Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27] >
Netflix Hermes test
Thread poster: Magdalena Adamus
Post removed: This post was hidden by a moderator or staff member for the following reason: The quoted post is deleted
Hurley815
Hurley815
Czech Republic
Czech translator Apr 27, 2017

Hi, guys. Just for comparison, I finished the test (for Czech translators) also about three weeks ago and received the results yesterday. The funny thing is that I passed with 83,3 % and a friend of mine didn't with 76 %, which leads me to believe that the threshold was probably around 80 %. So now hearing that in other countries the threshold was at about 90 %… seems I was more lucky than skilled. (this may be because statistic... See more
Hi, guys. Just for comparison, I finished the test (for Czech translators) also about three weeks ago and received the results yesterday. The funny thing is that I passed with 83,3 % and a friend of mine didn't with 76 %, which leads me to believe that the threshold was probably around 80 %. So now hearing that in other countries the threshold was at about 90 %… seems I was more lucky than skilled. (this may be because statistically, Czech Republic isn't as skilled in English language as other countries, but I'm not sure)Collapse


 
ChachiArcola
ChachiArcola
Italy
Local time: 09:36
Opinions requested Apr 27, 2017

Elena Pizzetti wrote:
From what I am reading from friends and colleagues, the minimum score changes from a language to the other. For example a 84 or 85% in Italian was not enough to pass, while an 80% was enough for other languages.


Elena, what do you think of the ambiguities in the first 2 sections? I may lack experience as a translator, but my knowledge of Italian is rather solid. Nevertheless, I often couldn't tell the difference between 2 of the proposed answers, they looked both right; better, they WERE. Guidelines stating that one had to choose based on "semantic, syntax or best practices" didn't help in my opinion.
Also, I'm under the (uncomfortable) impression that those who simply left the subtitling exercises blank consistently scored more than those who completed the test as per instructions. What do you think, based on your and your friends' experience?

And a last question. I'm surprised to notice that a random base of Italians scored in this test so well, that the threshold had to be raised so high. Again, what's your opinion on this as a professional?
For the English language, the average knowledge in Italy is among the worst in Europe, data say. For example, Italian students as a group constantly score less than average in the Cambridge English evaluation tests, where data are collected worldwide and tests carefully administered under the same conditions everywhere. Aren't you surprised of this very high threshold? I am.


 
Cristiana Coblis
Cristiana Coblis  Identity Verified
Romania
Local time: 10:36
Member (2004)
English to Romanian
+ ...
New languages, launched languages Apr 27, 2017

Hurley815 wrote:

Hi, guys. Just for comparison, I finished the test (for Czech translators) also about three weeks ago and received the results yesterday. The funny thing is that I passed with 83,3 % and a friend of mine didn't with 76 %, which leads me to believe that the threshold was probably around 80 %. So now hearing that in other countries the threshold was at about 90 %… seems I was more lucky than skilled. (this may be because statistically, Czech Republic isn't as skilled in English language as other countries, but I'm not sure)


Yes, I noticed that for Czech and Hungarian the pass threshold was lower, possibly around 80%. I am guessing that this is because the two are new languages and they need more people. For Romanian it was probably around 90% because a friend of mine got over 80% and did not pass unfortunately. Polish and Turkish had similar thresholds to Romanian. Polish and Turkish were launched recently, so they already have some people and probably need some more. For older languages, such as IT, PT and Spanish, where they probably need even fewer people, the threshold was also higher.

They are probably going to start evaluating the subtitling part for some of the tests. I suppose it was like a screening, there is no point to check the subtitling part, where they will need to pay a human, if a candidate got 30% is the previous parts.


 
Mirrors
Mirrors
Italy
Local time: 09:36
my take Apr 27, 2017

Does anyone else find the wording Netflix used to let us know we failed pretty interesting? It states that "at this time" our scores do not reach the required threshold and suggests saving our Hermes # for future identification purposes. It may be reaching, but to me it sounds like they're leaving the door open in case they have a change of heart.

ChachiArcola wrote:

Elena Pizzetti wrote:
From what I am reading from friends and colleagues, the minimum score changes from a language to the other. For example a 84 or 85% in Italian was not enough to pass, while an 80% was enough for other languages.


Elena, what do you think of the ambiguities in the first 2 sections? I may lack experience as a translator, but my knowledge of Italian is rather solid. Nevertheless, I often couldn't tell the difference between 2 of the proposed answers, they looked both right; better, they WERE. Guidelines stating that one had to choose based on "semantic, syntax or best practices" didn't help in my opinion.
Also, I'm under the (uncomfortable) impression that those who simply left the subtitling exercises blank consistently scored more than those who completed the test as per instructions. What do you think, based on your and your friends' experience?

