| Pages in topic: [1 2 3] > | Contests: Congratulations to winners in "A translator's life"! Thread poster: Ana Moirano
|
|---|
| | Zea_Mays 
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
| Points in winner pairs? | Oct 13, 2023 |
It seems points are not visible for winner combinations.
| | | | | Chaotic organisation | Oct 14, 2023 |
As many other users have said before, the organisation of contests is chaotic and the voting system flawed.
Let's start with the voting system:
If a contestant votes, they can only vote for other contestants, but who would want to benefit their competitors at the expense of themselves? Proz have claimed that the system balances out because this ‘loss’ is mutual, but that is only true in a contest such as the European Song Contest where it is enforced that all conte... See more As many other users have said before, the organisation of contests is chaotic and the voting system flawed.
Let's start with the voting system:
If a contestant votes, they can only vote for other contestants, but who would want to benefit their competitors at the expense of themselves? Proz have claimed that the system balances out because this ‘loss’ is mutual, but that is only true in a contest such as the European Song Contest where it is enforced that all contestants vote. In the Proz contest, a contestant can opt not to vote, so there is no guarantee that the ‘loss’ suffered when a contestant votes will be recovered from other contestants. This is a major mathematical flaw which discourages contestants from voting. If contestants can vote, it needs to be enforced that they do vote. This can be achieved by allocating a default vote from those who don’t vote, which would be spread evenly amongst the other contestants. A minor penalty could also be applied if they fail to vote. See below.
Then we have the non-contestants: what is their interest in voting? It requires a lot of time to study the entries enough to vote. For what benefit? Proz needs to compensate them in one way or another if their votes are needed.
It is unacceptable to accept entries and then tell the contestants that there is no winner. A proper voting system instead of the current voting mess could achieve that.
The other place where entries go to waste is in the initial phase, where enough people need to submit entries. But before I submit an entry, I would like to know that a winner will be chosen. I believe this could be sorted out by an initial round where potential contestants commit to submitting an entry before time as wasted doing it. This way, we will know which combinations can go ahead. If someone who has committed themselves to do so then fails to do it, I believe a penalty needs to be applied. This could be a deduction from Browniz or Kudoz points or both. This could mean limiting participation to users with enough points to ‘pay’ a penalty.
As Zea Mays pointed out, it would also seem appropriate to show the points achieved after the final round. Why are they hidden?
In any case, I’m not going to waste my time participating again as long is the system is such a mess as it is. Many users have said it before, but Proz are not listening. ▲ Collapse
| | | | |
|
|
|
expressisverbis
Portugal
Local time: 19:34
Member (2015)
English to Portuguese
+ ...
| ChatGPT as a contestant/finalist | Oct 14, 2023 |
Yes, I was told that in another topic. ChatGPT is a contestant and a finalist in some language pairs.
It doesn't surprise me...
[Edited at 2023-10-14 12:06 GMT]
| | | | | Another suggestion | Oct 14, 2023 |
I also suggest that all colleagues (only a very few do) justify their dislikes and not just say "mistranslation", "syntax", "spelling", "inconsistencies", "grammar error", etc.
| | | | Zea_Mays
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
| Points are shown in pairs with no winners | Oct 14, 2023 |
So why not in pairs with winners? Perhaps it's just a temporary situation
as I remember that at the end of the contest all entries were displayed in descending order based on points received.
| | | | Tiago Vieira
Brazil
Local time: 15:34
Member (2023)
English to Portuguese
+ ...
| First time I voted | Oct 14, 2023 |
It was an interesting Contest, I hope next one will be interesting if you add more language pairs.
| | |
|
|
|
Zea_Mays
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
| I had made 7 submissions in total | Oct 14, 2023 |
to see how things would go as I had a certain suspicion. 3 entries for my language combinations (EN>DE & IT, IT>DE)
and 4 entries with 2 source languages I do not work in but which I can read (FR>DE & IT, ES>DE & IT).
Surprise, surprise, in the pairs with a higher number of submissions (EN>DE & IT, 2 of my main working pairs),
I did not pass the first round because my entries had not been graded by enough people.*
(2 ratings for EN>IT and 6 for EN>D... See more to see how things would go as I had a certain suspicion. 3 entries for my language combinations (EN>DE & IT, IT>DE)
and 4 entries with 2 source languages I do not work in but which I can read (FR>DE & IT, ES>DE & IT).
