The game has ended for 8 days, and the discussion has almost quieted down.
\nFirst of all, congratulations to all the kagglers who won medals. However, there are still some people who have been mistakenly removed from the leaderboard and still haven't received any proper response.
\nTake our team as an example. We are removed from leaderboard just because we made one of our additional data sets public for a period of time after the game and then changed it to private. In the subsequent late submission, our rerun was successful and in compliance with the rules.
\nFor some other teams, it may because they deleted the dataset they had uploaded. And based on the communication with @authman ,
\n\n\nI also just now received the same email verbatim, sans the team name.
\n
we guess kaggle gave the same response even we are in a such different situations. Kaggle did not made any other detailed responses to us so far, including what will they do, or just do nothing. In the forum, they only responded to Aidyn Ubingazhibov’s post, which received over 200 upvote.
\nI sincerely thank the kaggle team for their dedication in this competition, including organizing this game and spending a lot of energy to clean up non-compliant submissions. But for me, it is totally unacceptable to be removed a precious medal without doing anything that obey the rules. If the kaggle team needs any evidence from team, please feel free to contact us.
\nI will keep attention to the results of this matter! I also hope that everyone can give us some support, because each of us has an obligation to promote the perfection of the rules. Only in this way can we avoid the same thing happening to ourselves in the future.
", "messages": [ { "id": "987445", "postDate": "08/27/2020 09:09:25", "content": "The game has ended for 8 days, and the discussion has almost quieted down.
\nFirst of all, congratulations to all the kagglers who won medals. However, there are still some people who have been mistakenly removed from the leaderboard and still haven't received any proper response.
\nTake our team as an example. We are removed from leaderboard just because we made one of our additional data sets public for a period of time after the game and then changed it to private. In the subsequent late submission, our rerun was successful and in compliance with the rules.
\nFor some other teams, it may because they deleted the dataset they had uploaded. And based on the communication with @authman ,
\n\n\nI also just now received the same email verbatim, sans the team name.
\n
we guess kaggle gave the same response even we are in a such different situations. Kaggle did not made any other detailed responses to us so far, including what will they do, or just do nothing. In the forum, they only responded to Aidyn Ubingazhibov’s post, which received over 200 upvote.
\nI sincerely thank the kaggle team for their dedication in this competition, including organizing this game and spending a lot of energy to clean up non-compliant submissions. But for me, it is totally unacceptable to be removed a precious medal without doing anything that obey the rules. If the kaggle team needs any evidence from team, please feel free to contact us.
\nI will keep attention to the results of this matter! I also hope that everyone can give us some support, because each of us has an obligation to promote the perfection of the rules. Only in this way can we avoid the same thing happening to ourselves in the future.
", "rawMarkdown": "The game has ended for 8 days, and the discussion has almost quieted down. \n\nFirst of all, **congratulations to all the kagglers who won medals**. However, there are still some people who have been mistakenly removed from the leaderboard and still haven't received any proper response. \n\nTake our team as an example. We are removed from leaderboard just because we made one of our additional data sets public for a period of time after the game and then changed it to private. In the subsequent late submission, our rerun was successful and in compliance with the rules.\n\nFor some other teams, it may because they deleted the dataset they had uploaded. And based on the communication with @authman ,\n> I also just now received the same email verbatim, sans the team name.\n\nwe guess kaggle gave the same response even we are in a such different situations. **Kaggle did not made any other detailed responses to us so far**, including what will they do, or just do nothing. In the forum, they only responded to Aidyn Ubingazhibov’s post, which received over 200 upvote.\n\nI sincerely thank the kaggle team for their dedication in this competition, including organizing this game and spending a lot of energy to clean up non-compliant submissions. **But for me, it is totally unacceptable to be removed a precious medal without doing anything that obey the rules. If the kaggle team needs any evidence from team, please feel free to contact us.** \n\nI will keep attention to the results of this matter! I also hope that everyone can give us some support, because each of us has an obligation to promote the perfection of the rules. Only in this way can we avoid the same thing happening to ourselves in the future.", "votes": null }, { "id": "987685", "postDate": "08/27/2020 12:51:36", "content": "@philculliton @maggiemd @bendvd @stavness
", "rawMarkdown": "philculliton @maggiemd @bendvd @stavness", "votes": null }, { "id": "991024", "postDate": "08/30/2020 04:24:14", "content": "@philculliton @maggiemd @bendvd @stavness Please help us!!
", "rawMarkdown": "philculliton @maggiemd @bendvd @stavness Please help us!!", "votes": null }, { "id": "992223", "postDate": "08/31/2020 04:06:05", "content": "Personally, I gave up. It's way past any reasonable period where they're going to do anything about it.
