![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1556833%2F08876786e3f30b032b36b59eb75e8c31%2FScreenshot%20from%202019-08-23%2017-33-32.png?generation=1566579228036880&alt=media\")
Sure, there are very talented people who do very good on their first competition but to have your first submission scores this high in such a short time after you signed up (<1h) is impossible IMHO. What do you guys think?
\n\nIs there any reasonable explanation for this?\n@LIUCHANG
\n\nUPDATE:\nThis is not witch hunting and it seems it is not a false alarm either. The account is now deleted and I would call this case BUSTED.
\n\n
[![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1556833%2F37155f4aa98bc90d1a3a38ed53641110%2FScreenshot%20from%202019-08-23%2017-33-18.png?generation=1566577947446567&alt=media\")
Sure, there are very talented people who do very good on their first competition but to have your first submission scores this high in such a short time after you signed up (<1h) is impossible IMHO. What do you guys think?
\n\nIs there any reasonable explanation for this?\n@LIUCHANG
\n\nUPDATE:\nThis is not witch hunting and it seems it is not a false alarm either. The account is now deleted and I would call this case BUSTED.
\n\n
[
A little real. \n- Registered at 8:23 on Aug. 19\n- First submission at 10:02 on Aug. 19
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2F66fb9282a8ce1c81d69e61fccf919964%2Fsuperman4.png?generation=1567442159044042&alt=media\")
EDITED\nIs 'lxxsgex' a famous word?\nI'm really sorry if it's irrelevant. I am very nervous.\nhttps://www.kaggle.com/lxxlxxsgex\nhttps://www.kaggle.com/lxxsgex\nhttps://www.kaggle.com/sgexwjy
", "rawMarkdown": "A little real. \n- Registered at 8:23 on Aug. 19\n- First submission at 10:02 on Aug. 19\n\n\n\n**EDITED**\nIs 'lxxsgex' a famous word?\nI'm really sorry if it's irrelevant. I am very nervous.\nhttps://www.kaggle.com/lxxlxxsgex\nhttps://www.kaggle.com/lxxsgex\nhttps://www.kaggle.com/sgexwjy", "votes": 9 }, { "id": 606955, "postDate": "2019-08-24T11:19:26.747Z", "content": "I'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.
\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.
\n\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.
\n\nWould you call all people with high score and low number of submissions cheaters?
\n\nIt is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?
", "rawMarkdown": "I'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.\n\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.\n\nWould you call all people with high score and low number of submissions cheaters?\n\nIt is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?", "votes": 7, "replies": [ { "id": 606960, "postDate": "2019-08-24T11:26:58.220Z", "content": "Usually I would agree with you, but in this case the account is already deleted, so he was right?
", "rawMarkdown": "Usually I would agree with you, but in this case the account is already deleted, so he was right?", "votes": 7 }, { "id": 606976, "postDate": "2019-08-24T11:49:02.353Z", "content": "In this case I agree. But I think it is better to write to admins with such accusations
", "rawMarkdown": "In this case I agree. But I think it is better to write to admins with such accusations", "votes": 8 }, { "id": 607037, "postDate": "2019-08-24T13:47:48.080Z", "content": "@Andrew I share your frustration for \"witch hunting\" but that's not the case here. As Psi and others pointed the account is already deleted.
\n\nTo answer your questions and comment your statements:\n```\nI'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.
\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.
\n\n```I do get heated up when I conclude that there is a high evidence of cheating as in this case. Don't you feel the same? I am not jealous nor angry because there is someone better. This is childish. Wonder what made you think this?
\n\n```\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.
\n\n```\nI agree with your statement that there is a lot of talent around but as @juliencs stated already I cannot understand how someone understand the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training in less than 30 minutes. Besides, to build a strong CV you should have at least a few submissions and if that would be the case I would likely not take any actions.
\n\n\n\n\n\n
Would you call all people with high score and low number of submissions cheaters?
Obviously not.
\n\n\n\n\n\n
It is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?
I took my time to think how someone could score this high as well as where my reaction could lead, including being wrong. It turns out i am guilty of being lazy to describe all my thoughts which I apologize for.
\n\nTo conclude my longest post ever, I am accused here of being jealous and angry person which has a pretty nice chance of getting the 1st. slander on kaggle :)
", "rawMarkdown": "[@Andrew](https://www.kaggle.com/artgor) I share your frustration for \"witch hunting\" but that's not the case here. As Psi and others pointed the account is already deleted. \n\nTo answer your questions and comment your statements:\n```\nI'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.\n\n```I do get heated up when I conclude that there is a high evidence of cheating as in this case. Don't you feel the same? I am not jealous nor angry because there is someone better. This is childish. Wonder what made you think this?\n\n```\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.\n\n```\nI agree with your statement that there is a lot of talent around but as @juliencs stated already I cannot understand how someone understand the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training in less than 30 minutes. Besides, to build a strong CV you should have at least a few submissions and if that would be the case I would likely not take any actions.\n\n> `Would you call all people with high score and low number of submissions cheaters?`\n\nObviously not.\n\n> `It is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?`\n\nI took my time to think how someone could score this high as well as where my reaction could lead, including being wrong. It turns out i am guilty of being lazy to describe all my thoughts which I apologize for.\n\nTo conclude my longest post ever, I am accused here of being jealous and angry person which has a pretty nice chance of getting the 1st. slander on kaggle :)", "votes": 1 }, { "id": 607043, "postDate": "2019-08-24T13:56:13.060Z", "content": "@valanm \nI think that in this case you are completely correct and I was wrong - I was heated from a similar discussion in another thread. I admit that my words were misguided and want to give a sincere apology.
", "rawMarkdown": "@valanm \nI think that in this case you are completely correct and I was wrong - I was heated from a similar discussion in another thread. I admit that my words were misguided and want to give a sincere apology.", "votes": 19 }, { "id": 607108, "postDate": "2019-08-24T16:06:47.293Z", "content": "I accept your apology. It's water under the bridge. I understand you had good intentions :)
", "rawMarkdown": "I accept your apology. It's water under the bridge. I understand you had good intentions :)", "votes": 16 }, { "id": 608076, "postDate": "2019-08-26T09:51:53.280Z", "content": "True, some people have been doing doing ML outside Kaggle for longer than several people that made a lot of submissions here. Its not suprising that they get higher ranking.
