| Pages in topic: [1 2] > |
Answers accuracy rating Thread poster: Dmitry Murzakov
|
|---|
Dmitry Murzakov 
Russian Federation
Local time: 12:00
English to Russian
+ ...
Hi, everybody!
I want to propose a new feature which can be displayed in users’ profiles – percentage ratio of the user’s answers chosen by askers as the most helpful (A) to the whole number of the user’s answers (B). It is calculated by the following formula – (100 * A) / B. For example my percentage ratio is (100*6) : 29 = 20,68 (%). You can provide a bright name for this new feature and choose a number of digits to be displayed after point.
Dmitry.
| | | |
Nicole Schnell
United States
Local time: 01:00
English to German
+ ...
In memoriam | Accuracy vs. helpful | Oct 24, 2010 |
1. Not every question is about technical terms, and there is no such thing as "accurate" answers to questions in the fields of literature, advertising or gaming, for example.
2. I can perfectly view your entire statistics in regard to KudoZ questions. Maybe this feature is available to paying members only, I can't tell.
| | | |
Nicole Schnell
United States
Local time: 01:00
English to German
+ ...
In memoriam
I omitted the topic of showing percentages on purpose because I can not agree with the term "accuracy rating".
| | | |
| Accuracy does not/can not apply | Oct 24, 2010 |
Nicole Schnell wrote: I omitted the topic of showing percentages on purpose because I can not agree with the term "accuracy rating".
Answers selected are 'helpful' in the eyes of the Asker. Which makes the degree of helpfulness completely subjective and dependent on Asker's very personal opinion.
The answers can be inaccurate, ridiculous, impossible etc but still be chosen as 'helpful' by Asker. I am referring to strict terminology questions and not things like ideas for advertising slogans which are just brain-picking/idea-stealing sessions with no right or wrong answer possible.
In the old days, when erroneous answers were chosen (again, strictly relating to terminology questions), people used to rush to the discussion box area to warn that the wrong answer had been chosen. Nowadays, this is a punishable offence so it's no longer done and even erroneous glossary entries are sacred and untouchable.
So 'accuracy' is in the eyes of the beholder and generally speaking, I feel that any 'ratio' based on answers chosen risks being totally misleading.
| | |
|
|
|
| Nothing is new under the moon | Oct 24, 2010 |
Your idea has been discussed ad nauseam on these forums, Dmitry. Please do a search for "ratio" and "Kudoz" and you will see.
| | | |
Dmitry Murzakov
Russian Federation
Local time: 12:00
English to Russian
+ ...
TOPIC STARTER | Have it your way | Oct 24, 2010 |
Mikhail Kropotov
I am new here that's why I could miss the relevant topic.
Nicole Schnell
I have hurried over the term "accuracy rating". I suppose the term like this "helpful anwsers percantage rating" would be better. I can agree that even erroneous answers can be chosen as most helpful but this rating will be useful for outsourcers first of all. Although every one can understand how much helpful answers were given by the user just looking a... See more Mikhail Kropotov
I am new here that's why I could miss the relevant topic.
Nicole Schnell
I have hurried over the term "accuracy rating". I suppose the term like this "helpful anwsers percantage rating" would be better. I can agree that even erroneous answers can be chosen as most helpful but this rating will be useful for outsourcers first of all. Although every one can understand how much helpful answers were given by the user just looking at his answers and scores statistics. ▲ Collapse
| | | |
Yaotl Altan
Mexico
Local time: 02:00
Member (2006)
English to Spanish
+ ...
Dmitry Murzakov wrote:
Hi, everybody!
I want to propose a new feature which can be displayed in users’ profiles – percentage ratio of the user’s answers chosen by askers as the most helpful (A) to the whole number of the user’s answers (B). It is calculated by the following formula – (100 * A) / B. For example my percentage ratio is (100*6) : 29 = 20,68 (%). You can provide a bright name for this new feature and choose a number of digits to be displayed after point.
Dmitry.
It's a great idea, i support it. That could be the main sort configuration as other cofigurations may be available, for instance, the number of answers or whatever you like. Proz has the resources to fix that and make it work as a spreadsheet, sorting columns according to several parameters.
I like your idea!
| | | |
Dmitry Murzakov
Russian Federation
Local time: 12:00
English to Russian
+ ...