\nIn this case, Kaggle executed very poorly in terms of getting this code competition finalized. I don't ever recall seeing final scoring delayed this long before. The lag resulted in a myriad of discrepancies due to people altering the status (private/public, versioned/unversioned, available/deleted) of their model datasets during the limbo period post closure, whereas under normal competition configuration, once the competition ended, nothing would have mattered.
\nEven though they failed hard here, they are making some changes. For example, I noticed the submission screen for the bird and the lyft code competitions have some dataset versioning added to it where that did not exist before. People can still get screwed hard if they delete their models after the competition ends, and from Inversion's post, they won't be altering that for some reason(?). Kaggle also failed hard here with their lack of communication to us, their users. Canned responses, if any, and very delayed. All quite disheartening.
\nFor me, this was a good wake up call. Kaggle has been instrumental to motivating me to become a better data scientist. But at the same token, it's important for me not to put it on a pedestal and to recognize it's just like any other company. It has an objective, namely to sustain itself and keep their hosts happy in as much as possible. If a handful of data scientists had their entries removed, who cares? The host still get what they want: solutions from hundreds of other developers to their problem. In any relationship, you will get hurt. It's an occupational hazard. Here on Kaggle, many of the greats had experiences where they were burned by either a cheating team mate, burned by Kaggle, or even burned by a competition host directly. The question is what are you going to do after that / how will you deal with it? I've decided to keep Kaggling because I have some specific goals I intend on accomplishing (I use Kaggle the same way they 'use' me) but I won't sing their praises.
\nThe one thing that makes the burn hurt is that as time goes by, it gets harder to level up. If this situation happened 2 years ago, maybe… no big deal? But it happening now causes more pain because getting a silver or top 50th position only gets tougher as better technology, more contestants, and stronger practitioners pile onto the platform.
", "rawMarkdown": "Personally, I gave up. It's way past any reasonable period where they're going to do anything about it.\n\nIn this case, Kaggle executed very poorly in terms of getting this code competition finalized. I don't ever recall seeing final scoring delayed this long before. The lag resulted in a myriad of discrepancies due to people altering the status (private/public, versioned/unversioned, available/deleted) of their model datasets during the limbo period post closure, whereas under normal competition configuration, once the competition ended, nothing would have mattered.\n\nEven though they failed hard here, they are making some changes. For example, I noticed the submission screen for the bird and the lyft code competitions have some dataset versioning added to it where that did not exist before. People can still get screwed hard if they delete their models after the competition ends, and from Inversion's post, they won't be altering that for some reason(?). Kaggle also failed hard here with their lack of communication to us, their users. Canned responses, if any, and very delayed. All quite disheartening.\n\nFor me, this was a good wake up call. Kaggle has been instrumental to motivating me to become a better data scientist. But at the same token, it's important for me not to put it on a pedestal and to recognize it's just like any other company. It has an objective, namely to sustain itself and keep their hosts happy in as much as possible. If a handful of data scientists had their entries removed, who cares? The host still get what they want: solutions from hundreds of other developers to their problem. In any relationship, you will get hurt. It's an occupational hazard. Here on Kaggle, many of the greats had experiences where they were burned by either a cheating team mate, burned by Kaggle, or even burned by a competition host directly. The question is what are you going to do after that / how will you deal with it? I've decided to keep Kaggling because I have some specific goals I intend on accomplishing (I use Kaggle the same way they 'use' me) but I won't sing their praises.\n\nThe one thing that makes the burn hurt is that as time goes by, it gets harder to level up. If this situation happened 2 years ago, maybe... no big deal? But it happening now causes more pain because getting a silver or top 50th position only gets tougher as better technology, more contestants, and stronger practitioners pile onto the platform.", "votes": null }, { "id": "992330", "postDate": "08/31/2020 05:32:01", "content": "There are several things that can happen. In some cases if a dataset that the kernel relies on to run is removed then the kernel will not run. In other cases there were issues with the code that did not allow the kernel to run completely. In those cases the people who were removed can look and see there there was an error message with their code. We did not actively remove those referenced, their kernel did not generate a score so there was nothing to display on the leaderboard.
\nWe do take everyone's work very seriously and we work really hard to set-up competitions but competitions are very complex and sometimes things happen that are beyond our control. Thank you for your understanding.