Surprise, surprise, in the pairs with a higher number of submissions (EN>DE & IT, 2 of my main working pairs),
I did not pass the first round because my entries had not been graded by enough people.*
(2 ratings for EN>IT and 6 for EN>DE).
This adds to the other rather questionable aspects of how the contest is run.
The irony is: In those pairs I wanted to see what would happen if I rated the entries of colleagues
(I try to be as sincere and fair as possible in that). I think I did it for all entries for one combination where I did not get to the finals,
for the other pair I don't remember. I also rated entries in some of the other combinations I had made submissions for.
Now, by rating them, in the first stage you contribute to other colleague's entries get through to the finals.
This means that in this stage, it is in large measure the number of voters of an entry that matters, not the quality of the entry.
And the more entries there are, the more likely it is that an entry won't make it to the finals because of this.
It seems that there are also very few people (who didn't take part in the contest) who are willing to vote.
As for AI taking part in the contest, this was said during the Translation Day Event.
The AI profile has been taking part since quite some years, becoming better every year. It's not the first time it made it to the finals.
This is not surprising, as MT tools like DeepL have become very good in some language combinations over the past few years.
If I remember correctly, it was the MT tool mentioned that was used for that AI profile, not ChatGPT.
So, if the AI profile entry made it to the finals, this could just mean that by chance it was seen by more people
(the entries are shown in random order), and/or it is telling us how, um, "good" many of us translators are...
*My entries in the other 5 combinations had good to quite good ratings and were rated by more people
than my entry for EN>IT, where there was a high number of submissions.
It's not the first time I've made submissions to a contest, but once you know how it's run and the related dynamics,
you're not that eager to take part again.
[Bearbeitet am 2023-10-14 14:44 GMT] ▲ Collapse
| | | |
I had a look at the FR-EU PT language pair (closed without a winner) and once again it has Brazilian entries (at least two).
| | | | Zea_Mays
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
| For the chronicles and the stats: | Oct 15, 2023 |
For my purpose, I tried to submit best possible translations (it was a quite unbusy time). See results below.
As I had to produce 7 pieces, quality may have suffered here and there, but it seems to have been quite consistent, though.
As said, my working pairs are EN>DE & IT, DE>IT, IT>DE (target language is mostly DE).
From the results, it seems I should consider working in ES/FR>DE & IT instead. 🤣
There were also an entire chapter to dedicate to the quite different b... See more For my purpose, I tried to submit best possible translations (it was a quite unbusy time). See results below.
As I had to produce 7 pieces, quality may have suffered here and there, but it seems to have been quite consistent, though.
As said, my working pairs are EN>DE & IT, DE>IT, IT>DE (target language is mostly DE).
From the results, it seems I should consider working in ES/FR>DE & IT instead. 🤣
There were also an entire chapter to dedicate to the quite different behaviour of DE vs IT fellow translators.
Point system in first stage: entries can be rated with up to 5 stars
for two categories: quality and accuracy.
Point system in final stage when choosing 1st, 2d and 3th best entry:
1st: 4 points; 2d: 2 points; 3th: 1 point
Results of my entries, with average "stars" in first stage and ratings in finals
IT>DE (no first stage, 3 entries in total)
Finals: 4 ratings, 16 points (4 x4)
Final ranking: first
FR>DE (no first stage, 6 entries in total)
Finals: 5 ratings, 17 points (4 x4, 1 x1)
Final ranking: first
ES>DE (no first stage, 4 entries in total)
Finals: 5 ratings, 16 points (3 x4, 2 x2)
Final ranking: first
ES>IT (no first stage, 6 entries in total)
Finals: 2 ratings, 8 points (2 x4)
Final ranking: 2d
FR>IT (10 entries in total, 4 in final round)
First stage: 12 ratings, 3.17/5 stars > finalist
Finals: 4 ratings, 11 points (2 x4, 1 x2, 1 x1)
Final ranking: 3th
EN>DE (16 entries in total, 6 in final round)
First stage: 6 ratings, 3.08/5 stars
Kicked out
EN>IT (22 entries in total, 7 in final round)
First stage: 2 ratings, 1.50/5 stars 
Kicked out
[Bearbeitet am 2023-10-15 08:26 GMT] ▲ Collapse
| | | | Zea_Mays
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
| Bot participant | Oct 15, 2023 |
https://www.proz.com/profile/1127212
In 2019, DeepL has been used, in 2023, ChatGPT in 30 pairs.