TOPIC STARTER
Yaotl Altan
Than you for your support! I hope ProZ administration will hear me.
| | |
|
|
|
Ildiko Santana
United States
Local time: 01:00
Member (2002)
Hungarian to English
+ ...
MODERATOR | Askers ask for help, because they don't know/understand, so how can they judge accuracy of answers? | Oct 24, 2010 |
writeaway wrote:
Answers selected are 'helpful' in the eyes of the Asker. Which makes the degree of helpfulness completely subjective and dependent on Asker's very personal opinion.
The answers can be inaccurate, ridiculous, impossible etc but still be chosen as 'helpful' by Asker. I am referring to strict terminology questions and not things like ideas for advertising slogans which are just brain-picking/idea-stealing sessions with no right or wrong answer possible.
In the old days, when erroneous answers were chosen (again, strictly relating to terminology questions), people used to rush to the discussion box area to warn that the wrong answer had been chosen. Nowadays, this is a punishable offence so it's no longer done and even erroneous glossary entries are sacred and untouchable.
So 'accuracy' is in the eyes of the beholder and generally speaking, I feel that any 'ratio' based on answers chosen risks being totally misleading.
I couldn't agree more!
| | | |
Yaotl Altan
Mexico
Local time: 02:00
Member (2006)
English to Spanish
+ ...
Dmitry:
Take a look on "New forum for submitting suggestions and ideas" at the root of this forum. perhaps you can post your idea or support it if someone else has done it.
| | | |
| The decision is too subjective and accuracy plays no role in it | Oct 24, 2010 |
writeaway wrote: Nicole Schnell wrote: I omitted the topic of showing percentages on purpose because I can not agree with the term "accuracy rating". Answers selected are 'helpful' in the eyes of the Asker. Which makes the degree of helpfulness completely subjective and dependent on Asker's very personal opinion. The answers can be inaccurate, ridiculous, impossible etc but still be chosen as 'helpful' by Asker. I am referring to strict terminology questions and not things like ideas for advertising slogans which are just brain-picking/idea-stealing sessions with no right or wrong answer possible. In the old days, when erroneous answers were chosen (again, strictly relating to terminology questions), people used to rush to the discussion box area to warn that the wrong answer had been chosen. Nowadays, this is a punishable offence so it's no longer done and even erroneous glossary entries are sacred and untouchable. So 'accuracy' is in the eyes of the beholder and generally speaking, I feel that any 'ratio' based on answers chosen risks being totally misleading.
I am against this percentage ratio proposed here. The decision is too subjective and accuracy plays no role in it.
As it has been said, it should be about accuracy not helpfulness, since, in the KudoZ system, helpfulness is decided upon by the asker.
Nicole and Writeaway, above, summed it up very well.
| | | |
Dmitry Murzakov
Russian Federation
Local time: 12:00
English to Russian
+ ...
TOPIC STARTER | Helpful answers percentage ratio | Oct 25, 2010 |
Miguel Carmona
I have already said that formulation of the answer was made in a hurry. I selected another name for it – “helpful answers percentage ratio”. Let’s speak about it now if you want. This ratio will not change anything but it is a comfortable feature to evaluate translator’s professional level.
| | |
|
|
|
Dmitry Murzakov
Russian Federation
Local time: 12:00
English to Russian
+ ...
TOPIC STARTER | This feature has so many disadvantages as a number of scores. | Oct 25, 2010 |
Your arguments are really very persuasive. I think that this feature has so many disadvantages as a number of scores. For example, a number of scores earned through Kudoz doesn't tell also that somebody is better in this industry because he managed he was quicker. And I didn’t mention to attribute more important role to this feature but only a statistics instrument.
| | | |
| The key word is probably Acceptance | Oct 25, 2010 |
I believe a very accurate name for this statistic is "Acceptance ratio." That's what it has been called from time to time in this community.
| | | |
| Honestly I don't think this feature would help to present fairly translator's professional approach. | Oct 25, 2010 |
Dmitry Murzakov wrote:
I selected another name for it – “helpful answers percentage ratio”. Let’s speak about it now if you want. This ratio will not change anything but it is a comfortable feature to evaluate translator’s professional level.
I've seen too many occasions when colleagues rush in with calque translations (while unfailingly opting for highest confidence level!) only to make sure they get there first. It may be a doable strategy to score thousands of Kudoz points, but would you look at it as a mark of professionalism?
| | | |
| Pages in topic: [1 2] > |
Login to reply/comment 
