It's suggested that the final score be calculated by combining both the Public Score and the Private Score to balance the differences in dataset distribution. What do you think?
", "messages": [ { "id": "3398362", "postDate": "01/29/2026 03:07:41", "content": "It's suggested that the final score be calculated by combining both the Public Score and the Private Score to balance the differences in dataset distribution. What do you think?
", "rawMarkdown": "It's suggested that the final score be calculated by combining both the Public Score and the Private Score to balance the differences in dataset distribution. What do you think?", "votes": null }, { "id": "3398367", "postDate": "01/29/2026 03:16:51", "content": "Isn't the testing phase over ? This ain't the final score? Or we can have more shakeups?
", "rawMarkdown": "Isn't the testing phase over ? This ain't the final score? Or we can have more shakeups?", "votes": null }, { "id": "3398368", "postDate": "01/29/2026 03:19:55", "content": "Public datasets can be overfitted to improve scores, but private datasets cannot, in order to measure the true performance of the model.
", "rawMarkdown": "Public datasets can be overfitted to improve scores, but private datasets cannot, in order to measure the true performance of the model.", "votes": null }, { "id": "3398394", "postDate": "01/29/2026 04:07:10", "content": "Why is there such a huge gap between the public score and the private score? It must be related to the dataset, right? I've competed before, and the larger the dataset, the smaller the shake. How is the private dataset split?
", "rawMarkdown": "Why is there such a huge gap between the public score and the private score? It must be related to the dataset, right? I've competed before, and the larger the dataset, the smaller the shake. How is the private dataset split?", "votes": null }, { "id": "3398841", "postDate": "01/29/2026 19:15:16", "content": "Its done - kaggle might review for cheating, etc but no more real changes - you might improve a spot or two.
", "rawMarkdown": "Its done - kaggle might review for cheating, etc but no more real changes - you might improve a spot or two.", "votes": null }, { "id": "3398983", "postDate": "01/30/2026 02:27:27", "content": "Are you the competition organizer?
", "rawMarkdown": "Are you the competition organizer?", "votes": null }, { "id": "3398984", "postDate": "01/30/2026 02:29:15", "content": "May I direct my appeal regarding the revoked competition results to you?
", "rawMarkdown": "May I direct my appeal regarding the revoked competition results to you?", "votes": null }, { "id": "3399028", "postDate": "01/30/2026 04:34:08", "content": "Not the organizer. Just a person who's been on the site for 8 years.
\nI saw your question in another discussion post - you likely have / will receive an email about revoked status - what was the approximate leader board position just before the end?
\nThe email will likely tell you why. I got an email for being a rude person the other day in discussion post - the links to appeal were a bit messy but you could get there. An AI will handle the initial part mostly likely.
\nIf no email there's a drop down on the left panel - click More and than support. in the support pages you can follow to appeal. But wait a day or two for the email would be my advice.
", "rawMarkdown": "Not the organizer. Just a person who's been on the site for 8 years.\n\nI saw your question in another discussion post - you likely have / will receive an email about revoked status - what was the approximate leader board position just before the end? \n\nThe email will likely tell you why. I got an email for being a rude person the other day in discussion post - the links to appeal were a bit messy but you could get there. An AI will handle the initial part mostly likely.\n\nIf no email there's a drop down on the left panel - click More and than support. in the [support](https://www.kaggle.com/contact#/account/enforcement-action) pages you can follow to appeal. But wait a day or two for the email would be my advice.", "votes": null }, { "id": "3399075", "postDate": "01/30/2026 06:36:03", "content": "Thank you. I achieved a public leaderboard score of over 0.74 and ranked 123rd. I selected the two submissions with the highest public leaderboard scores as my final entries, but their corresponding private leaderboard score was only 0.60. However, among all my submissions, there was one with a private leaderboard score of 0.61, yet its public leaderboard score was merely 0.71, which is why I didn't choose it. It’s a great pity that my competition results are now gone. I feel quite disappointed, as I made more than 140 submissions and dedicated over two months to rigorous model training.@addisonhoward @qiyuliaocsiro
", "rawMarkdown": "Thank you. I achieved a public leaderboard score of over 0.74 and ranked 123rd. I selected the two submissions with the highest public leaderboard scores as my final entries, but their corresponding private leaderboard score was only 0.60. However, among all my submissions, there was one with a private leaderboard score of 0.61, yet its public leaderboard score was merely 0.71, which is why I didn't choose it. It’s a great pity that my competition results are now gone. I feel quite disappointed, as I made more than 140 submissions and dedicated over two months to rigorous model training.@addisonhoward @qiyuliaocsiro", "votes": null }, { "id": "3399253", "postDate": "01/30/2026 14:03:34", "content": "We've cleaned the leaderboard and disqualified some teams that have violated the rules. If you think you were removed by mistake, or believe you have evidence that suggests another team cheated, please contact compliance. Please fill in all the fields honestly.
", "rawMarkdown": "We've cleaned the leaderboard and disqualified some teams that have violated the rules. If you think you were removed by mistake, or believe you have evidence that suggests another team cheated, please contact [compliance](https://www.kaggle.com/compliance). Please fill in all the fields honestly.", "votes": null } ], "comments": [ { "id": 3398367, "author_name": "suman2208", "author_url": "", "post_date": "01/29/2026 03:16:51", "content": "Isn't the testing phase over ? This ain't the final score? Or we can have more shakeups?
