{ "id": 384482, "title": "So cheater GrandMaster Survive? Great job", "url": "/competitions/otto-recommender-system/discussion/384482", "author_name": "", "post_date": "2023-02-08T03:45:09.045461200Z", "votes": 48, "comment_count": 37, "views": 0, "content": "

Congratulations

", "messages": [ { "id": "2134516", "postDate": "02/08/2023 03:45:09", "content": "

Congratulations

", "rawMarkdown": "Congratulations", "votes": null }, { "id": "2134817", "postDate": "02/08/2023 09:17:40", "content": "

As the title of this thread says, LB was finalized with suspected cheaters remaining, despite the issue being raised in the following discussions.

\n\n

Whether they are banned or not does not change the color of our team medal. But how about the teams on boarderlines? I feel so sorry for them.

\n

And if there is no explanation or action on this issue, this is good news not only for the users who may have cheated this time, but also for those who may cheat in future competitions.

\n

For players who have done the hard work, those who were brave enough to raise the issue, and the value of kaggle competitions, I hope that at least kaggle team will provide some explanation for this issue.

", "rawMarkdown": "As the title of this thread says, LB was finalized with suspected cheaters remaining, despite the issue being raised in the following discussions.\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381318\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381360\n\nWhether they are banned or not does not change the color of our team medal. But how about the teams on boarderlines? I feel so sorry for them.\n\nAnd if there is no explanation or action on this issue, this is good news not only for the users who may have cheated this time, but also for those who may cheat in future competitions.\n\nFor players who have done the hard work, those who were brave enough to raise the issue, and the value of kaggle competitions, I hope that at least kaggle team will provide some explanation for this issue.", "votes": null }, { "id": "2134836", "postDate": "02/08/2023 09:28:04", "content": "

I second this.
\nCheaters not being punished is really demotivating for people that compete honestly, and opens the door for more abuse.
\nThis is also really bad for Kaggle's and image. The competition was really good, it's a shame that people will remember OTTO as the competition where private sharing was rewarded.

\n

One of the cheaters jumped to rank 16 in the overall competitions rankings, and the other one got competition GM.

\n

Poke @inversion

", "rawMarkdown": "I second this.\nCheaters not being punished is really demotivating for people that compete honestly, and opens the door for more abuse. \nThis is also really bad for Kaggle's and image. The competition was really good, it's a shame that people will remember OTTO as the competition where private sharing was rewarded.\n\nOne of the cheaters jumped to rank 16 in the overall competitions rankings, and the other one got competition GM. \n\nPoke @inversion", "votes": null }, { "id": "2134909", "postDate": "02/08/2023 10:25:38", "content": "

This reminds me this
\nhttps://www.kaggle.com/competitions/siim-acr-pneumothorax-segmentation/discussion/97499

", "rawMarkdown": "This reminds me this\nhttps://www.kaggle.com/competitions/siim-acr-pneumothorax-segmentation/discussion/97499", "votes": null }, { "id": "2134929", "postDate": "02/08/2023 10:50:13", "content": "

After 2 years of spending countless hours on Kaggle, seems I was doing it all wrong 🤦‍♂️ there is a better way to becoming GM. 🤷‍♂️

", "rawMarkdown": "After 2 years of spending countless hours on Kaggle, seems I was doing it all wrong 🤦‍♂️ there is a better way to becoming GM. 🤷‍♂️", "votes": null }, { "id": "2134962", "postDate": "02/08/2023 11:06:58", "content": "

I'm disappointed at the decision of Kaggle staffs.
\nStealing the submission from other team is more deserved of the gold model than 16th place solution ?

", "rawMarkdown": "I'm disappointed at the decision of Kaggle staffs.\nStealing the submission from other team is more deserved of the gold model than 16th place solution ?", "votes": null }, { "id": "2135014", "postDate": "02/08/2023 11:57:09", "content": "

It's more comfortable to send a few emails than to spend money on renting a server and thinking about how to improve day and night. SAD.

", "rawMarkdown": "It's more comfortable to send a few emails than to spend money on renting a server and thinking about how to improve day and night. SAD.", "votes": null }, { "id": "2135255", "postDate": "02/08/2023 14:30:48", "content": "

I just got potential winners email from kaggle. It means they decided not to remove them.

", "rawMarkdown": "I just got potential winners email from kaggle. It means they decided not to remove them.", "votes": null }, { "id": "2135297", "postDate": "02/08/2023 14:52:10", "content": "

I thought I'd weigh in on the decision here. First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed. Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks. Appropriate actions will be taken when sufficient evidence is obtained. Lastly, yes, there are are various activities that look suspicious, but these don't necessarily mean a team has cheated. We do take action when we believe there is sufficient evidence to do so.