And a last question. I'm surprised to notice that a random base of Italians scored in this test so well, that the threshold had to be raised so high. Again, what's your opinion on this as a professional?
For the English language, the average knowledge in Italy is among the worst in Europe, data say. For example, Italian students as a group constantly score less than average in the Cambridge English evaluation tests, where data are collected worldwide and tests carefully administered under the same conditions everywhere. Aren't you surprised of this very high threshold? I am.


I got 86% and didn't pass either, whilst I read someone with 88.something% did, so I'm assuming the Italian threshold was either 87 or 88%- being this close to passing and missing it for either one or two percent is, I'm not going to lie, fairly disappointing.

As far as the first two sections of the test go, I think that some answers really were too similar to be able to nail them without any context; in some cases, the difference was subtle enough that one couldn't confidently tell 'wrong' from 'right' at all. I find that to be the case with a few multiple choice tests, unfortunately, which is why I personally don't like them much; they are, though, much easier to grade and review, which is why they're so favoured imho. This Hermes test was clearly no exception.

I'm also interested in the answer to your last question- I was under the same impression myself. My assumption is that they simply needed fewer Italian translators than, say, Hungarian ones and the easier way to go about selecting them was making sure the test was harder to pass.


 
Anna Voskou
Anna Voskou
United Kingdom
Local time: 08:36
English to Greek
+ ...
Greek pass grade? Apr 27, 2017

Hello, I did this test in Greek and got a passing grade of 86%, from what I understand, the grade is different for every language?

I've seen people with other working languages with the same grade and failing, so I'd rather not have Netflix contact me about my grade being a mistake...!

Thank you


 
Magdalena Adamus
Magdalena Adamus
Poland
Local time: 09:36
English to Polish
TOPIC STARTER
Don't overanalyse it. Apr 27, 2017

Mirrors wrote:

Does anyone else find the wording Netflix used to let us know we failed pretty interesting? It states that "at this time" our scores do not reach the required threshold and suggests saving our Hermes # for future identification purposes. It may be reaching, but to me it sounds like they're leaving the door open in case they have a change of heart.



I'm sorry to be a sceptic here but after working in Vendor Management for quite some time now I can tell you that this is just a more euphemistic way of saying: You failed. Bye.


 
Jocelin Meunier
Jocelin Meunier  Identity Verified
France
Local time: 09:36
English to French
+ ...
Finally I could pass the test Apr 27, 2017

and fail with 65,33 %. Now, I know I can do mistakes but I'm a professional translator and subtitler. My work is appreciated, I know I make a good job so I can't help but think that this test is seriously flawed (not surprising since it comes from Sfera, but still).
I took it in French and many answers were weird. They didn't sound natural or could have easily been shortened. Sometimes there were two good answers and there was no indication on wether you have to choose the shortest or the
... See more
and fail with 65,33 %. Now, I know I can do mistakes but I'm a professional translator and subtitler. My work is appreciated, I know I make a good job so I can't help but think that this test is seriously flawed (not surprising since it comes from Sfera, but still).
I took it in French and many answers were weird. They didn't sound natural or could have easily been shortened. Sometimes there were two good answers and there was no indication on wether you have to choose the shortest or the most truthful (chose the shortest everytime, seemed the msot logical choice for a subtitling test).
As for the 4th part, there were many subtitles with two issues but you could only choose one. Now that I learn here that the part 5 and 6, the most important, are overlooked completely, I can definitely say that this is not a good test. Of course, I'm not the most perfect subtitler either, but it's really not doing a good job at testing a subtitler. Congratulations to those who passed the test, honestly.
Collapse


 
saretton
saretton
Italy
English to Italian
+ ...
As a matter of fact... Apr 27, 2017

As a student of both English and Russian language, I selected more than one language option in the starting form before getting the test on Hermes, therefore two different tests appeared on my dashboard - one from English to Italian and one from English to Russian. Being Italian my mother tongue, I've taken the Italian one first and completed the Russian one only halfway, also because I'm aware I'm definitely not half as good at Russian language as I am at English; I still have to complete part ... See more
As a student of both English and Russian language, I selected more than one language option in the starting form before getting the test on Hermes, therefore two different tests appeared on my dashboard - one from English to Italian and one from English to Russian. Being Italian my mother tongue, I've taken the Italian one first and completed the Russian one only halfway, also because I'm aware I'm definitely not half as good at Russian language as I am at English; I still have to complete part 4, 5 and 6 of it.