", "rawMarkdown": "There are several things that can happen. In some cases if a dataset that the kernel relies on to run is removed then the kernel will not run. In other cases there were issues with the code that did not allow the kernel to run completely. In those cases the people who were removed can look and see there there was an error message with their code. We did not actively remove those referenced, their kernel did not generate a score so there was nothing to display on the leaderboard. \n\nWe do take everyone's work very seriously and we work really hard to set-up competitions but competitions are very complex and sometimes things happen that are beyond our control. Thank you for your understanding.", "votes": null }, { "id": "993535", "postDate": "09/01/2020 02:17:13", "content": "Ok.. Thanks for reply firstly @maggiemd .But our situation is a little different here: we didn't delete datasets, we just adjusted them from public to private after competition. Is that also a problem? I don't think we have other errors, because our kernel runs successfully with late submit without any change(even dataset).
\nSo you mean that if we are wrongly removed, we won't get any compensation, right?🤕
", "rawMarkdown": "Ok.. Thanks for reply firstly @maggiemd .But our situation is a little different here: we didn't delete datasets, we just adjusted them from public to private after competition. Is that also a problem? I don't think we have other errors, because our kernel runs successfully with late submit without any change(even dataset).\n\nSo you mean that if we are wrongly removed, we won't get any compensation, right?🤕", "votes": null }, { "id": "993538", "postDate": "09/01/2020 02:19:37", "content": "I’m almost give up too.. That's so sad.
\nThe reason I continue to post in this competition is that I don't want this situation to happen in the future again.
@philculliton @maggiemd @bendvd @stavness
", "votes": null, "replies": [] }, { "id": 991024, "author_name": "marcowu", "author_url": "", "post_date": "08/30/2020 04:24:14", "content": "@philculliton @maggiemd @bendvd @stavness Please help us!!
", "votes": null, "replies": [] }, { "id": 992223, "author_name": "authman", "author_url": "", "post_date": "08/31/2020 04:06:05", "content": "Personally, I gave up. It's way past any reasonable period where they're going to do anything about it.
\nIn this case, Kaggle executed very poorly in terms of getting this code competition finalized. I don't ever recall seeing final scoring delayed this long before. The lag resulted in a myriad of discrepancies due to people altering the status (private/public, versioned/unversioned, available/deleted) of their model datasets during the limbo period post closure, whereas under normal competition configuration, once the competition ended, nothing would have mattered.
\nEven though they failed hard here, they are making some changes. For example, I noticed the submission screen for the bird and the lyft code competitions have some dataset versioning added to it where that did not exist before. People can still get screwed hard if they delete their models after the competition ends, and from Inversion's post, they won't be altering that for some reason(?). Kaggle also failed hard here with their lack of communication to us, their users. Canned responses, if any, and very delayed. All quite disheartening.
\nFor me, this was a good wake up call. Kaggle has been instrumental to motivating me to become a better data scientist. But at the same token, it's important for me not to put it on a pedestal and to recognize it's just like any other company. It has an objective, namely to sustain itself and keep their hosts happy in as much as possible. If a handful of data scientists had their entries removed, who cares? The host still get what they want: solutions from hundreds of other developers to their problem. In any relationship, you will get hurt. It's an occupational hazard. Here on Kaggle, many of the greats had experiences where they were burned by either a cheating team mate, burned by Kaggle, or even burned by a competition host directly. The question is what are you going to do after that / how will you deal with it? I've decided to keep Kaggling because I have some specific goals I intend on accomplishing (I use Kaggle the same way they 'use' me) but I won't sing their praises.
\nThe one thing that makes the burn hurt is that as time goes by, it gets harder to level up. If this situation happened 2 years ago, maybe… no big deal? But it happening now causes more pain because getting a silver or top 50th position only gets tougher as better technology, more contestants, and stronger practitioners pile onto the platform.
", "votes": null, "replies": [ { "id": 993538, "author_name": "li325040229", "author_url": "", "post_date": "09/01/2020 02:19:37", "content": "I’m almost give up too.. That's so sad.
\nThe reason I continue to post in this competition is that I don't want this situation to happen in the future again.
There are several things that can happen. In some cases if a dataset that the kernel relies on to run is removed then the kernel will not run. In other cases there were issues with the code that did not allow the kernel to run completely. In those cases the people who were removed can look and see there there was an error message with their code. We did not actively remove those referenced, their kernel did not generate a score so there was nothing to display on the leaderboard.
\nWe do take everyone's work very seriously and we work really hard to set-up competitions but competitions are very complex and sometimes things happen that are beyond our control. Thank you for your understanding.
", "votes": null, "replies": [ { "id": 993535, "author_name": "li325040229", "author_url": "", "post_date": "09/01/2020 02:17:13", "content": "Ok.. Thanks for reply firstly @maggiemd .But our situation is a little different here: we didn't delete datasets, we just adjusted them from public to private after competition. Is that also a problem? I don't think we have other errors, because our kernel runs successfully with late submit without any change(even dataset).