", "rawMarkdown": "True, some people have been doing doing ML outside Kaggle for longer than several people that made a lot of submissions here. Its not suprising that they get higher ranking.", "votes": 2 } ] }, { "id": 615044, "postDate": "2019-09-01T11:53:29.840Z", "content": "EDITED: The suspicious account has disappeared.
\n\nI found a new suspicion. Is this possible? What do you think?\n(I don't know where else to report. Sorry.)
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2Ffbe2cba3d576949e0eff79511c05c775%2Fsuperman2.png?generation=1567338632664932&alt=media\")
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2F55b66fb52f9e7df6290e4e3b0a7b39d1%2Fsuperman2_excel.jpg?generation=1567338691792886&alt=media\")
wow, no doubt, a genius..
", "rawMarkdown": "wow, no doubt, a genius.." }, { "id": 615180, "postDate": "2019-09-01T15:10:54.337Z", "content": "The genius account has gone.
", "rawMarkdown": "The genius account has gone.", "votes": 2 }, { "id": 615197, "postDate": "2019-09-01T15:41:13.877Z", "content": "I am quite sure this \"genius\" is still around :) this is 2nd account. Kaggle should be able to trace back via device IP and find him as well as others. Not? But this happens after the competition when a bunch of teams get filtered out and only a few complain and get back to the LB
", "rawMarkdown": "I am quite sure this \"genius\" is still around :) this is 2nd account. Kaggle should be able to trace back via device IP and find him as well as others. Not? But this happens after the competition when a bunch of teams get filtered out and only a few complain and get back to the LB" }, { "id": 615201, "postDate": "2019-09-01T15:48:31.127Z", "content": "It obviously is someone from top rankings, so expect to be busted :D
", "rawMarkdown": "It obviously is someone from top rankings, so expect to be busted :D", "votes": 3 }, { "id": 615210, "postDate": "2019-09-01T16:01:42.193Z", "content": "3 accounts with 0.817+ scores are busted which could easily be a single person
", "rawMarkdown": " 3 accounts with 0.817+ scores are busted which could easily be a single person", "votes": 1 }, { "id": 615535, "postDate": "2019-09-02T04:13:16.557Z", "content": "Lol I would expect someone to do this kind of thing near the end of the competition. Even I feel 5 submissions per day is not enough for me :D
", "rawMarkdown": "Lol I would expect someone to do this kind of thing near the end of the competition. Even I feel 5 submissions per day is not enough for me :D" } ] }, { "id": 616050, "postDate": "2019-09-02T16:27:30.477Z", "content": "How about this?\n- Registered at 1:13 on Aug. 25\n- First submission at 1:20 on Aug. 25
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2F68daaed9190f2bd99d24b688ec28597e%2Fsuperman3.png?generation=1567441628616531&alt=media\")
man, this is some serious eyeing
", "rawMarkdown": "man, this is some serious eyeing ", "votes": 3 } ] }, { "id": 606655, "postDate": "2019-08-23T21:07:30.953Z", "content": "Seems pretty obvious cheating to me. He has to download the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training, and understand how Kaggle kernels work, while getting a score which is frankly unattainable on a first try with only the 2019 dataset in my opinion.
\n\nI guess he's gonna stay visible until the end of the competition and disappear once rankings are final.
\n\nEdit : ok it may not be cheating. Someone decided it was a good idea to post a top 3% kernel right before the end of the competition.
\n\nEdit 2 : seems the kernel has been removed ^^
", "rawMarkdown": "Seems pretty obvious cheating to me. He has to download the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training, and understand how Kaggle kernels work, while getting a score which is frankly unattainable on a first try with only the 2019 dataset in my opinion.\n\nI guess he's gonna stay visible until the end of the competition and disappear once rankings are final.\n\nEdit : ok it may not be cheating. Someone decided it was a good idea to post a top 3% kernel right before the end of the competition.\n\nEdit 2 : seems the kernel has been removed ^^", "votes": 3 }, { "id": 606560, "postDate": "2019-08-23T18:26:37.527Z", "content": "and here I am hustling for more than 45 days🤕
", "rawMarkdown": "and here I am hustling for more than 45 days🤕 ", "votes": 4, "replies": [ { "id": 621235, "postDate": "2019-09-08T11:11:49.900Z", "content": "Hey man, I am a research intern at IITB, if you want to work on kaggle competitions let me know
", "rawMarkdown": "Hey man, I am a research intern at IITB, if you want to work on kaggle competitions let me know" } ] }, { "id": 616067, "postDate": "2019-09-02T16:47:41.547Z", "content": "I have found several other suspicious accounts.\nI know here is not a place to uncover suspicions, but I don't know where to report.
\n\nHi @sohier or other administrators.\nWhat do you think? Could you check suspicious accounts later?\nThank you.
", "rawMarkdown": "I have found several other suspicious accounts.\nI know here is not a place to uncover suspicions, but I don't know where to report.\n\nHi @sohier or other administrators.\nWhat do you think? Could you check suspicious accounts later?\nThank you.", "votes": 1, "replies": [ { "id": 616125, "postDate": "2019-09-02T18:01:56.763Z", "content": "do you let the community know who was filtered out with your detection?
", "rawMarkdown": "do you let the community know who was filtered out with your detection?" }, { "id": 616140, "postDate": "2019-09-02T18:23:32.703Z", "content": "When we finalize the competition, anyone filtered out is simply dropped from the leaderboard.
", "rawMarkdown": "When we finalize the competition, anyone filtered out is simply dropped from the leaderboard.", "votes": 3 }, { "id": 616245, "postDate": "2019-09-02T22:21:47.827Z", "content": "@sohier Thank you for your kind replying. I am relieved to hear that.\nPlease let me know If my previous report appears to be an inappropriate attack.\nIf inappropriate, I will erase the previous comments and not do comment it like this in the future.
