | | Mónica Machado
Reino Unido
Local time: 07:03
inglés al portugués
+ ...
| Abolish KudoZ points | Mar 28, 2003 |
Hello All,
I think we should help when we can (full stop).
Earning points is nice but sometimes makes people shooting answers in all directions just to see if they get any points back. That\'s why some weeks ago I decided stopping that - I have deleted all my kudoZ points. Now I am at the end of the list (with no points) - before I have 89 and send my help directly to the posters if I think it can help them.
I am not on the run to be... See more Hello All,
I think we should help when we can (full stop).
Earning points is nice but sometimes makes people shooting answers in all directions just to see if they get any points back. That\'s why some weeks ago I decided stopping that - I have deleted all my kudoZ points. Now I am at the end of the list (with no points) - before I have 89 and send my help directly to the posters if I think it can help them.
I am not on the run to be seen - true, but I don\'t worry that much when I see funny answers on KudoZ section, either.
Since for me Proz is just a place to change opinions and help others and not a place to get work from, I think it was the best to do.
I still stink KudoZ should be given for other purposes - wonderful people that on this site try to help others all the time, with hints about translation, software and other advices.
All the best
Mónica ▲ Collapse
| | | | | Páginas sobre el tema: [1 2] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Proposal: KudoZ Won/Total Answers Percentage | Trados Business Manager Lite | Create customer quotes and invoices from within Trados Studio
Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.
More info » |
| | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