", "votes": null, "replies": [ { "id": 3398841, "author_name": "pcjimmmy", "author_url": "", "post_date": "01/29/2026 19:15:16", "content": "Its done - kaggle might review for cheating, etc but no more real changes - you might improve a spot or two.
", "votes": null, "replies": [ { "id": 3398983, "author_name": "yiyanghu", "author_url": "", "post_date": "01/30/2026 02:27:27", "content": "Are you the competition organizer?
", "votes": null, "replies": [] }, { "id": 3398984, "author_name": "yiyanghu", "author_url": "", "post_date": "01/30/2026 02:29:15", "content": "May I direct my appeal regarding the revoked competition results to you?
", "votes": null, "replies": [ { "id": 3399028, "author_name": "pcjimmmy", "author_url": "", "post_date": "01/30/2026 04:34:08", "content": "Not the organizer. Just a person who's been on the site for 8 years.
\nI saw your question in another discussion post - you likely have / will receive an email about revoked status - what was the approximate leader board position just before the end?
\nThe email will likely tell you why. I got an email for being a rude person the other day in discussion post - the links to appeal were a bit messy but you could get there. An AI will handle the initial part mostly likely.
\nIf no email there's a drop down on the left panel - click More and than support. in the support pages you can follow to appeal. But wait a day or two for the email would be my advice.
", "votes": null, "replies": [ { "id": 3399075, "author_name": "yiyanghu", "author_url": "", "post_date": "01/30/2026 06:36:03", "content": "Thank you. I achieved a public leaderboard score of over 0.74 and ranked 123rd. I selected the two submissions with the highest public leaderboard scores as my final entries, but their corresponding private leaderboard score was only 0.60. However, among all my submissions, there was one with a private leaderboard score of 0.61, yet its public leaderboard score was merely 0.71, which is why I didn't choose it. It’s a great pity that my competition results are now gone. I feel quite disappointed, as I made more than 140 submissions and dedicated over two months to rigorous model training.@addisonhoward @qiyuliaocsiro
", "votes": null, "replies": [ { "id": 3399253, "author_name": "addisonhoward", "author_url": "", "post_date": "01/30/2026 14:03:34", "content": "We've cleaned the leaderboard and disqualified some teams that have violated the rules. If you think you were removed by mistake, or believe you have evidence that suggests another team cheated, please contact compliance. Please fill in all the fields honestly.
", "votes": null, "replies": [] } ] } ] } ] } ] } ] }, { "id": 3398368, "author_name": "yunsuxiaozi", "author_url": "", "post_date": "01/29/2026 03:19:55", "content": "Public datasets can be overfitted to improve scores, but private datasets cannot, in order to measure the true performance of the model.
", "votes": null, "replies": [] }, { "id": 3398394, "author_name": "yiyanghu", "author_url": "", "post_date": "01/29/2026 04:07:10", "content": "Why is there such a huge gap between the public score and the private score? It must be related to the dataset, right? I've competed before, and the larger the dataset, the smaller the shake. How is the private dataset split?
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3398362": "It's suggested that the final score be calculated by combining both the Public Score and the Private Score to balance the differences in dataset distribution. What do you think?", "3398367": "Isn't the testing phase over ? This ain't the final score? Or we can have more shakeups?", "3398368": "Public datasets can be overfitted to improve scores, but private datasets cannot, in order to measure the true performance of the model.", "3398394": "Why is there such a huge gap between the public score and the private score? It must be related to the dataset, right? I've competed before, and the larger the dataset, the smaller the shake. How is the private dataset split?", "3398841": "Its done - kaggle might review for cheating, etc but no more real changes - you might improve a spot or two.", "3398983": "Are you the competition organizer?", "3398984": "May I direct my appeal regarding the revoked competition results to you?", "3399028": "Not the organizer. Just a person who's been on the site for 8 years.\n\nI saw your question in another discussion post - you likely have / will receive an email about revoked status - what was the approximate leader board position just before the end? \n\nThe email will likely tell you why. I got an email for being a rude person the other day in discussion post - the links to appeal were a bit messy but you could get there. An AI will handle the initial part mostly likely.\n\nIf no email there's a drop down on the left panel - click More and than support. in the [support](https://www.kaggle.com/contact#/account/enforcement-action) pages you can follow to appeal. But wait a day or two for the email would be my advice.", "3399075": "Thank you. I achieved a public leaderboard score of over 0.74 and ranked 123rd. I selected the two submissions with the highest public leaderboard scores as my final entries, but their corresponding private leaderboard score was only 0.60. However, among all my submissions, there was one with a private leaderboard score of 0.61, yet its public leaderboard score was merely 0.71, which is why I didn't choose it. It’s a great pity that my competition results are now gone. I feel quite disappointed, as I made more than 140 submissions and dedicated over two months to rigorous model training.@addisonhoward @qiyuliaocsiro", "3399253": "We've cleaned the leaderboard and disqualified some teams that have violated the rules. If you think you were removed by mistake, or believe you have evidence that suggests another team cheated, please contact [compliance](https://www.kaggle.com/compliance). Please fill in all the fields honestly." }, "source": "meta" }