", "rawMarkdown": "@sohier Thank you for your kind replying. I am relieved to hear that.\nPlease let me know If my previous report appears to be an inappropriate attack.\nIf inappropriate, I will erase the previous comments and not do comment it like this in the future." }, { "id": 617566, "postDate": "2019-09-04T09:05:59.430Z", "content": "![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F448347%2F7976ffc91133be6f10156fa71fffdd65%2F2019-09-04%2017.56.34.png?generation=1567587416606223&alt=media\")
@octpath0302 \nCould you please also check this 2 guys' submission log?\nI find them usually improved their LB at the same time.
\n\nBut I have no evidence.
", "rawMarkdown": "\n\n@octpath0302 \nCould you please also check this 2 guys' submission log?\nI find them usually improved their LB at the same time.\n\n**But I have no evidence.**", "votes": 3 }, { "id": 617574, "postDate": "2019-09-04T09:17:43.660Z", "content": "I think you wanted to also tag @sohier. You are right that you have no evidence, but this is indeed suspicious as they have collaborated on past competition.
", "rawMarkdown": "I think you wanted to also tag @sohier. You are right that you have no evidence, but this is indeed suspicious as they have collaborated on past competition." }, { "id": 617646, "postDate": "2019-09-04T10:47:37.380Z", "content": "I don't have the intuition that I should start researching because I don't have much experience of Kaggle.\nAnd I want to concentrate on the last effort.\nSorry!
", "rawMarkdown": "I don't have the intuition that I should start researching because I don't have much experience of Kaggle.\nAnd I want to concentrate on the last effort.\nSorry!", "votes": 1 }, { "id": 620810, "postDate": "2019-09-08T01:21:30.917Z", "content": "And these 2 guys I mentioned before, have shared the same Private LB rank shaking. (+5 from Public LB to Private LB)\n@sohier @philippsinger @octpath0302
", "rawMarkdown": "And these 2 guys I mentioned before, have shared the same Private LB rank shaking. (+5 from Public LB to Private LB)\n@sohier @philippsinger @octpath0302 " }, { "id": 620894, "postDate": "2019-09-08T03:43:36.047Z", "content": "@haqishen Due to some big ShakeDown, I lost my energy to look for suspicious people.😂 Sorry!
", "rawMarkdown": "@haqishen Due to some big ShakeDown, I lost my energy to look for suspicious people.😂 Sorry!", "votes": 1 }, { "id": 621093, "postDate": "2019-09-08T07:41:15.893Z", "content": "I am sorry to hear that... But I'm still interested in your method. Would you mind to share your main idea to get to Public LB 0.84?
", "rawMarkdown": "I am sorry to hear that... But I'm still interested in your method. Would you mind to share your main idea to get to Public LB 0.84?" } ] }, { "id": 607835, "postDate": "2019-08-26T00:46:06.453Z", "content": "next superman? @alialzboon20
", "rawMarkdown": "next superman? @alialzboon20", "votes": 1, "replies": [ { "id": 607893, "postDate": "2019-08-26T03:56:33.257Z", "content": "Superman's account was registered at 14:15 on August 25, and it seems that the submission was successful 2, 9, 12 minutes after registration.\nWow!
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2F2f2e1fe71f1fe7f5e7de73526cb93f73%2F.png?generation=1566791300878639&alt=media\")
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2208353%2F6694271dd7a08561644c5c1c4fd8a5d6%2Falialzboon_SharedScreenshot.jpg?generation=1566791318958523&alt=media\")
Just like Molecule somebody published a high scoring kernel for 2 hours and some people probably have used it. Another one in Pneumothorax, 4 days before the the deadline. It's becoming the norm.
", "rawMarkdown": "Just like Molecule somebody published a high scoring kernel for 2 hours and some people probably have used it. Another one in Pneumothorax, 4 days before the the deadline. It's becoming the norm.", "votes": 1, "replies": [ { "id": 608227, "postDate": "2019-08-26T14:10:29.547Z", "content": "FYI, the Pneumothorax kernel you are referring to was in fact published 3weeks ago. But it had a bug which got spotted by one of the forkers and reported back to everyone. So, I would not put it in the same category as this case.
", "rawMarkdown": "FYI, the Pneumothorax kernel you are referring to was in fact published 3weeks ago. But it had a bug which got spotted by one of the forkers and reported back to everyone. So, I would not put it in the same category as this case. ", "votes": 1 }, { "id": 608269, "postDate": "2019-08-26T15:16:38.077Z", "content": "I can read the dates thanks, but the score increase was 3 days ago and the reason does not matter anyway.
", "rawMarkdown": "I can read the dates thanks, but the score increase was 3 days ago and the reason does not matter anyway. ", "votes": -1 } ] }, { "id": 606922, "postDate": "2019-08-24T10:04:05.937Z", "content": "Thanks for bringing this to notice. It may be because of recent Kernel restrictions. Someone may have created another account to get more trials. It's unfair for hard and genuine working people. One person occupies two or more spots on the leaderboard unnecessarily, which forbids someone deserving.
", "rawMarkdown": "Thanks for bringing this to notice. It may be because of recent Kernel restrictions. Someone may have created another account to get more trials. It's unfair for hard and genuine working people. One person occupies two or more spots on the leaderboard unnecessarily, which forbids someone deserving. ", "votes": 1 }, { "id": 606891, "postDate": "2019-08-24T09:05:46.520Z", "content": "This is the case of creating second account and submitting through it. @valanm the profile is deleted now.