", "rawMarkdown": "I thought I'd weigh in on the decision here. First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed. Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks. Appropriate actions will be taken when sufficient evidence is obtained. Lastly, yes, there are are various activities that look suspicious, but these don't necessarily mean a team has cheated. We do take action when we believe there is sufficient evidence to do so.", "votes": null }, { "id": "2135304", "postDate": "02/08/2023 14:53:30", "content": "

I can assure you that while there are cheaters that fly under the radar for some time, there does come a time when there is sufficient for action to be taken.

", "rawMarkdown": "I can assure you that while there are cheaters that fly under the radar for some time, there does come a time when there is sufficient for action to be taken.", "votes": null }, { "id": "2135403", "postDate": "02/08/2023 16:13:25", "content": "

@inversion
\nWhen?
\nAfter they retire from their well paid job obtained based on their fake Kaggle GM ?

", "rawMarkdown": "inversion \nWhen?\nAfter they retire from their well paid job obtained based on their fake Kaggle GM ?", "votes": null }, { "id": "2135637", "postDate": "02/08/2023 19:01:35", "content": "

@inversion

\n
\n

Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks.

\n
\n

Is it safe to assume that this will be reinvestigated for past competitions as well?
\nThe allegations could affect not only OTTO participants, but also participants from several past competitions(not only kaggle).

\n

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321

\n
\n

We do take action when we believe there is sufficient evidence to do so.

\n
\n

Apparently someone was accidentally banned once, was there sufficient evidence at the time? I demand an explanation.

\n

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2118580

", "rawMarkdown": "inversion \n\n> Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks.\n\nIs it safe to assume that this will be reinvestigated for past competitions as well?\nThe allegations could affect not only OTTO participants, but also participants from several past competitions(not only kaggle).\n\nhttps://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n\n> We do take action when we believe there is sufficient evidence to do so.\n\nApparently someone was accidentally banned once, was there sufficient evidence at the time? I demand an explanation.\n\nhttps://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2118580", "votes": null }, { "id": "2135692", "postDate": "02/08/2023 20:06:30", "content": "

It is quite disappointing that kaggle did not post an official message even a week after the competition. Feels that the team is not taking this incident seriously. (If you're not in the prize zone, you can easily get away with cheating?)

\n

I think a lot of people would be walking out of the platform if \"appropriate\" action is not taken. It feels GM and gold medals reputation has sunk.

", "rawMarkdown": "It is quite disappointing that kaggle did not post an official message even a week after the competition. Feels that the team is not taking this incident seriously. (If you're not in the prize zone, you can easily get away with cheating?)\n\nI think a lot of people would be walking out of the platform if \"appropriate\" action is not taken. It feels GM and gold medals reputation has sunk.", "votes": null }, { "id": "2135698", "postDate": "02/08/2023 20:11:16", "content": "

meanwhile you won your gold having @poteman in team. You could leave a comment here about that experience at least :)

", "rawMarkdown": "meanwhile you won your gold having @poteman in team. You could leave a comment here about that experience at least :)", "votes": null }, { "id": "2135705", "postDate": "02/08/2023 20:18:21", "content": "

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2116033
\nI've wrote it here

", "rawMarkdown": "https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2116033\nI've wrote it here", "votes": null }, { "id": "2135747", "postDate": "02/08/2023 21:28:21", "content": "

Now he knows crime pays, and he'll keep on cheating.

\n

Really sad that he is still on kaggle.

\n

But I'm happy that his inocent team members were not punished. They did deserve the medal and prize.

", "rawMarkdown": "Now he knows crime pays, and he'll keep on cheating.\n\nReally sad that he is still on kaggle.\n\nBut I'm happy that his inocent team members were not punished. They did deserve the medal and prize.", "votes": null }, { "id": "2135823", "postDate": "02/08/2023 22:06:56", "content": "

@inversion

\n
\n

First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed.

\n
\n

Have you read the posts about suspicious cheating activity on this competition^1, ^2, ^3? (If you want to have a grasp of this issue first, you can refer to a sum-up of this issue^4.)

\n

We are not talking about a cheating that all of teammates are cheaters, but a cheater hiding in a team, and stealing their team-mate's solution, then shares the solutions to his cheater's pals who is also hiding in other innocent team. So above cheater screening process do not work for this case. Because their (innocent) team mate do have their works.

\n

If these reports are true, the suspected two users are highly possible to cheat using the same scheme repeatedly to many teams and competitions over years (and this cheat made them Grandmaster and ranked high on total score). If Kaggle do nothing about them, they will continue to cheat for the future.

\n

It is not only the issue of this single competition, this is an issue of many past and future competitions.

\n

I hope Kaggle will take this issue properly.