Eliminated in the qualification phase of 20 pairs, it seems it was finalist in target languages like Arabian and Hindi.
For those of you who have a plus subscription, you can watch ... See more https://www.proz.com/profile/1127212
In 2019, DeepL has been used, in 2023, ChatGPT in 30 pairs.
Eliminated in the qualification phase of 20 pairs, it seems it was finalist in target languages like Arabian and Hindi.
For those of you who have a plus subscription, you can watch what Proz said about this here again:
https://www.proz.com/tv/ITD2023,
chapter "Proz.com industry report for freelance translators and interpreters, 2023",
starting at around 29:00 (click on the arrow at the right until you get to this part). ▲ Collapse
| | |
|
|
|
Maybe someone can explain how points are given please. In my main language pair (EN < MT), there are 3 contestants. One can be eliminated as it has lots of errors, and no Likes - I could not find anything to like. My entry as well as the other entry both have 4 points and my entry has 2 likes and 8 Dislikes, the other entrant's has 1 like and 10 dislikes, so how is it that the other entry has ind of moved up a place? Mine used to appear before the other entrant's, at least when I checked. Also i... See more Maybe someone can explain how points are given please. In my main language pair (EN < MT), there are 3 contestants. One can be eliminated as it has lots of errors, and no Likes - I could not find anything to like. My entry as well as the other entry both have 4 points and my entry has 2 likes and 8 Dislikes, the other entrant's has 1 like and 10 dislikes, so how is it that the other entry has ind of moved up a place? Mine used to appear before the other entrant's, at least when I checked. Also it appears then that other people besides the ones who took part, did vote, as, otherwise there would not have been 4 votes for my entry and 4 votes for the other entrant's! Maybe someone can explain. I did not enter for the other competitions because I did not like the source texts. ▲ Collapse
| | | | Zea_Mays
Italy
Local time: 20:34
Member (2009)
English to German
+ ...
You find an explanation on how points are awarded in my forlast post above yours.
Regarding your entry, as you can see, 1 person has awarded 4 points to you,
and 1 person did the same for the other participant. So you are ex-aequo.
Likes/dislikes do not count for ranking.
IMO, likes/dislikes can be quite misleading as often they are very subjective.
At least a third, "orange" tag would be required for preferential notes here.
[Bearbeit... See more You find an explanation on how points are awarded in my forlast post above yours.
Regarding your entry, as you can see, 1 person has awarded 4 points to you,
and 1 person did the same for the other participant. So you are ex-aequo.
Likes/dislikes do not count for ranking.
IMO, likes/dislikes can be quite misleading as often they are very subjective.
At least a third, "orange" tag would be required for preferential notes here.
[Bearbeitet am 2023-10-15 13:30 GMT] ▲ Collapse
| | | | Barbara Carrara
Italy
Local time: 20:34
Member (2008)
English to Italian
+ ...
| Speechless. But with a lot to say, alas | Oct 15, 2023 |
Education
PhD - Learning from others
CV/Resume
Fools. The entire internet is my CV and cannot be contained by your puny file system.
Portfolio
'First mistake.
It was in the first days of April 1813, as promised the morning one Sunday, one of those beautiful days when the Paris to the first time in any dirt on the pavement and no clouds in the Sky looks.'
(For the original French, https://tinyurl.com/5y59pcfa )
The above is a taste of how the bot 'translated' the opening of Balzac's (!) La femme de trente ans from German (!?) into its 'native' (!?) English.
At this point, I'd very much like to know WHO is responsible for filling the bot profile page and what's the reasoning behind selecting literary excerpts for the bot to 'translate' (Literature? Piece of cake!).
Because a human must have done that, right?
| | | | | Pages in topic: [1 2 3] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Contests: Congratulations to winners in "A translator's life"! | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
| | Protemos translation business management system | Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!
The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