", "rawMarkdown": "This is the case of creating second account and submitting through it. @valanm the profile is deleted now.", "votes": 1 }, { "id": 606723, "postDate": "2019-08-24T01:36:16.003Z", "content": "There are 2 possibilities:\n- He has talent and experiences. And he read dicussions which contain hints and public kernels more than enough to achieve 0.80+\n- 2nd account for x2 submission (Violation of the rules)
", "rawMarkdown": "There are 2 possibilities:\n- He has talent and experiences. And he read dicussions which contain hints and public kernels more than enough to achieve 0.80+\n- 2nd account for x2 submission (Violation of the rules)", "votes": 1 }, { "id": 606563, "postDate": "2019-08-23T18:33:48.943Z", "content": "Maybe it's someone's 2nd acc for x2 submission? Also with recent change in GPU limit that's not impossible that people want more instances :/
", "rawMarkdown": "Maybe it's someone's 2nd acc for x2 submission? Also with recent change in GPU limit that's not impossible that people want more instances :/", "votes": 1, "replies": [ { "id": 606576, "postDate": "2019-08-23T18:54:22.780Z", "content": "that is violation of rules
", "rawMarkdown": "that is violation of rules" }, { "id": 606609, "postDate": "2019-08-23T19:37:53.553Z", "content": "I agree!
", "rawMarkdown": "I agree!" } ] }, { "id": 607096, "postDate": "2019-08-24T15:54:01.983Z", "content": "The account is now deleted and I would call this case BUSTED.
\n\nAs this is likely an extra account used for LB probing there is a fair chance that the one behind it:\n1.) has a score above 0.817\n2.) slightly below (pretty sure it's above 0.8)
\n\nI am sure kaggle will handle this anyway and remove the team who naively violated the rules so I am quite curious what's gonna happen.
", "rawMarkdown": "The account is now deleted and I would call this case BUSTED. \n\nAs this is likely an extra account used for LB probing there is a fair chance that the one behind it:\n1.) has a score above 0.817\n2.) slightly below (pretty sure it's above 0.8)\n\nI am sure kaggle will handle this anyway and remove the team who naively violated the rules so I am quite curious what's gonna happen.", "votes": 2, "replies": [ { "id": 608148, "postDate": "2019-08-26T12:05:24.647Z", "content": "Observing what looks like cheating is infuriating but this kind of public shaming for its risk of false accusations is unproductive and could lead to unfortunate accidents. I think it's better just report the suspected accounts silently and enjoy the competition. Even if you have one in a thousand chance of being wrong, this behavior would be interpreted as harassment.
", "rawMarkdown": "Observing what looks like cheating is infuriating but this kind of public shaming for its risk of false accusations is unproductive and could lead to unfortunate accidents. I think it's better just report the suspected accounts silently and enjoy the competition. Even if you have one in a thousand chance of being wrong, this behavior would be interpreted as harassment.", "votes": 2 }, { "id": 608220, "postDate": "2019-08-26T14:05:30.123Z", "content": "I am not sure I understand correctly. Do you refer to my comment above or the whole topic?
", "rawMarkdown": "I am not sure I understand correctly. Do you refer to my comment above or the whole topic?" }, { "id": 608273, "postDate": "2019-08-26T15:19:18.183Z", "content": "Your choice.
", "rawMarkdown": "Your choice." }, { "id": 608298, "postDate": "2019-08-26T15:48:15.927Z", "content": "I understand my action has it's negative side, but I am strongly convinced bringing it up has far more positive. TBH, I do not see what's wrong with my comment.
", "rawMarkdown": "I understand my action has it's negative side, but I am strongly convinced bringing it up has far more positive. TBH, I do not see what's wrong with my comment.", "votes": 1 }, { "id": 608315, "postDate": "2019-08-26T16:05:14.550Z", "content": "I didn't say in this specific case you are wrong. I'm just saying maybe he was the smartest data scientist you and I have ever seen and without any evidence or reason we shouldn't accuse anybody. Even if you are sure why not report it quietly.
", "rawMarkdown": "I didn't say in this specific case you are wrong. I'm just saying maybe he was the smartest data scientist you and I have ever seen and without any evidence or reason we shouldn't accuse anybody. Even if you are sure why not report it quietly." }, { "id": 608323, "postDate": "2019-08-26T16:15:33.160Z", "content": "Even if you are sure why not report it quietly.\nu r correct. next time
There was no chance this account was legit as well as the other one with 2 minutes for downloading data and training models...why such a rush
", "rawMarkdown": "`Even if you are sure why not report it quietly.`\nu r correct. next time\n\nThere was no chance this account was legit as well as the other one with 2 minutes for downloading data and training models...why such a rush ", "votes": 2 } ] }, { "id": 606583, "postDate": "2019-08-23T19:03:44.137Z", "content": "why would this topic deserve a downvote? :)
", "rawMarkdown": "why would this topic deserve a downvote? :)", "votes": 2, "replies": [ { "id": 607196, "postDate": "2019-08-24T18:59:59.187Z", "content": "I am still puzzled by the amount of downvotes this topic got. I get notified when people upvote but the count (X) sometimes shows X+1, rarely X-1, and sometimes stays the same as before notification :D
", "rawMarkdown": "I am still puzzled by the amount of downvotes this topic got. I get notified when people upvote but the count (X) sometimes shows X+1, rarely X-1, and sometimes stays the same as before notification :D" } ] }, { "id": 608529, "postDate": "2019-08-26T22:04:14.713Z", "content": "hahahaha
", "rawMarkdown": "hahahaha" }, { "id": 608043, "postDate": "2019-08-26T08:43:51.480Z", "content": "maybe
", "rawMarkdown": "maybe" }, { "id": 608015, "postDate": "2019-08-26T08:01:05.017Z", "content": "CHEATERS are everywhere ? 😑
what are the odds that this \"X-men\" is back with 3rd account? The same pattern (new account and submission within couple of minutes) with the same score (0.817) just appeared on the LB.
", "rawMarkdown": "what are the odds that this \"X-men\" is back with 3rd account? The same pattern (new account and submission within couple of minutes) with the same score (0.817) just appeared on the LB. " }, { "id": 606718, "postDate": "2019-08-24T01:24:04.367Z", "content": "It could be either talent or team formation(without forming a team)
", "rawMarkdown": "It could be either talent or team formation(without forming a team)" }, { "id": 606564, "postDate": "2019-08-23T18:34:17.803Z", "content": "maybe the code leaked?