", "rawMarkdown": "inversion \n\n> First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed.\n\nHave you read the posts about suspicious cheating activity on this competition[^1], [^2], [^3]? (If you want to have a grasp of this issue first, you can refer to a sum-up of this issue[^4].)\n\nWe are not talking about a cheating that all of teammates are cheaters, but a cheater hiding in a team, and stealing their team-mate's solution, then shares the solutions to his cheater's pals who is also hiding in other innocent team. So above cheater screening process do not work for this case. Because their (innocent) team mate do have their works.\n\nIf these reports are true, the suspected two users are highly possible to cheat using the same scheme repeatedly to many teams and competitions over years (and this cheat made them Grandmaster and ranked high on total score). If Kaggle do nothing about them, they will continue to cheat for the future. \n\nIt is not only the issue of this single competition, this is an issue of many past and future competitions.\n\nI hope Kaggle will take this issue properly.\n\n[^1]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381318\n[^2]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n[^3]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381360\n[^4]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381527", "votes": null }, { "id": "2135878", "postDate": "02/08/2023 23:22:39", "content": "

Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.

\n

However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.

\n

I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.

", "rawMarkdown": "Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.\n\nHowever, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.\n\nI strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.", "votes": null }, { "id": "2135883", "postDate": "02/08/2023 23:34:13", "content": "

Agreed. It is just baffling that there is no official statement and someone had to make a post so we can all have a clue of what is happening.

", "rawMarkdown": "Agreed. It is just baffling that there is no official statement and someone had to make a post so we can all have a clue of what is happening.", "votes": null }, { "id": "2135885", "postDate": "02/08/2023 23:36:45", "content": "

And after they have created many more fake Masters or GMs

", "rawMarkdown": "And after they have created many more fake Masters or GMs", "votes": null }, { "id": "2135913", "postDate": "02/09/2023 00:35:26", "content": "

I am still thinking Kaggle team is taking actions and need more time for investigations.

", "rawMarkdown": "I am still thinking Kaggle team is taking actions and need more time for investigations.", "votes": null }, { "id": "2136029", "postDate": "02/09/2023 03:53:43", "content": "

@poteman and other cheaters keep up your great work! Kaggle recognizes your great job! :D

", "rawMarkdown": "poteman and other cheaters keep up your great work! Kaggle recognizes your great job! :D", "votes": null }, { "id": "2137327", "postDate": "02/09/2023 21:25:25", "content": "

you were right :D

", "rawMarkdown": "you were right :D", "votes": null }, { "id": "2137351", "postDate": "02/09/2023 22:18:32", "content": "
\n

Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.

\n

However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.

\n

I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.

\n
\n

How did people cheat?

", "rawMarkdown": "> Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.\n> \n> However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.\n> \n> I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.\n\nHow did people cheat?", "votes": null }, { "id": "2137359", "postDate": "02/09/2023 22:33:54", "content": "

Hi @thrill1, welcome to Kaggle as you join the platform 2 hours ago, just after the accounts of the 2 cheaters were terminated.
\nOr should we call you poteman2 or RunningZ2 ?
\nUnfortunately it is so easy to create a new account.

", "rawMarkdown": "Hi @thrill1, welcome to Kaggle as you join the platform 2 hours ago, just after the accounts of the 2 cheaters were terminated.\nOr should we call you poteman2 or RunningZ2 ?\nUnfortunately it is so easy to create a new account.", "votes": null }, { "id": "2137382", "postDate": "02/09/2023 23:48:17", "content": "

Update:

\n

I'm glad to hear about my solo Gold and GM promotion from Kaggle's recent judgment update!
\nI think it's a great move for the overall health of the Kaggle community to eliminate teams with cheating suspicions. And, my heart goes out to those innocent participants who unintentionally merged with cheaters.

", "rawMarkdown": "Update:\n\nI'm glad to hear about my solo Gold and GM promotion from Kaggle's recent judgment update! \nI think it's a great move for the overall health of the Kaggle community to eliminate teams with cheating suspicions. And, my heart goes out to those innocent participants who unintentionally merged with cheaters.", "votes": null }, { "id": "2137384", "postDate": "02/09/2023 23:53:44", "content": "

Congrats! Pay high tribute to these innocent participants.

", "rawMarkdown": "Congrats! Pay high tribute to these innocent participants.", "votes": null }, { "id": "2137385", "postDate": "02/09/2023 23:56:53", "content": "

Sorry, I am new to the platform. They asked me to comment in the checklist and this was an active discussion. Just learning about Kaggle and thinking about posting a competition for my company.

", "rawMarkdown": "Sorry, I am new to the platform. They asked me to comment in the checklist and this was an active discussion. Just learning about Kaggle and thinking about posting a competition for my company.", "votes": null }, { "id": "2137394", "postDate": "02/10/2023 00:09:02", "content": "

If poteman is not native english speaker, it does not mean that he is a cheater. The reproach about the lack of response from poteman ----Okey---Kaggle has been changed. I see it.