\nSo you mean that if we are wrongly removed, we won't get any compensation, right?🤕
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "987445": "The game has ended for 8 days, and the discussion has almost quieted down. \n\nFirst of all, **congratulations to all the kagglers who won medals**. However, there are still some people who have been mistakenly removed from the leaderboard and still haven't received any proper response. \n\nTake our team as an example. We are removed from leaderboard just because we made one of our additional data sets public for a period of time after the game and then changed it to private. In the subsequent late submission, our rerun was successful and in compliance with the rules.\n\nFor some other teams, it may because they deleted the dataset they had uploaded. And based on the communication with @authman ,\n> I also just now received the same email verbatim, sans the team name.\n\nwe guess kaggle gave the same response even we are in a such different situations. **Kaggle did not made any other detailed responses to us so far**, including what will they do, or just do nothing. In the forum, they only responded to Aidyn Ubingazhibov’s post, which received over 200 upvote.\n\nI sincerely thank the kaggle team for their dedication in this competition, including organizing this game and spending a lot of energy to clean up non-compliant submissions. **But for me, it is totally unacceptable to be removed a precious medal without doing anything that obey the rules. If the kaggle team needs any evidence from team, please feel free to contact us.** \n\nI will keep attention to the results of this matter! I also hope that everyone can give us some support, because each of us has an obligation to promote the perfection of the rules. Only in this way can we avoid the same thing happening to ourselves in the future.", "987685": "philculliton @maggiemd @bendvd @stavness", "991024": "philculliton @maggiemd @bendvd @stavness Please help us!!", "992223": "Personally, I gave up. It's way past any reasonable period where they're going to do anything about it.\n\nIn this case, Kaggle executed very poorly in terms of getting this code competition finalized. I don't ever recall seeing final scoring delayed this long before. The lag resulted in a myriad of discrepancies due to people altering the status (private/public, versioned/unversioned, available/deleted) of their model datasets during the limbo period post closure, whereas under normal competition configuration, once the competition ended, nothing would have mattered.\n\nEven though they failed hard here, they are making some changes. For example, I noticed the submission screen for the bird and the lyft code competitions have some dataset versioning added to it where that did not exist before. People can still get screwed hard if they delete their models after the competition ends, and from Inversion's post, they won't be altering that for some reason(?). Kaggle also failed hard here with their lack of communication to us, their users. Canned responses, if any, and very delayed. All quite disheartening.\n\nFor me, this was a good wake up call. Kaggle has been instrumental to motivating me to become a better data scientist. But at the same token, it's important for me not to put it on a pedestal and to recognize it's just like any other company. It has an objective, namely to sustain itself and keep their hosts happy in as much as possible. If a handful of data scientists had their entries removed, who cares? The host still get what they want: solutions from hundreds of other developers to their problem. In any relationship, you will get hurt. It's an occupational hazard. Here on Kaggle, many of the greats had experiences where they were burned by either a cheating team mate, burned by Kaggle, or even burned by a competition host directly. The question is what are you going to do after that / how will you deal with it? I've decided to keep Kaggling because I have some specific goals I intend on accomplishing (I use Kaggle the same way they 'use' me) but I won't sing their praises.\n\nThe one thing that makes the burn hurt is that as time goes by, it gets harder to level up. If this situation happened 2 years ago, maybe... no big deal? But it happening now causes more pain because getting a silver or top 50th position only gets tougher as better technology, more contestants, and stronger practitioners pile onto the platform.", "992330": "There are several things that can happen. In some cases if a dataset that the kernel relies on to run is removed then the kernel will not run. In other cases there were issues with the code that did not allow the kernel to run completely. In those cases the people who were removed can look and see there there was an error message with their code. We did not actively remove those referenced, their kernel did not generate a score so there was nothing to display on the leaderboard. \n\nWe do take everyone's work very seriously and we work really hard to set-up competitions but competitions are very complex and sometimes things happen that are beyond our control. Thank you for your understanding.", "993535": "Ok.. Thanks for reply firstly @maggiemd .But our situation is a little different here: we didn't delete datasets, we just adjusted them from public to private after competition. Is that also a problem? I don't think we have other errors, because our kernel runs successfully with late submit without any change(even dataset).\n\nSo you mean that if we are wrongly removed, we won't get any compensation, right?🤕", "993538": "I’m almost give up too.. That's so sad.\nThe reason I continue to post in this competition is that I don't want this situation to happen in the future again." }, "source": "meta" }