| Pages in topic: < [1 2 3 4 5 6] > | 7th ProZ.com Translation Contest: announcement of winners! Thread poster: RominaZ
|
|---|
| Good idea but... | Jul 4, 2008 |
Good idea Scott to show the names, but still, I think that the contest should be judged by an impartial board of judges. But this has been discussed also in the previous contest, and it was agreed that it's not possible to do so, thus...
I wonder if diasqualification could be an option, in case of evidence of foul play...
| | | | | Impartial judges would be nice. If not possible, then publish refs/reasons and (optionally) names | Jul 5, 2008 |
Liliana Roman-Hamilton wrote:
...I think that the contest should be judged by an impartial board of judges. But this has been discussed also in the previous contest, and it was agreed that it's not possible to do so, thus...
Yes, it would nice if we somehow had an impartial board of judges. It can feel uncomfortable voting for or against competitors.
If it's not possible to have an impartial board of judges, then maybe the next-best thing would be to publicize more information (similar to the way Kudoz are done): display the references/reasons accompanying the Like/Dislike tags (and the Agree/Disagree tags), and maybe show the names of the voters (or make this optional, so voters could choose whether or not to have their name displayed).
Sometimes voters posted a Like/Dislike tag which maybe some of the other voters couldn't really understand unless the reasons/references for it were also displayed. For example, a native Russian speaker has now pointed out in the comments that my winning RU>EN entry mistranslated a certain word. I think he probably tagged this word during the voting but since there was no way for him to state a reason, his tag was ignored. Now that the contest is over and he is able to post a complete comment explaining the meaning of this word, more people understand his point. So I think that tags plus the references/reasons accompanying them should be publicized so that if just one voter notices something bad (or good) that nobody else notices, then there is a way to inform the other voters about it.
And I personally would have no problem letting my name be displayed along with my votes (the same way that answers and comments in Kudoz also have the names displayed). I think this only encourages good voting behavior. It can be exciting to be a contestant - but maybe it would also be more exciting to be a voter if your name could be shown with your votes. Some people might not feel like entering the contests but they might actually be motivated to vote if their votes displayed their names - similar to the way that Kudoz works. And maybe we could even give Browniz to voters to further encourage them to vote.
And if names were displayed with votes, then it would discourage unfair voting strategies, and encourage voters to do their best as we way of gaining better credibility as contest judges.
| | | | | mechanism and quality of writing | Jul 5, 2008 |
I am pleased to see that other people share my impressions regarding both the voting procedure and the quality of writing.
I took part in every contest and almost every time it seemed to me that “peculiar things” happened. The time I won my translation got in the final round one hour before the closing; by contrast, another time my translation had been qualified together with only 3 morefor several days, until half an hour before the closing, when about ten more were added. This t... See more I am pleased to see that other people share my impressions regarding both the voting procedure and the quality of writing.
I took part in every contest and almost every time it seemed to me that “peculiar things” happened. The time I won my translation got in the final round one hour before the closing; by contrast, another time my translation had been qualified together with only 3 morefor several days, until half an hour before the closing, when about ten more were added. This time I had 5 dislikes and 1 like, which compared to the general score should have placed my work in the final round, but it was not so. Transparency would be very useful to make us understand the selection mechanism.
As of the quality of writing, the Italian group participating in the forum knows that when I criticize a text I do not do it out of personal frustration, and I can always back up my claim. The winning entry in the EN/IT pair contains one terrible mistake and several less serious ones, besides mistranslations. Is it fair towards those who have better linguistic abilities that it should win?
One thing would be if all the voters met one day in a room and a supervisor made sure that each person left a comment and a mark for every translation. Another thing is to rely on people’s will to commit themselves. We do not know where the comments come from, how many people had the time and fancied leaving comments and votes during the weeks of the contest. We especially do not know how prepared the voters are from a narrative, literary point of view. Both now and in the past several people have expressed their doubts. For these reasons I hope the idea of an external judge will be considered soon. Even one person per language pair would be enough in my opinion, providing that he/she knows the rules of the language. ▲ Collapse
| | |
|
|
|
| You are right | Jul 5, 2008 |
Yes, you are right.
I am tired of saying that the anonimous voting is either incompetent or dishonest (or both, at times), and any sober-minded person will notice regularity in the irregularity of the results in way of poor translations (to say the least about their quality) becoming the winners. No need to give examples: they are way too many.
You are utterly right: even one judge who is fair and competent would be much better than what we witness now.
| | | | | "Anonymous voting is either incompetent or dishonest" | Jul 5, 2008 |
Yuri Smirnov wrote:
Yes, you are right.
I am tired of saying that the anonimous voting is either incompetent or dishonest (or both, at times), and any sober-minded person will notice regularity in the irregularity of the results in way of poor translations (to say the least about their quality) becoming the winners. No need to give examples: they are way too many.