", "rawMarkdown": "If poteman is not native english speaker, it does not mean that he is a cheater. The reproach about the lack of response from poteman ----Okey---Kaggle has been changed. I see it.", "votes": null }, { "id": "2137398", "postDate": "02/10/2023 00:16:04", "content": "

Are you another account of poteman?

", "rawMarkdown": "Are you another account of poteman?", "votes": null }, { "id": "2137399", "postDate": "02/10/2023 00:21:34", "content": "

Of course, I'm poteman. You immediately recogranise it, because you are the expert here!

", "rawMarkdown": "Of course, I'm poteman. You immediately recogranise it, because you are the expert here!", "votes": null }, { "id": "2137411", "postDate": "02/10/2023 01:02:58", "content": "

Congratulations~

", "rawMarkdown": "Congratulations~", "votes": null }, { "id": "2137479", "postDate": "02/10/2023 03:48:34", "content": "

Congrats!!

", "rawMarkdown": "Congrats!!", "votes": null }, { "id": "2139719", "postDate": "02/11/2023 04:36:56", "content": "

I will tell u my story. I have been using Kaggle for the past 3 years and I have forked many notebooks for practice. I will never make a forked notebook public and showcase as my work. I have this consciousness of not to get creds for others' work! However, I recently thought of participating in a competition, I completed the work and saw that I got placed in the last 100s. I was a bit disappointed yet when I see other's solutions I could see at least 10 people had the same solution WITHOUT EVEN CHANGING THE COMMENTS! and they were placed in the top 10%. Kaggle should take serious steps about this code plagiarism and encourage upcoming enthusiasts to get recognized for their work. I may be at the last, but I can happily say that the work was completed by me!

\n

Kudos to all the enthusiasts like me who are grinding hard and publishing your own work. Lets make it to grand master one day!

", "rawMarkdown": "I will tell u my story. I have been using Kaggle for the past 3 years and I have forked many notebooks for practice. I will never make a forked notebook public and showcase as my work. I have this consciousness of not to get creds for others' work! However, I recently thought of participating in a competition, I completed the work and saw that I got placed in the last 100s. I was a bit disappointed yet when I see other's solutions I could see at least 10 people had the same solution WITHOUT EVEN CHANGING THE COMMENTS! and they were placed in the top 10%. Kaggle should take serious steps about this code plagiarism and encourage upcoming enthusiasts to get recognized for their work. I may be at the last, but I can happily say that the work was completed by me! \n\nKudos to all the enthusiasts like me who are grinding hard and publishing your own work. Lets make it to grand master one day!", "votes": null }, { "id": "2139880", "postDate": "02/11/2023 08:55:18", "content": "

Ok this is great

", "rawMarkdown": "Ok this is great", "votes": null }, { "id": "2140462", "postDate": "02/11/2023 19:02:44", "content": "

I am a newcomer to Kaggle. Please describe more!

", "rawMarkdown": "I am a newcomer to Kaggle. Please describe more!", "votes": null }, { "id": "2142122", "postDate": "02/13/2023 10:29:06", "content": "

Yeah it's unfortunate. If cheating goes unchecked, it undermines the hard work of fair players and raises doubts about the authenticity of competition outcomes. This can be discouraging for those who compete with honesty and integrity.

", "rawMarkdown": "Yeah it's unfortunate. If cheating goes unchecked, it undermines the hard work of fair players and raises doubts about the authenticity of competition outcomes. This can be discouraging for those who compete with honesty and integrity.", "votes": null }, { "id": "2143790", "postDate": "02/14/2023 14:05:58", "content": "

I am a newcomer also!

", "rawMarkdown": "I am a newcomer also!", "votes": null } ], "comments": [ { "id": 2134817, "author_name": "chiakiichimura", "author_url": "", "post_date": "02/08/2023 09:17:40", "content": "

As the title of this thread says, LB was finalized with suspected cheaters remaining, despite the issue being raised in the following discussions.

\n\n

Whether they are banned or not does not change the color of our team medal. But how about the teams on boarderlines? I feel so sorry for them.

\n

And if there is no explanation or action on this issue, this is good news not only for the users who may have cheated this time, but also for those who may cheat in future competitions.

\n

For players who have done the hard work, those who were brave enough to raise the issue, and the value of kaggle competitions, I hope that at least kaggle team will provide some explanation for this issue.

", "votes": null, "replies": [ { "id": 2134836, "author_name": "theoviel", "author_url": "", "post_date": "02/08/2023 09:28:04", "content": "

I second this.
\nCheaters not being punished is really demotivating for people that compete honestly, and opens the door for more abuse.
\nThis is also really bad for Kaggle's and image. The competition was really good, it's a shame that people will remember OTTO as the competition where private sharing was rewarded.

\n

One of the cheaters jumped to rank 16 in the overall competitions rankings, and the other one got competition GM.

