{ "id": 519811, "title": "👀 It's time for shakeup 🌀", "url": "/competitions/leap-atmospheric-physics-ai-climsim/discussion/519811", "author_name": "mh", "post_date": "2024-07-13T01:49:04.251000", "votes": -12, "comment_count": 8, "views": 0, "content": "

This topic is not discussed much, but I think there is a pretty big shakeup considering that 80% of the test data is calculated with private LB.
\nHow can we use 2 subs effectively?

", "messages": [ { "id": 2919540, "postDate": "2024-07-13T01:59:12.563Z", "content": "

Some selects leak, and others do not. It would be the shakeup.

", "rawMarkdown": "Some selects leak, and others do not. It would be the shakeup.", "votes": 5, "replies": [ { "id": 2919667, "postDate": "2024-07-13T05:30:08.683Z", "content": "

what will happen if select leak and no leak solutions? Does it mean hosts will disqualify team because the leak usage? even though there is exist selected solution with no leak? I mean team is using leak data as fact,and stuff or hosts already doesnt trust this team no matter what and thats affecting leaderboard, some teams not use leak submission, ambiguity exist…

", "rawMarkdown": "what will happen if select leak and no leak solutions? Does it mean hosts will disqualify team because the leak usage? even though there is exist selected solution with no leak? I mean team is using leak data as fact,and stuff or hosts already doesnt trust this team no matter what and thats affecting leaderboard, some teams not use leak submission, ambiguity exist...", "votes": 1, "replies": [ { "id": 2921914, "postDate": "2024-07-14T18:27:46.860Z", "content": "

It doesn't make sense that people are still trying leaky solutions given the risk of disqualification and removal from the leaderboard.

", "rawMarkdown": "It doesn't make sense that people are still trying leaky solutions given the risk of disqualification and removal from the leaderboard.", "votes": 5 }, { "id": 2921958, "postDate": "2024-07-14T19:11:19.983Z", "content": "

@jerrylin96 - some people are asking for a reason (theyre likely breaking the competition rules)

", "rawMarkdown": "@jerrylin96 - some people are asking for a reason (theyre likely breaking the competition rules)" }, { "id": 2921969, "postDate": "2024-07-14T19:18:34.950Z", "content": "

@jerrylin96 Can you work with Kaggle to do a random spot-check of a number of submission in silver and bronze or a limitted targetted check of those who have indicated (or suggested) they are using the leak? This should be enough to disincentivize cheating. Some will still take the risk, at least this gives participants abiding by the rules a more level playing field.

", "rawMarkdown": "@jerrylin96 Can you work with Kaggle to do a random spot-check of a number of submission in silver and bronze or a limitted targetted check of those who have indicated (or suggested) they are using the leak? This should be enough to disincentivize cheating. Some will still take the risk, at least this gives participants abiding by the rules a more level playing field.", "votes": 3 } ] } ] }, { "id": 2919532, "postDate": "2024-07-13T01:50:34.047Z", "content": "

No, there won't be a huge shakeup. Unless huge amounts of participants discard their morals and use the leak.

", "rawMarkdown": "No, there won't be a huge shakeup. Unless huge amounts of participants discard their morals and use the leak.", "votes": 5 }, { "id": 2919705, "postDate": "2024-07-13T06:31:44.403Z", "content": "

I don't think the shake up will be big either but since the scores of top teams are so close to each other, there might be +-5 change in their positions.

", "rawMarkdown": "I don't think the shake up will be big either but since the scores of top teams are so close to each other, there might be +-5 change in their positions.", "votes": 1 }, { "id": 2919531, "postDate": "2024-07-13T01:49:04.250Z", "content": "

This topic is not discussed much, but I think there is a pretty big shakeup considering that 80% of the test data is calculated with private LB.
\nHow can we use 2 subs effectively?

", "rawMarkdown": "This topic is not discussed much, but I think there is a pretty big shakeup considering that 80% of the test data is calculated with private LB.\nHow can we use 2 subs effectively?", "votes": -12 }, { "id": 2921313, "postDate": "2024-07-14T10:07:34.933Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 2919540, "author_name": "ONODERA", "author_url": "", "post_date": "2024-07-13T01:59:12.563000", "content": "

Some selects leak, and others do not. It would be the shakeup.

", "votes": 5, "replies": [ { "id": 2919667, "author_name": "VITALIY", "author_url": "", "post_date": "2024-07-13T05:30:08.683000", "content": "

what will happen if select leak and no leak solutions? Does it mean hosts will disqualify team because the leak usage? even though there is exist selected solution with no leak? I mean team is using leak data as fact,and stuff or hosts already doesnt trust this team no matter what and thats affecting leaderboard, some teams not use leak submission, ambiguity exist…

", "votes": 1, "replies": [ { "id": 2921914, "author_name": "Jerry Lin", "author_url": "", "post_date": "2024-07-14T18:27:46.860000", "content": "

It doesn't make sense that people are still trying leaky solutions given the risk of disqualification and removal from the leaderboard.

", "votes": 5, "replies": [] }, { "id": 2921958, "author_name": "Sean McManus", "author_url": "", "post_date": "2024-07-14T19:11:19.983000", "content": "

@jerrylin96 - some people are asking for a reason (theyre likely breaking the competition rules)

", "votes": 0, "replies": [] }, { "id": 2921969, "author_name": "Sean McManus", "author_url": "", "post_date": "2024-07-14T19:18:34.950000", "content": "

@jerrylin96 Can you work with Kaggle to do a random spot-check of a number of submission in silver and bronze or a limitted targetted check of those who have indicated (or suggested) they are using the leak? This should be enough to disincentivize cheating. Some will still take the risk, at least this gives participants abiding by the rules a more level playing field.

", "votes": 3, "replies": [] } ] } ] }, { "id": 2919532, "author_name": "ForcewithMe", "author_url": "", "post_date": "2024-07-13T01:50:34.047000", "content": "

No, there won't be a huge shakeup. Unless huge amounts of participants discard their morals and use the leak.

", "votes": 5, "replies": [] }, { "id": 2919705, "author_name": "Gunes Evitan", "author_url": "", "post_date": "2024-07-13T06:31:44.403000", "content": "

I don't think the shake up will be big either but since the scores of top teams are so close to each other, there might be +-5 change in their positions.

", "votes": 1, "replies": [] }, { "id": 2921313, "author_name": "", "author_url": "", "post_date": "2024-07-14T10:07:34.933000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2919540": "Some selects leak, and others do not. It would be the shakeup.", "2919532": "No, there won't be a huge shakeup. Unless huge amounts of participants discard their morals and use the leak.", "2919705": "I don't think the shake up will be big either but since the scores of top teams are so close to each other, there might be +-5 change in their positions.", "2919531": "This topic is not discussed much, but I think there is a pretty big shakeup considering that 80% of the test data is calculated with private LB.\nHow can we use 2 subs effectively?", "2921313": "" } }