", "rawMarkdown": "maybe the code leaked?", "replies": [ { "id": 606584, "postDate": "2019-08-23T19:07:06.210Z", "content": "but you would still need to at least train the model/models.
", "rawMarkdown": "but you would still need to at least train the model/models.", "votes": 1 }, { "id": 606586, "postDate": "2019-08-23T19:07:28.100Z", "content": "maybe trained models leaked too :)
", "rawMarkdown": "maybe trained models leaked too :)" } ] }, { "id": 616256, "postDate": "2019-09-02T22:46:16.510Z", "rawMarkdown": "", "isDeleted": true }, { "id": 606861, "postDate": "2019-08-24T08:16:07.117Z", "rawMarkdown": "", "isDeleted": true }, { "id": 606802, "postDate": "2019-08-24T05:28:41.633Z", "rawMarkdown": "", "votes": -4, "isDeleted": true }, { "id": 606739, "postDate": "2019-08-24T02:19:36.703Z", "rawMarkdown": "", "votes": 1, "isDeleted": true, "replies": [ { "id": 607111, "postDate": "2019-08-24T16:07:55.867Z", "content": "It's not OK. I think this is written in the rules
", "rawMarkdown": "It's not OK. I think this is written in the rules" }, { "id": 607269, "postDate": "2019-08-24T23:35:03.133Z", "content": "why would you even have multiple accounts if it wasnt for more submissions though?
", "rawMarkdown": "why would you even have multiple accounts if it wasnt for more submissions though?", "votes": 1 }, { "id": 607350, "postDate": "2019-08-25T04:24:43.177Z", "content": "To have access to more GPUs and experiment more. This is one of the reasons why they set the limit to 2commits 1 interactive
", "rawMarkdown": "To have access to more GPUs and experiment more. This is one of the reasons why they set the limit to 2commits 1 interactive" } ] }, { "id": 606721, "postDate": "2019-08-24T01:33:16.040Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 616059, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-02T16:39:54.073000", "content": "A little real. \n- Registered at 8:23 on Aug. 19\n- First submission at 10:02 on Aug. 19
\n\n
EDITED\nIs 'lxxsgex' a famous word?\nI'm really sorry if it's irrelevant. I am very nervous.\nhttps://www.kaggle.com/lxxlxxsgex\nhttps://www.kaggle.com/lxxsgex\nhttps://www.kaggle.com/sgexwjy
", "votes": 9, "replies": [] }, { "id": 606955, "author_name": "Andrey Lukyanenko", "author_url": "", "post_date": "2019-08-24T11:19:26.747000", "content": "I'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.
\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.
\n\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.
\n\nWould you call all people with high score and low number of submissions cheaters?
\n\nIt is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?
", "votes": 7, "replies": [ { "id": 606960, "author_name": "Psi", "author_url": "", "post_date": "2019-08-24T11:26:58.220000", "content": "Usually I would agree with you, but in this case the account is already deleted, so he was right?
", "votes": 7, "replies": [] }, { "id": 606976, "author_name": "Andrey Lukyanenko", "author_url": "", "post_date": "2019-08-24T11:49:02.353000", "content": "In this case I agree. But I think it is better to write to admins with such accusations
", "votes": 8, "replies": [] }, { "id": 607037, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-24T13:47:48.080000", "content": "@Andrew I share your frustration for \"witch hunting\" but that's not the case here. As Psi and others pointed the account is already deleted.
\n\nTo answer your questions and comment your statements:\n```\nI'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.
\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.
\n\n```I do get heated up when I conclude that there is a high evidence of cheating as in this case. Don't you feel the same? I am not jealous nor angry because there is someone better. This is childish. Wonder what made you think this?
\n\n```\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.
\n\n```\nI agree with your statement that there is a lot of talent around but as @juliencs stated already I cannot understand how someone understand the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training in less than 30 minutes. Besides, to build a strong CV you should have at least a few submissions and if that would be the case I would likely not take any actions.
\n\n\n\n\n\n
Would you call all people with high score and low number of submissions cheaters?
Obviously not.
\n\n\n\n\n\n
It is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?
I took my time to think how someone could score this high as well as where my reaction could lead, including being wrong. It turns out i am guilty of being lazy to describe all my thoughts which I apologize for.
\n\nTo conclude my longest post ever, I am accused here of being jealous and angry person which has a pretty nice chance of getting the 1st. slander on kaggle :)
", "votes": 1, "replies": [] }, { "id": 607043, "author_name": "Andrey Lukyanenko", "author_url": "", "post_date": "2019-08-24T13:56:13.060000", "content": "@valanm \nI think that in this case you are completely correct and I was wrong - I was heated from a similar discussion in another thread. I admit that my words were misguided and want to give a sincere apology.
", "votes": 19, "replies": [] }, { "id": 607108, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-24T16:06:47.293000", "content": "I accept your apology. It's water under the bridge. I understand you had good intentions :)
", "votes": 16, "replies": [] }, { "id": 608076, "author_name": "Magall George", "author_url": "", "post_date": "2019-08-26T09:51:53.280000", "content": "True, some people have been doing doing ML outside Kaggle for longer than several people that made a lot of submissions here. Its not suprising that they get higher ranking.
", "votes": 2, "replies": [] } ] }, { "id": 615044, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-01T11:53:29.840000", "content": "EDITED: The suspicious account has disappeared.
\n\nI found a new suspicion. Is this possible? What do you think?\n(I don't know where else to report. Sorry.)
\n\n\n
wow, no doubt, a genius..
", "votes": 0, "replies": [] }, { "id": 615180, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-01T15:10:54.337000", "content": "The genius account has gone.