\n

Poke @inversion

", "votes": null, "replies": [ { "id": 2135304, "author_name": "inversion", "author_url": "", "post_date": "02/08/2023 14:53:30", "content": "

I can assure you that while there are cheaters that fly under the radar for some time, there does come a time when there is sufficient for action to be taken.

", "votes": null, "replies": [ { "id": 2135403, "author_name": "blankaf", "author_url": "", "post_date": "02/08/2023 16:13:25", "content": "

@inversion
\nWhen?
\nAfter they retire from their well paid job obtained based on their fake Kaggle GM ?

", "votes": null, "replies": [ { "id": 2135885, "author_name": "hinepo", "author_url": "", "post_date": "02/08/2023 23:36:45", "content": "

And after they have created many more fake Masters or GMs

", "votes": null, "replies": [] } ] } ] } ] } ] }, { "id": 2134909, "author_name": "onodera", "author_url": "", "post_date": "02/08/2023 10:25:38", "content": "

This reminds me this
\nhttps://www.kaggle.com/competitions/siim-acr-pneumothorax-segmentation/discussion/97499

", "votes": null, "replies": [] }, { "id": 2134929, "author_name": "crodoc", "author_url": "", "post_date": "02/08/2023 10:50:13", "content": "

After 2 years of spending countless hours on Kaggle, seems I was doing it all wrong 🤦‍♂️ there is a better way to becoming GM. 🤷‍♂️

", "votes": null, "replies": [ { "id": 2140462, "author_name": "alirezasalehan", "author_url": "", "post_date": "02/11/2023 19:02:44", "content": "

I am a newcomer to Kaggle. Please describe more!

", "votes": null, "replies": [] }, { "id": 2143790, "author_name": "xyxowen123navercom", "author_url": "", "post_date": "02/14/2023 14:05:58", "content": "

I am a newcomer also!

", "votes": null, "replies": [] } ] }, { "id": 2134962, "author_name": "toshik", "author_url": "", "post_date": "02/08/2023 11:06:58", "content": "

I'm disappointed at the decision of Kaggle staffs.
\nStealing the submission from other team is more deserved of the gold model than 16th place solution ?

", "votes": null, "replies": [] }, { "id": 2135014, "author_name": "hengzheng", "author_url": "", "post_date": "02/08/2023 11:57:09", "content": "

It's more comfortable to send a few emails than to spend money on renting a server and thinking about how to improve day and night. SAD.

", "votes": null, "replies": [] }, { "id": 2135255, "author_name": "onodera", "author_url": "", "post_date": "02/08/2023 14:30:48", "content": "

I just got potential winners email from kaggle. It means they decided not to remove them.

", "votes": null, "replies": [] }, { "id": 2135297, "author_name": "inversion", "author_url": "", "post_date": "02/08/2023 14:52:10", "content": "

I thought I'd weigh in on the decision here. First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed. Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks. Appropriate actions will be taken when sufficient evidence is obtained. Lastly, yes, there are are various activities that look suspicious, but these don't necessarily mean a team has cheated. We do take action when we believe there is sufficient evidence to do so.

", "votes": null, "replies": [ { "id": 2135637, "author_name": "yukkyo", "author_url": "", "post_date": "02/08/2023 19:01:35", "content": "

@inversion

\n
\n

Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks.

\n
\n

Is it safe to assume that this will be reinvestigated for past competitions as well?
\nThe allegations could affect not only OTTO participants, but also participants from several past competitions(not only kaggle).

\n

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321

\n
\n

We do take action when we believe there is sufficient evidence to do so.

\n
\n

Apparently someone was accidentally banned once, was there sufficient evidence at the time? I demand an explanation.

\n

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2118580

", "votes": null, "replies": [] }, { "id": 2135692, "author_name": "kyoshioka47", "author_url": "", "post_date": "02/08/2023 20:06:30", "content": "

It is quite disappointing that kaggle did not post an official message even a week after the competition. Feels that the team is not taking this incident seriously. (If you're not in the prize zone, you can easily get away with cheating?)

\n

I think a lot of people would be walking out of the platform if \"appropriate\" action is not taken. It feels GM and gold medals reputation has sunk.

", "votes": null, "replies": [ { "id": 2135698, "author_name": "kvlmll", "author_url": "", "post_date": "02/08/2023 20:11:16", "content": "

meanwhile you won your gold having @poteman in team. You could leave a comment here about that experience at least :)

", "votes": null, "replies": [ { "id": 2135705, "author_name": "kyoshioka47", "author_url": "", "post_date": "02/08/2023 20:18:21", "content": "

https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2116033
\nI've wrote it here

", "votes": null, "replies": [] }, { "id": 2136029, "author_name": "projdev", "author_url": "", "post_date": "02/09/2023 03:53:43", "content": "

@poteman and other cheaters keep up your great work! Kaggle recognizes your great job! :D

", "votes": null, "replies": [] } ] } ] }, { "id": 2135823, "author_name": "tatamikenn", "author_url": "", "post_date": "02/08/2023 22:06:56", "content": "

@inversion

\n
\n

First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed.