Here's the paradox, though. As I've already written, I failed at Italian - 75.33%, even though I completed the test entirely, translating the subtitles and everything. However, in the light of our recent findings on the matter written in this topic, I could now complete the multiple choice part in Russian, skipping altogether part 5 and 6, and still pass. And I repeat that I'm aware of not being half as good at Russian as I am with English. (Actually, in short, I don't know whether to take the Russian test or not).

Should I be graded positively, it would really upset me (another paradox). I get that they are adjusting the selection criteria bit by bit, as they move from one point to another in the process, yet it starts to seem unprofessional to me. I would have liked better if those criteria had been disclosed to us before we took the test, not changing them mid-process. It's all smoky: no official statements from Netflix concerning the problems this Hermes thing has been having, no clear threshold for any language (we are left to deduce them on our own, together)... nothing.
Collapse


 
Willian Meirinho
Willian Meirinho
Brazil
Local time: 05:36
Different results. Timed sessions doubt Apr 28, 2017

Hi everyone!

I did the test twice and i'm a bit annoyed with the results.

First time i did it, it was everything swift, just to evaluate myself. One of the clips that i had to subtitle, i didn't finish and it wasn't in line with the standard of my language. Still, I've got a 74% score. This first try was arond 4 weeks ago and they took some 3 weeks to evaluate the test ang give me the results.

Second time. This one was 2 days ago. This time, i took time to
... See more
Hi everyone!

I did the test twice and i'm a bit annoyed with the results.

First time i did it, it was everything swift, just to evaluate myself. One of the clips that i had to subtitle, i didn't finish and it wasn't in line with the standard of my language. Still, I've got a 74% score. This first try was arond 4 weeks ago and they took some 3 weeks to evaluate the test ang give me the results.

Second time. This one was 2 days ago. This time, i took time to study the rules of subtitling. I did all the multiple choice questions, even double checked with Google. Aside the part you need to spot the error on the subtitle cutscene, i'm pretty sure i nailed it 100%. At the subtitling part, this time i filled all the lines. I did everything according with Netflix requirements and double checked as well. Now the problem: My result appeared instantly after i submitted and it was just 78%. They didn't not even check the most important part of the test and NO WAY i got more than 10 questions wrong.

I was wondering if they take the time i took answering into consideration, because i double checked everything and submitted with little time left... I don't know. The whole thing seemed very unfair for me.

I'd like to know your opinion about it.

Congrats for everyone who passed.
Collapse


 
Butze
Butze
Italy
German to Italian
Clarification Apr 28, 2017

Magdalena Bardalinska wrote:

Mirrors wrote:

Does anyone else find the wording Netflix used to let us know we failed pretty interesting? It states that "at this time" our scores do not reach the required threshold and suggests saving our Hermes # for future identification purposes. It may be reaching, but to me it sounds like they're leaving the door open in case they have a change of heart.



I'm sorry to be a sceptic here but after working in Vendor Management for quite some time now I can tell you that this is just a more euphemistic way of saying: You failed. Bye.


Hi,
Just to be clear. You think the message for those who have failed is just a more euphemistic wording style but you believe (you were the one to notify us about it) that they may actually reconsider the threshold and thus review fail/pass results? This kind of sounds like a contradiction to me, and I would like to know more about it.


 
Magdalena Adamus
Magdalena Adamus
Poland
Local time: 09:36
English to Polish
TOPIC STARTER
I can only assume what was Netflix intentions with accouncing results of only ABCD part Apr 28, 2017

Butze wrote:

Hi,
Just to be clear. You think the message for those who have failed is just a more euphemistic wording style but you believe (you were the one to notify us about it) that they may actually reconsider the threshold and thus review fail/pass results? This kind of sounds like a contradiction to me, and I would like to know more about it.



This is because I find it highly unlikely that they would go back to evaluating subtitles part in the applications that they failed. Why? Because I think that after they were scared by the amounts of applications they got, they wanted to limit the number of tests that actual human being would need to check. This is why they took only the test part into account when failing/passing candidates. If now they start to evaluate subtitles part in failed test, what was this fuss about with announcing scoring two days ago? This would make no sense whatsoever. Still, I really would like to be wrong - I really hope that they would change some peoples' results after checking their subs even if they failed the ABCD part - that would be the only fair way.


 
rekape
rekape
Hungary
English to Hungarian
+ ...
Strange experience Apr 28, 2017

Hello everybody!