", "votes": 2, "replies": [] }, { "id": 615197, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-09-01T15:41:13.877000", "content": "I am quite sure this \"genius\" is still around :) this is 2nd account. Kaggle should be able to trace back via device IP and find him as well as others. Not? But this happens after the competition when a bunch of teams get filtered out and only a few complain and get back to the LB
", "votes": 0, "replies": [] }, { "id": 615201, "author_name": "Psi", "author_url": "", "post_date": "2019-09-01T15:48:31.127000", "content": "It obviously is someone from top rankings, so expect to be busted :D
", "votes": 3, "replies": [] }, { "id": 615210, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-09-01T16:01:42.193000", "content": "3 accounts with 0.817+ scores are busted which could easily be a single person
", "votes": 1, "replies": [] }, { "id": 615535, "author_name": "Quan", "author_url": "", "post_date": "2019-09-02T04:13:16.557000", "content": "Lol I would expect someone to do this kind of thing near the end of the competition. Even I feel 5 submissions per day is not enough for me :D
", "votes": 0, "replies": [] } ] }, { "id": 616050, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-02T16:27:30.477000", "content": "How about this?\n- Registered at 1:13 on Aug. 25\n- First submission at 1:20 on Aug. 25
\n\n
man, this is some serious eyeing
", "votes": 3, "replies": [] } ] }, { "id": 606655, "author_name": "juliencs", "author_url": "", "post_date": "2019-08-23T21:07:30.953000", "content": "Seems pretty obvious cheating to me. He has to download the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training, and understand how Kaggle kernels work, while getting a score which is frankly unattainable on a first try with only the 2019 dataset in my opinion.
\n\nI guess he's gonna stay visible until the end of the competition and disappear once rankings are final.
\n\nEdit : ok it may not be cheating. Someone decided it was a good idea to post a top 3% kernel right before the end of the competition.
\n\nEdit 2 : seems the kernel has been removed ^^
", "votes": 3, "replies": [] }, { "id": 606560, "author_name": "Chinmay Bharti", "author_url": "", "post_date": "2019-08-23T18:26:37.527000", "content": "and here I am hustling for more than 45 days🤕
", "votes": 4, "replies": [ { "id": 621235, "author_name": "Krishna Adithya", "author_url": "", "post_date": "2019-09-08T11:11:49.900000", "content": "Hey man, I am a research intern at IITB, if you want to work on kaggle competitions let me know
", "votes": 0, "replies": [] } ] }, { "id": 616067, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-02T16:47:41.547000", "content": "I have found several other suspicious accounts.\nI know here is not a place to uncover suspicions, but I don't know where to report.
\n\nHi @sohier or other administrators.\nWhat do you think? Could you check suspicious accounts later?\nThank you.
", "votes": 1, "replies": [ { "id": 616125, "author_name": "Sohier Dane", "author_url": "", "post_date": "2019-09-02T18:01:56.763000", "content": "do you let the community know who was filtered out with your detection?
", "votes": 0, "replies": [] }, { "id": 616140, "author_name": "Sohier Dane", "author_url": "", "post_date": "2019-09-02T18:23:32.703000", "content": "When we finalize the competition, anyone filtered out is simply dropped from the leaderboard.
", "votes": 3, "replies": [] }, { "id": 616245, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-02T22:21:47.827000", "content": "@sohier Thank you for your kind replying. I am relieved to hear that.\nPlease let me know If my previous report appears to be an inappropriate attack.\nIf inappropriate, I will erase the previous comments and not do comment it like this in the future.
", "votes": 0, "replies": [] }, { "id": 617566, "author_name": "Qishen Ha", "author_url": "", "post_date": "2019-09-04T09:05:59.430000", "content": "
@octpath0302 \nCould you please also check this 2 guys' submission log?\nI find them usually improved their LB at the same time.
\n\nBut I have no evidence.
", "votes": 3, "replies": [] }, { "id": 617574, "author_name": "Psi", "author_url": "", "post_date": "2019-09-04T09:17:43.660000", "content": "I think you wanted to also tag @sohier. You are right that you have no evidence, but this is indeed suspicious as they have collaborated on past competition.
", "votes": 0, "replies": [] }, { "id": 617646, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-04T10:47:37.380000", "content": "I don't have the intuition that I should start researching because I don't have much experience of Kaggle.\nAnd I want to concentrate on the last effort.\nSorry!
", "votes": 1, "replies": [] }, { "id": 620810, "author_name": "Qishen Ha", "author_url": "", "post_date": "2019-09-08T01:21:30.917000", "content": "And these 2 guys I mentioned before, have shared the same Private LB rank shaking. (+5 from Public LB to Private LB)\n@sohier @philippsinger @octpath0302
", "votes": 0, "replies": [] }, { "id": 620894, "author_name": "oct_path", "author_url": "", "post_date": "2019-09-08T03:43:36.047000", "content": "@haqishen Due to some big ShakeDown, I lost my energy to look for suspicious people.😂 Sorry!
", "votes": 1, "replies": [] }, { "id": 621093, "author_name": "Qishen Ha", "author_url": "", "post_date": "2019-09-08T07:41:15.893000", "content": "I am sorry to hear that... But I'm still interested in your method. Would you mind to share your main idea to get to Public LB 0.84?
", "votes": 0, "replies": [] } ] }, { "id": 607835, "author_name": "hirune924", "author_url": "", "post_date": "2019-08-26T00:46:06.453000", "content": "next superman? @alialzboon20
", "votes": 1, "replies": [ { "id": 607893, "author_name": "oct_path", "author_url": "", "post_date": "2019-08-26T03:56:33.257000", "content": "Superman's account was registered at 14:15 on August 25, and it seems that the submission was successful 2, 9, 12 minutes after registration.\nWow!
\n\n\n
Just like Molecule somebody published a high scoring kernel for 2 hours and some people probably have used it. Another one in Pneumothorax, 4 days before the the deadline. It's becoming the norm.
", "votes": 1, "replies": [ { "id": 608227, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-26T14:10:29.547000", "content": "FYI, the Pneumothorax kernel you are referring to was in fact published 3weeks ago. But it had a bug which got spotted by one of the forkers and reported back to everyone. So, I would not put it in the same category as this case.