\n
\n

Have you read the posts about suspicious cheating activity on this competition^1, ^2, ^3? (If you want to have a grasp of this issue first, you can refer to a sum-up of this issue^4.)

\n

We are not talking about a cheating that all of teammates are cheaters, but a cheater hiding in a team, and stealing their team-mate's solution, then shares the solutions to his cheater's pals who is also hiding in other innocent team. So above cheater screening process do not work for this case. Because their (innocent) team mate do have their works.

\n

If these reports are true, the suspected two users are highly possible to cheat using the same scheme repeatedly to many teams and competitions over years (and this cheat made them Grandmaster and ranked high on total score). If Kaggle do nothing about them, they will continue to cheat for the future.

\n

It is not only the issue of this single competition, this is an issue of many past and future competitions.

\n

I hope Kaggle will take this issue properly.

", "votes": null, "replies": [] } ] }, { "id": 2135747, "author_name": "hinepo", "author_url": "", "post_date": "02/08/2023 21:28:21", "content": "

Now he knows crime pays, and he'll keep on cheating.

\n

Really sad that he is still on kaggle.

\n

But I'm happy that his inocent team members were not punished. They did deserve the medal and prize.

", "votes": null, "replies": [] }, { "id": 2135878, "author_name": "tommy1028", "author_url": "", "post_date": "02/08/2023 23:22:39", "content": "

Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.

\n

However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.

\n

I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.

", "votes": null, "replies": [ { "id": 2135883, "author_name": "hinepo", "author_url": "", "post_date": "02/08/2023 23:34:13", "content": "

Agreed. It is just baffling that there is no official statement and someone had to make a post so we can all have a clue of what is happening.

", "votes": null, "replies": [] }, { "id": 2137351, "author_name": "thrill1", "author_url": "", "post_date": "02/09/2023 22:18:32", "content": "
\n

Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.

\n

However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.

\n

I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.

\n
\n

How did people cheat?

", "votes": null, "replies": [ { "id": 2137359, "author_name": "blankaf", "author_url": "", "post_date": "02/09/2023 22:33:54", "content": "

Hi @thrill1, welcome to Kaggle as you join the platform 2 hours ago, just after the accounts of the 2 cheaters were terminated.
\nOr should we call you poteman2 or RunningZ2 ?
\nUnfortunately it is so easy to create a new account.

", "votes": null, "replies": [ { "id": 2137385, "author_name": "thrill1", "author_url": "", "post_date": "02/09/2023 23:56:53", "content": "

Sorry, I am new to the platform. They asked me to comment in the checklist and this was an active discussion. Just learning about Kaggle and thinking about posting a competition for my company.

", "votes": null, "replies": [] } ] } ] }, { "id": 2137382, "author_name": "tommy1028", "author_url": "", "post_date": "02/09/2023 23:48:17", "content": "

Update:

\n

I'm glad to hear about my solo Gold and GM promotion from Kaggle's recent judgment update!
\nI think it's a great move for the overall health of the Kaggle community to eliminate teams with cheating suspicions. And, my heart goes out to those innocent participants who unintentionally merged with cheaters.

", "votes": null, "replies": [ { "id": 2137384, "author_name": "hengzheng", "author_url": "", "post_date": "02/09/2023 23:53:44", "content": "

Congrats! Pay high tribute to these innocent participants.

", "votes": null, "replies": [] }, { "id": 2137411, "author_name": "earlee0412", "author_url": "", "post_date": "02/10/2023 01:02:58", "content": "

Congratulations~

", "votes": null, "replies": [] }, { "id": 2137479, "author_name": "junaries", "author_url": "", "post_date": "02/10/2023 03:48:34", "content": "

Congrats!!

", "votes": null, "replies": [] } ] }, { "id": 2142122, "author_name": "carolineignore", "author_url": "", "post_date": "02/13/2023 10:29:06", "content": "

Yeah it's unfortunate. If cheating goes unchecked, it undermines the hard work of fair players and raises doubts about the authenticity of competition outcomes. This can be discouraging for those who compete with honesty and integrity.

", "votes": null, "replies": [] } ] }, { "id": 2135913, "author_name": "hydantess", "author_url": "", "post_date": "02/09/2023 00:35:26", "content": "

I am still thinking Kaggle team is taking actions and need more time for investigations.

", "votes": null, "replies": [ { "id": 2137327, "author_name": "christofhenkel", "author_url": "", "post_date": "02/09/2023 21:25:25", "content": "

you were right :D

", "votes": null, "replies": [] } ] }, { "id": 2137394, "author_name": "derkanat", "author_url": "", "post_date": "02/10/2023 00:09:02", "content": "

If poteman is not native english speaker, it does not mean that he is a cheater. The reproach about the lack of response from poteman ----Okey---Kaggle has been changed. I see it.