I think you 'hit the nail on the head,' Yuri. Anonymous voting may be important when voting to pick a politician for office - but when posting comments and references about a translation, the anonymity seems counterproductive in this case, and allows incompetent and/or dishonest voting to happen.
Displaying voters' identities (either during or after the contest) is probably the simplest and most powerful way to improve the voting system. Two approaches could be taken:
- adding voter ids to the current system involving tags and popups (and also displaying the references and ratings in the current system); or
- designing a contest voting system similar to the Kudoz system (which might be simpler from a programming point of view, avoiding the use of popups, which seem to be liable to bugs and errors).
On top of the problems with anonymous voting, we also have inconsistencies with the different ways in which voting information is currently used:
- the somewhat detailed voting info (tags and their categories) is displayed but not used in voting
- the most detailed voting info (references/reasons) neither displayed nor used in voting
- the un-detailed, un-displayed voting info (ratings or stars) determines the winners
Not displaying references/reasons is causing major headaches - sometimes a voter will find a subtle but serious error in a translation (which the other voters would never find) and give it a tag saying 'Dislike'/Mistranslations or 'Dislike'/Grammar and adding a reference explaining why - but because the other voters can't see the reference/reason, they remain uninformed about the reference/reason accompanying the 'Dislike' tag, and they may end up voting to 'Disagree' with a perfectly reasonable 'Dislike' tag because they had no idea why the tag was posted in the first place!
(It's rather odd: Why is the reference/reason accompanying a tag not displayed? Does anybody see it? If anybody sees it, do they act on it? And if nobody sees or acts on the reference/reason, then whey are taggers being asked to supply it? As far as I can see, the reference/reason accompanying a tag is discarded - which bizarre behavior for a software system: asking users to input information which is thrown in the trash, unseen and unused.)
And remember the rule (up until the 6th contest) which said that if a voter voted to disqualify an entry during the qualifying round - and that entry actually went on to *win* - then that voter would be barred from voting in future qualifying rounds. This was a good rule! It discouraged dishonest or incompetent voting behavior. But now it's possible to vote against a 'strong' entry 'with impunity' during the qualification phase - just give it 1 star.
I understand the site admins and programmers did their best to develop a new system for the 7th contest, and the new system is quite sophisticated in its capabilities. But, because this is a piece of 'social software,' we need to think about designing a proper set of rules to encourage or reward competent or honest voting behavior (and to discourage or punish incompetent or dishonest voting behavior).
Finding one impartial, capable judge for so many language pairs might be an impossibility, and would prevent the contest from spontaneously expanding to include more language pairs the way it currently does (i.e., currently a contest for a language pair can come into existence whenever enough entries are submitted for that pair, without any additional effort involving in finding a judge for that pair). I believe in the concept of 'distributed moderation' from social software - reflected in the saying 'Many hands make light work.' There ought to be a way to develop a good voting system based on group participation without taking recourse to centralized judges (because having judges could bring a whole additional set of headaches). The key is to design the rules of the voting system to encourage good voting behavior and discourage bad voting behavior.
Perhaps the simplest two rules would be:
(1) Make voter ids public (either during or after voting)
(2) Make voter comments public
There could be two ways of 'making voter ids public' - displaying the names of voters along with their comments immediately - or displaying the names of voters after the voting has ended. (There could even be a third possibility: we could give the voter the option of selecting to be anonymous or to have their id public - but we could 'weight' public votes much more heavily than anonymous votes. This might provide a comfortable way for some people to 'get their votes/comments out there' even if they are hesitant about having their id associated with their votes/comments.) Knowing that sooner or later your name will be seen attached to your votes/comments will encourage people to vote more carefully and responsibly. The publicity might even attract a new set of participants - people who aren't interested in showing their abilities as contestants, but who are interested in showing their abilities (and gaining credibility on the site) as good contest judges.
Rather than getting bogged down in adding all sorts of new rules and procedures to the system, we could simply copy much of the existing framework from the Kudoz voting system:
- get rid of the anonymity for voters
- make comments visible (the reasons/references why a voter likes or dislikes a term)
- maybe award Browniz to encourage voting participation
A simple comment system similar to Kudoz (with many fewer 'bells and whistles' than the current contest voting system with all its buttons and popups, which looks like it was an extremely complex programming job using Javascript or Ajax involving many database tables and columns) might work quite well actually.
So what I'm saying is, copy the model of Kudoz as much as possible, which is simple and transparent and works rather well to come up with the 'best' translation of the source term.
Yes, a contest text is significantly longer than a single Kudoz term - but much of the voting framework from Kudoz could be used as-is for the contests - just with the expectation that there will be many more comments per source text, because a contest source text is much longer than a Kudoz term. But the Kudoz framework itself could be applied quite successfully to contests (keeping the contestants' names anonymous until the contest is over, of course).