", "votes": 1, "replies": [] }, { "id": 608269, "author_name": "Sia", "author_url": "", "post_date": "2019-08-26T15:16:38.077000", "content": "I can read the dates thanks, but the score increase was 3 days ago and the reason does not matter anyway.
", "votes": -1, "replies": [] } ] }, { "id": 606922, "author_name": "Filemon", "author_url": "", "post_date": "2019-08-24T10:04:05.937000", "content": "Thanks for bringing this to notice. It may be because of recent Kernel restrictions. Someone may have created another account to get more trials. It's unfair for hard and genuine working people. One person occupies two or more spots on the leaderboard unnecessarily, which forbids someone deserving.
", "votes": 1, "replies": [] }, { "id": 606891, "author_name": "NB(Nishant_Bhadauria)", "author_url": "", "post_date": "2019-08-24T09:05:46.520000", "content": "This is the case of creating second account and submitting through it. @valanm the profile is deleted now.
", "votes": 1, "replies": [] }, { "id": 606723, "author_name": "Quan", "author_url": "", "post_date": "2019-08-24T01:36:16.003000", "content": "There are 2 possibilities:\n- He has talent and experiences. And he read dicussions which contain hints and public kernels more than enough to achieve 0.80+\n- 2nd account for x2 submission (Violation of the rules)
", "votes": 1, "replies": [] }, { "id": 606563, "author_name": "Brian Lee", "author_url": "", "post_date": "2019-08-23T18:33:48.943000", "content": "Maybe it's someone's 2nd acc for x2 submission? Also with recent change in GPU limit that's not impossible that people want more instances :/
", "votes": 1, "replies": [ { "id": 606576, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-23T18:54:22.780000", "content": "that is violation of rules
", "votes": 0, "replies": [] }, { "id": 606609, "author_name": "Brian Lee", "author_url": "", "post_date": "2019-08-23T19:37:53.553000", "content": "I agree!
", "votes": 0, "replies": [] } ] }, { "id": 607096, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-24T15:54:01.983000", "content": "The account is now deleted and I would call this case BUSTED.
\n\nAs this is likely an extra account used for LB probing there is a fair chance that the one behind it:\n1.) has a score above 0.817\n2.) slightly below (pretty sure it's above 0.8)
\n\nI am sure kaggle will handle this anyway and remove the team who naively violated the rules so I am quite curious what's gonna happen.
", "votes": 2, "replies": [ { "id": 608148, "author_name": "Sia", "author_url": "", "post_date": "2019-08-26T12:05:24.647000", "content": "Observing what looks like cheating is infuriating but this kind of public shaming for its risk of false accusations is unproductive and could lead to unfortunate accidents. I think it's better just report the suspected accounts silently and enjoy the competition. Even if you have one in a thousand chance of being wrong, this behavior would be interpreted as harassment.
", "votes": 2, "replies": [] }, { "id": 608220, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-26T14:05:30.123000", "content": "I am not sure I understand correctly. Do you refer to my comment above or the whole topic?
", "votes": 0, "replies": [] }, { "id": 608273, "author_name": "Sia", "author_url": "", "post_date": "2019-08-26T15:19:18.183000", "content": "Your choice.
", "votes": 0, "replies": [] }, { "id": 608298, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-26T15:48:15.927000", "content": "I understand my action has it's negative side, but I am strongly convinced bringing it up has far more positive. TBH, I do not see what's wrong with my comment.
", "votes": 1, "replies": [] }, { "id": 608315, "author_name": "Sia", "author_url": "", "post_date": "2019-08-26T16:05:14.550000", "content": "I didn't say in this specific case you are wrong. I'm just saying maybe he was the smartest data scientist you and I have ever seen and without any evidence or reason we shouldn't accuse anybody. Even if you are sure why not report it quietly.
", "votes": 0, "replies": [] }, { "id": 608323, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-26T16:15:33.160000", "content": "Even if you are sure why not report it quietly.\nu r correct. next time
There was no chance this account was legit as well as the other one with 2 minutes for downloading data and training models...why such a rush
", "votes": 2, "replies": [] } ] }, { "id": 606583, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-23T19:03:44.137000", "content": "why would this topic deserve a downvote? :)
", "votes": 2, "replies": [ { "id": 607196, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-24T18:59:59.187000", "content": "I am still puzzled by the amount of downvotes this topic got. I get notified when people upvote but the count (X) sometimes shows X+1, rarely X-1, and sometimes stays the same as before notification :D
", "votes": 0, "replies": [] } ] }, { "id": 608529, "author_name": "ImperadorHades", "author_url": "", "post_date": "2019-08-26T22:04:14.713000", "content": "hahahaha
", "votes": 0, "replies": [] }, { "id": 608043, "author_name": "hoangton", "author_url": "", "post_date": "2019-08-26T08:43:51.480000", "content": "maybe
", "votes": 0, "replies": [] }, { "id": 608015, "author_name": "Ashikur Rashid Khan", "author_url": "", "post_date": "2019-08-26T08:01:05.017000", "content": "CHEATERS are everywhere ? 😑
what are the odds that this \"X-men\" is back with 3rd account? The same pattern (new account and submission within couple of minutes) with the same score (0.817) just appeared on the LB.
", "votes": 0, "replies": [] }, { "id": 606718, "author_name": "Bibek", "author_url": "", "post_date": "2019-08-24T01:24:04.367000", "content": "It could be either talent or team formation(without forming a team)
", "votes": 0, "replies": [] }, { "id": 606564, "author_name": "leixiang@AInnovation", "author_url": "", "post_date": "2019-08-23T18:34:17.803000", "content": "maybe the code leaked?
", "votes": 0, "replies": [ { "id": 606584, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-23T19:07:06.210000", "content": "but you would still need to at least train the model/models.