", "votes": null, "replies": [ { "id": 2137398, "author_name": "onodera", "author_url": "", "post_date": "02/10/2023 00:16:04", "content": "

Are you another account of poteman?

", "votes": null, "replies": [ { "id": 2137399, "author_name": "derkanat", "author_url": "", "post_date": "02/10/2023 00:21:34", "content": "

Of course, I'm poteman. You immediately recogranise it, because you are the expert here!

", "votes": null, "replies": [] } ] } ] }, { "id": 2139719, "author_name": "rohithmahadevan", "author_url": "", "post_date": "02/11/2023 04:36:56", "content": "

I will tell u my story. I have been using Kaggle for the past 3 years and I have forked many notebooks for practice. I will never make a forked notebook public and showcase as my work. I have this consciousness of not to get creds for others' work! However, I recently thought of participating in a competition, I completed the work and saw that I got placed in the last 100s. I was a bit disappointed yet when I see other's solutions I could see at least 10 people had the same solution WITHOUT EVEN CHANGING THE COMMENTS! and they were placed in the top 10%. Kaggle should take serious steps about this code plagiarism and encourage upcoming enthusiasts to get recognized for their work. I may be at the last, but I can happily say that the work was completed by me!

\n

Kudos to all the enthusiasts like me who are grinding hard and publishing your own work. Lets make it to grand master one day!

", "votes": null, "replies": [] }, { "id": 2139880, "author_name": "bernalp", "author_url": "", "post_date": "02/11/2023 08:55:18", "content": "