Whatever system is adopted, the important features to add (similar to Kudoz) are:
- visible voter ids (either immediately or delayed, or anonymous but 'weighted' much less); and
- visible comments (reasons/references)
Then voters will be motivated to behave competently and honestly when they are voting publicly - and a lot of the "peculiar things" which unadiluna and I and others have noticed about the voting will go away.
[Edited at 2008-07-05 15:18]
| | | | | 'Disagree' tags without visible references/reasons are useless | Jul 5, 2008 |
Liliana Roman-Hamilton wrote:
Well, those mistakes have been one by one disagreed to.
Looks fishy to you too, now? Well, welcome to my club!
Although I don't speak Italian, I understand what you are saying Liliana about the 'Disagree' tags.
A 'Disagree' tag (or 'Agree' tag) has almost no impact if the reason/reference accompanying it is not displayed.
For some strange reason, the current contest voting system does not display the reference/reason accompanying the 'Dislike' or 'Like' tag.
The consequence is that people who want to comment on the contest entries are forced to find other avenues to do so: either in this Forum, or in the Notes which can be added to the contest page for each language pair (after the contest is finished - when it's too late to have any impact on the voting!).
The reference or reason accompanying a tag should be displayed.
| | |
|
|
|
| not fair play.. raus! | Jul 5, 2008 |
Again, I agree with you Scott. There are several bugs to fix in the contests, to maintain some sort of fairplay.
No fairplay? Then disqualify, as in any other game, contest, or competition, otherwise it's useless to spend time to take part in it, vote or select the winners, when at the bottom there is no fairness.
What astonishes me most is that the lack of fairness comes from adults AND professional translators in a translation contest. It really st... See more Again, I agree with you Scott. There are several bugs to fix in the contests, to maintain some sort of fairplay.
No fairplay? Then disqualify, as in any other game, contest, or competition, otherwise it's useless to spend time to take part in it, vote or select the winners, when at the bottom there is no fairness.
What astonishes me most is that the lack of fairness comes from adults AND professional translators in a translation contest. It really stinks, and I'm very, very disappointed. ▲ Collapse
| | | | | I cannot agree more | Jul 5, 2008 |
I cannot agree more.
We've witnessed just the opposite in the English-Russian pair. When a case of 'unfair play' was not 'raus-ed', to use your words, it ended in several people either saying they would have done the same or all the blame went onto those who mention this 'foul play'. Sin unpunished becomes sin triumphant. That's a universal rule.
[Edited at 2008-07-05 16:04]
| | | | Yaotl Altan 
Mexico
Local time: 07:34
Member (2006)
English to Spanish
+ ...
I share almost every single comment stated here about the several ways we can implement to get real good contests. That's why I participated only in the first one.
By the way, don't yout think 7 contests are too much since they were created? This is: don't you think we are sufferein a severe contestitis ? I think there's no hurry to announce the 8 contest. If we schedule contests properly, with these corrections implemented, may be we can improve them and increase quality and user... See more I share almost every single comment stated here about the several ways we can implement to get real good contests. That's why I participated only in the first one.
By the way, don't yout think 7 contests are too much since they were created? This is: don't you think we are sufferein a severe contestitis ? I think there's no hurry to announce the 8 contest. If we schedule contests properly, with these corrections implemented, may be we can improve them and increase quality and users satisfaction. ▲ Collapse
| | |
|
|
|
ebru-d
Local time: 16:34
English to Turkish
| Good opportunity!! | Jul 5, 2008 |
I'd really thank you for organizing such a contest. Unfortunately I've just been informed. I'm a new member and I actually didn't think this site would do any good to me. But I've realised that even reading the comments is enough for someone to improve her or himself. So, thank you for this! And please let me know about the contests next time:-)
Ebru Demirkol
| | | |
Hallo,
I understand my posts were banned, as voting is underway.
So I had a look to the 6th contest: the winner's text has very poor quality, I'm afraid.
It looks like a contest for Uni students not pros. Sorry.
I reiterate what I wrote: this can turn Proz into a "language killer"
and as a paying member I guess I have the right to express my fears.
Regards,
Paola
| | | | | One more connotation | Jul 6, 2008 |
Just one more connotation.
I know at least a couple of very good literary translators who have many books translated (and, of course, proofread and edited as is the technology pattern in books translation). After having attempted to participate one or two times they are now very sceptical about the contest seeing what is ... See more Just one more connotation.
I know at least a couple of very good literary translators who have many books translated (and, of course, proofread and edited as is the technology pattern in books translation). After having attempted to participate one or two times they are now very sceptical about the contest seeing what is required to win the contest. What we get as a result is not an objective picture of the quality of translation: other, weird factors are determining who comes first.
[Edited at 2008-07-06 07:55] ▲ Collapse
| | | | | Pages in topic: < [1 2 3 4 5 6] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » 7th ProZ.com Translation Contest: announcement of winners! | 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! » |
| | TM-Town | Manage your TMs and Terms ... and boost your translation business
Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