", "votes": 1, "replies": [] }, { "id": 606586, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-23T19:07:28.100000", "content": "maybe trained models leaked too :)
", "votes": 0, "replies": [] } ] }, { "id": 616256, "author_name": "", "author_url": "", "post_date": "2019-09-02T22:46:16.510000", "content": "", "votes": 0, "replies": [] }, { "id": 606861, "author_name": "", "author_url": "", "post_date": "2019-08-24T08:16:07.117000", "content": "", "votes": 0, "replies": [] }, { "id": 606802, "author_name": "", "author_url": "", "post_date": "2019-08-24T05:28:41.633000", "content": "", "votes": -4, "replies": [] }, { "id": 606739, "author_name": "", "author_url": "", "post_date": "2019-08-24T02:19:36.703000", "content": "", "votes": 1, "replies": [ { "id": 607111, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-24T16:07:55.867000", "content": "It's not OK. I think this is written in the rules
", "votes": 0, "replies": [] }, { "id": 607269, "author_name": "sh", "author_url": "", "post_date": "2019-08-24T23:35:03.133000", "content": "why would you even have multiple accounts if it wasnt for more submissions though?
", "votes": 1, "replies": [] }, { "id": 607350, "author_name": "Miroslav Valan", "author_url": "", "post_date": "2019-08-25T04:24:43.177000", "content": "To have access to more GPUs and experiment more. This is one of the reasons why they set the limit to 2commits 1 interactive
", "votes": 0, "replies": [] } ] }, { "id": 606721, "author_name": "", "author_url": "", "post_date": "2019-08-24T01:33:16.040000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "606493": "Sure, there are very talented people who do very good on their first competition but to have your first submission scores this high in such a short time after you signed up (<1h) is impossible IMHO. What do you guys think? \n\nIs there any reasonable explanation for this?\n[@LIUCHANG](https://www.kaggle.com/chang522492401) \n\nUPDATE:\nThis is not witch hunting and it seems it is not a false alarm either. The account is now deleted and I would call this case BUSTED.\n\n\n\n\n[\n", "616059": "A little real. \n- Registered at 8:23 on Aug. 19\n- First submission at 10:02 on Aug. 19\n\n\n\n**EDITED**\nIs 'lxxsgex' a famous word?\nI'm really sorry if it's irrelevant. I am very nervous.\nhttps://www.kaggle.com/lxxlxxsgex\nhttps://www.kaggle.com/lxxsgex\nhttps://www.kaggle.com/sgexwjy", "606955": "I'm sick of this. Really sick.\nMore and more often I see that some people turn to witch hunting and jealousy.\n\nTrue, there were, there are and there will be cheaters. But if you label every person who got a high score from a couple submissions, I pity you - it seems you are angry that there is someone better than you.\n\nThe world is big. There are bright minds, there are people with domain expertise, there are people with published papers in the relevant domain, there are those who create a stable validation and validate on CV and so on.\n\nWould you call all people with high score and low number of submissions cheaters?\n\nIt is so easy to accuse other people of cheating. But \"for some reason\" I don't see any threads with apologizes for wrong and baseless accusations. Anyone ready to be held responsible for slander?", "615044": "**EDITED: The suspicious account has disappeared.**\n\nI found a new suspicion. Is this possible? What do you think?\n(I don't know where else to report. Sorry.)\n\n- Registered at 15:05 on August 28\n- First submission at 15:28 on August 28\n\n\n\n", "616050": "How about this?\n- Registered at 1:13 on Aug. 25\n- First submission at 1:20 on Aug. 25\n\n", "606655": "Seems pretty obvious cheating to me. He has to download the data, decide how to treat the problem (classification/regression, which loss, etc), perform the preprocessing, run the training, and understand how Kaggle kernels work, while getting a score which is frankly unattainable on a first try with only the 2019 dataset in my opinion.\n\nI guess he's gonna stay visible until the end of the competition and disappear once rankings are final.\n\nEdit : ok it may not be cheating. Someone decided it was a good idea to post a top 3% kernel right before the end of the competition.\n\nEdit 2 : seems the kernel has been removed ^^", "606560": "and here I am hustling for more than 45 days🤕 ", "616067": "I have found several other suspicious accounts.\nI know here is not a place to uncover suspicions, but I don't know where to report.\n\nHi @sohier or other administrators.\nWhat do you think? Could you check suspicious accounts later?\nThank you.", "607835": "next superman? @alialzboon20", "607053": "Just like Molecule somebody published a high scoring kernel for 2 hours and some people probably have used it. Another one in Pneumothorax, 4 days before the the deadline. It's becoming the norm.", "606922": "Thanks for bringing this to notice. It may be because of recent Kernel restrictions. Someone may have created another account to get more trials. It's unfair for hard and genuine working people. One person occupies two or more spots on the leaderboard unnecessarily, which forbids someone deserving. ", "606891": "This is the case of creating second account and submitting through it. @valanm the profile is deleted now.", "606723": "There are 2 possibilities:\n- He has talent and experiences. And he read dicussions which contain hints and public kernels more than enough to achieve 0.80+\n- 2nd account for x2 submission (Violation of the rules)", "606563": "Maybe it's someone's 2nd acc for x2 submission? Also with recent change in GPU limit that's not impossible that people want more instances :/", "607096": "The account is now deleted and I would call this case BUSTED. \n\nAs this is likely an extra account used for LB probing there is a fair chance that the one behind it:\n1.) has a score above 0.817\n2.) slightly below (pretty sure it's above 0.8)\n\nI am sure kaggle will handle this anyway and remove the team who naively violated the rules so I am quite curious what's gonna happen.", "606583": "why would this topic deserve a downvote? :)", "608529": "hahahaha", "608043": "maybe", "608015": "``CHEATERS are everywhere ? 😑 ``", "607719": "what are the odds that this \"X-men\" is back with 3rd account? The same pattern (new account and submission within couple of minutes) with the same score (0.817) just appeared on the LB. ", "606718": "It could be either talent or team formation(without forming a team)", "606564": "maybe the code leaked?", "616256": "", "606861": "", "606802": "", "606739": "", "606721": "" } }