Ok this is great

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2134516": "Congratulations", "2134817": "As the title of this thread says, LB was finalized with suspected cheaters remaining, despite the issue being raised in the following discussions.\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381318\n- https://www.kaggle.com/competitions/otto-recommender-system/discussion/381360\n\nWhether they are banned or not does not change the color of our team medal. But how about the teams on boarderlines? I feel so sorry for them.\n\nAnd if there is no explanation or action on this issue, this is good news not only for the users who may have cheated this time, but also for those who may cheat in future competitions.\n\nFor players who have done the hard work, those who were brave enough to raise the issue, and the value of kaggle competitions, I hope that at least kaggle team will provide some explanation for this issue.", "2134836": "I second this.\nCheaters not being punished is really demotivating for people that compete honestly, and opens the door for more abuse. \nThis is also really bad for Kaggle's and image. The competition was really good, it's a shame that people will remember OTTO as the competition where private sharing was rewarded.\n\nOne of the cheaters jumped to rank 16 in the overall competitions rankings, and the other one got competition GM. \n\nPoke @inversion", "2134909": "This reminds me this\nhttps://www.kaggle.com/competitions/siim-acr-pneumothorax-segmentation/discussion/97499", "2134929": "After 2 years of spending countless hours on Kaggle, seems I was doing it all wrong 🤦‍♂️ there is a better way to becoming GM. 🤷‍♂️", "2134962": "I'm disappointed at the decision of Kaggle staffs.\nStealing the submission from other team is more deserved of the gold model than 16th place solution ?", "2135014": "It's more comfortable to send a few emails than to spend money on renting a server and thinking about how to improve day and night. SAD.", "2135255": "I just got potential winners email from kaggle. It means they decided not to remove them.", "2135297": "I thought I'd weigh in on the decision here. First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed. Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks. Appropriate actions will be taken when sufficient evidence is obtained. Lastly, yes, there are are various activities that look suspicious, but these don't necessarily mean a team has cheated. We do take action when we believe there is sufficient evidence to do so.", "2135304": "I can assure you that while there are cheaters that fly under the radar for some time, there does come a time when there is sufficient for action to be taken.", "2135403": "inversion \nWhen?\nAfter they retire from their well paid job obtained based on their fake Kaggle GM ?", "2135637": "inversion \n\n> Second, some investigations take longer, and we made the decision to not hold up the leaderboard finalization for a process that make take weeks.\n\nIs it safe to assume that this will be reinvestigated for past competitions as well?\nThe allegations could affect not only OTTO participants, but also participants from several past competitions(not only kaggle).\n\nhttps://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n\n> We do take action when we believe there is sufficient evidence to do so.\n\nApparently someone was accidentally banned once, was there sufficient evidence at the time? I demand an explanation.\n\nhttps://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2118580", "2135692": "It is quite disappointing that kaggle did not post an official message even a week after the competition. Feels that the team is not taking this incident seriously. (If you're not in the prize zone, you can easily get away with cheating?)\n\nI think a lot of people would be walking out of the platform if \"appropriate\" action is not taken. It feels GM and gold medals reputation has sunk.", "2135698": "meanwhile you won your gold having @poteman in team. You could leave a comment here about that experience at least :)", "2135705": "https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321#2116033\nI've wrote it here", "2135747": "Now he knows crime pays, and he'll keep on cheating.\n\nReally sad that he is still on kaggle.\n\nBut I'm happy that his inocent team members were not punished. They did deserve the medal and prize.", "2135823": "inversion \n\n> First, the winner team validation process will kick into place, and teams will have to be able to show their work. If indeed, this is not possible (e.g., if there was cheating), the team will be removed.\n\nHave you read the posts about suspicious cheating activity on this competition[^1], [^2], [^3]? (If you want to have a grasp of this issue first, you can refer to a sum-up of this issue[^4].)\n\nWe are not talking about a cheating that all of teammates are cheaters, but a cheater hiding in a team, and stealing their team-mate's solution, then shares the solutions to his cheater's pals who is also hiding in other innocent team. So above cheater screening process do not work for this case. Because their (innocent) team mate do have their works.\n\nIf these reports are true, the suspected two users are highly possible to cheat using the same scheme repeatedly to many teams and competitions over years (and this cheat made them Grandmaster and ranked high on total score). If Kaggle do nothing about them, they will continue to cheat for the future. \n\nIt is not only the issue of this single competition, this is an issue of many past and future competitions.\n\nI hope Kaggle will take this issue properly.\n\n[^1]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381318\n[^2]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381321\n[^3]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381360\n[^4]: https://www.kaggle.com/competitions/otto-recommender-system/discussion/381527", "2135878": "Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.\n\nHowever, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.\n\nI strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.", "2135883": "Agreed. It is just baffling that there is no official statement and someone had to make a post so we can all have a clue of what is happening.", "2135885": "And after they have created many more fake Masters or GMs", "2135913": "I am still thinking Kaggle team is taking actions and need more time for investigations.", "2136029": "poteman and other cheaters keep up your great work! Kaggle recognizes your great job! :D", "2137327": "you were right :D", "2137351": "> Did I miss out on becoming a Grand Master because of cheaters or just due to bad luck (0.001%)? To be honest, I am personally disappointed that I did not receive the gold medal.\n> \n> However, what is of greater concern for the community is that Kaggle's integrity has been called into question more than ever. If cheating is overlooked, even with sufficient circumstantial evidence, then those who approached the competition with sincerity will not be properly recognized.\n> \n> I strongly believe that Kaggle users deserve greater transparency. Kaggle should provide an official explanation instead of only occasionally responding to discussions.\n\nHow did people cheat?", "2137359": "Hi @thrill1, welcome to Kaggle as you join the platform 2 hours ago, just after the accounts of the 2 cheaters were terminated.\nOr should we call you poteman2 or RunningZ2 ?\nUnfortunately it is so easy to create a new account.", "2137382": "Update:\n\nI'm glad to hear about my solo Gold and GM promotion from Kaggle's recent judgment update! \nI think it's a great move for the overall health of the Kaggle community to eliminate teams with cheating suspicions. And, my heart goes out to those innocent participants who unintentionally merged with cheaters.", "2137384": "Congrats! Pay high tribute to these innocent participants.", "2137385": "Sorry, I am new to the platform. They asked me to comment in the checklist and this was an active discussion. Just learning about Kaggle and thinking about posting a competition for my company.", "2137394": "If poteman is not native english speaker, it does not mean that he is a cheater. The reproach about the lack of response from poteman ----Okey---Kaggle has been changed. I see it.", "2137398": "Are you another account of poteman?", "2137399": "Of course, I'm poteman. You immediately recogranise it, because you are the expert here!", "2137411": "Congratulations~", "2137479": "Congrats!!", "2139719": "I will tell u my story. I have been using Kaggle for the past 3 years and I have forked many notebooks for practice. I will never make a forked notebook public and showcase as my work. I have this consciousness of not to get creds for others' work! However, I recently thought of participating in a competition, I completed the work and saw that I got placed in the last 100s. I was a bit disappointed yet when I see other's solutions I could see at least 10 people had the same solution WITHOUT EVEN CHANGING THE COMMENTS! and they were placed in the top 10%. Kaggle should take serious steps about this code plagiarism and encourage upcoming enthusiasts to get recognized for their work. I may be at the last, but I can happily say that the work was completed by me! \n\nKudos to all the enthusiasts like me who are grinding hard and publishing your own work. Lets make it to grand master one day!", "2139880": "Ok this is great", "2140462": "I am a newcomer to Kaggle. Please describe more!", "2142122": "Yeah it's unfortunate. If cheating goes unchecked, it undermines the hard work of fair players and raises doubts about the authenticity of competition outcomes. This can be discouraging for those who compete with honesty and integrity.", "2143790": "I am a newcomer also!" }, "source": "meta" }