I took the test yesterday for both Hungarian and Romanian languages.
My mother tongue is Hungarian but I grew up in Romania, so I thought I give it a try.

I found the Hungarian test rather ambiguous, at the multiple choice questions some sentences meant the same thing. I didn't really know which one I should choose as I felt more options were correct.
At the part where we had to give the meanings of idioms, I googled everything, just to be
... See more
Hello everybody!

I took the test yesterday for both Hungarian and Romanian languages.
My mother tongue is Hungarian but I grew up in Romania, so I thought I give it a try.

I found the Hungarian test rather ambiguous, at the multiple choice questions some sentences meant the same thing. I didn't really know which one I should choose as I felt more options were correct.
At the part where we had to give the meanings of idioms, I googled everything, just to be sure. I am sure I got them all right.
I feel that the part where only one subtitle comes in, and I have to spot the mistake, was the hardest.
One or two mistakes were obvious (subtitle came in too early, punctuation mistake), but I couldn't find any mistake in several subtitles.
The two translation tests were ok, I could finish them in time.

The Romanian test seemed a little bit easier for me, perhaps I got used to being tested by the time I got there.
The multiple choice questions were ambiguous here as well, but not as much as in Hungarian.
I have googled the idioms again, so those were right.
The part where I had to spot the mistake in the subtitles - this was easier as well. No question mark, subtitle enters too early, leaves too soon, it could be spotted easily.
The last video I had to translate was the famous American football one - I could finish it in time.

A little bit about me: I worked in the subtitling business 18 years ago, and started again a few weeks ago. My mother tongue is Hungarian, and Romanian is a second language for me, meaning I am not as proficient in it as I am in Hungarian.

Now my results: I failed both tests.
The thing what I find rather strange and annoying, is that I got slightly better results in Romanian than I did in Hungarian. Hungarian: 76, 67%, Romanian 77,50%.
According to this test, I can speak a second language better than my mother tongue. This is crazy.
Collapse


 
Marijej
Marijej
Netherlands
Local time: 09:36
English to Dutch
+ ...
85,33%? Apr 28, 2017

Hi, the English to Dutch threshold was 86% and I didn't pass with my 'meagre' 85,33%. There seem to be a lot of subtitlers with that exact score, which supports the assumption that only the multiple choice parts have been evaluated. As the multiple choice answers were so similar and open to multiple interpretations and it looks as though a computer evaluated the first four phases of the test I would like to argue for a human recount, but I'm guessing Netflix won't be willing to look into that op... See more
Hi, the English to Dutch threshold was 86% and I didn't pass with my 'meagre' 85,33%. There seem to be a lot of subtitlers with that exact score, which supports the assumption that only the multiple choice parts have been evaluated. As the multiple choice answers were so similar and open to multiple interpretations and it looks as though a computer evaluated the first four phases of the test I would like to argue for a human recount, but I'm guessing Netflix won't be willing to look into that option...Collapse


 
Butze
Butze
Italy
German to Italian
Yes but... Apr 28, 2017

Magdalena Bardalinska wrote:

Butze wrote:

Hi,
Just to be clear. You think the message for those who have failed is just a more euphemistic wording style but you believe (you were the one to notify us about it) that they may actually reconsider the threshold and thus review fail/pass results? This kind of sounds like a contradiction to me, and I would like to know more about it.



This is because I find it highly unlikely that they would go back to evaluating subtitles part in the applications that they failed. Why? Because I think that after they were scared by the amounts of applications they got, they wanted to limit the number of tests that actual human being would need to check. This is why they took only the test part into account when failing/passing candidates. If now they start to evaluate subtitles part in failed test, what was this fuss about with announcing scoring two days ago? This would make no sense whatsoever. Still, I really would like to be wrong - I really hope that they would change some peoples' results after checking their subs even if they failed the ABCD part - that would be the only fair way.


But do you think they would actually un-pass a pass? This is the absurdity of it all. You cannot announce a pass and then take a step back, what does this say about them?
Of course, it doesn't make sense that they would waste time evaluating the translation part of a clear fail. Also, they could have decided midway to actually make a pass in the first part a requirement but then why go back on your own words/decisions? This is what, to me, doesn't make any sense. If the pass we got was not definitive, then they could have said so, "you passed the first part, now we are going to evaluate the second part and we will get back at you". Since they haven't said so, I am really hoping they won't just change it later.


 
Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27] >


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

Netflix Hermes test







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! »