{ "id": 588272, "title": "Prize question for hosts", "url": "/competitions/drw-crypto-market-prediction/discussion/588272", "author_name": "", "post_date": "2025-07-05T10:32:19.747153Z", "votes": 2, "comment_count": 20, "views": 0, "content": "

@drwtrading : If a top 5 team doesn't submit their code/solution (no offense intended), will the prize pass to the next highest-ranked team that does submit?
\nOur team wants to understand how this affects prize eligibility. Thanks!

", "messages": [ { "id": "3241890", "postDate": "07/05/2025 10:32:19", "content": "

@drwtrading : If a top 5 team doesn't submit their code/solution (no offense intended), will the prize pass to the next highest-ranked team that does submit?
\nOur team wants to understand how this affects prize eligibility. Thanks!

", "rawMarkdown": "drwtrading : If a top 5 team doesn't submit their code/solution (no offense intended), will the prize pass to the next highest-ranked team that does submit?\nOur team wants to understand how this affects prize eligibility. Thanks!", "votes": null }, { "id": "3241905", "postDate": "07/05/2025 10:52:10", "content": "

Why do you consider this scenario? It's too early to talk about prizes. Do you assume / know your solution will be in top?

", "rawMarkdown": "Why do you consider this scenario? It's too early to talk about prizes. Do you assume / know your solution will be in top?", "votes": null }, { "id": "3241916", "postDate": "07/05/2025 11:06:57", "content": "

This is just one consideration. If the top-ranked teams using leaked data are not disqualified (and rewards are not passed down), then there’s no point for us to continue the competition.

", "rawMarkdown": "This is just one consideration. If the top-ranked teams using leaked data are not disqualified (and rewards are not passed down), then there’s no point for us to continue the competition.", "votes": null }, { "id": "3241924", "postDate": "07/05/2025 11:12:13", "content": "

In my opinion, the private leaderboard will be cleared and verified. To claim prize, you should provide a solution. Do you score 14th place without any external dataset?

", "rawMarkdown": "In my opinion, the private leaderboard will be cleared and verified. To claim prize, you should provide a solution. Do you score 14th place without any external dataset?", "votes": null }, { "id": "3241933", "postDate": "07/05/2025 11:17:25", "content": "

We only used external datasets permitted by the competition rules, and we can provide both the solution and the code for verification

", "rawMarkdown": "We only used external datasets permitted by the competition rules, and we can provide both the solution and the code for verification", "votes": null }, { "id": "3241934", "postDate": "07/05/2025 11:17:52", "content": "

If you believe in your solution, you should continue.

", "rawMarkdown": "If you believe in your solution, you should continue.", "votes": null }, { "id": "3241943", "postDate": "07/05/2025 11:32:08", "content": "

Interesting, how do you know an external dataset is permitted? We don't know timestamps, so how to determine what is future data? I understand you have to add extra columns to both train and test data frames, but I can't imagine on what criteria.

", "rawMarkdown": "Interesting, how do you know an external dataset is permitted? We don't know timestamps, so how to determine what is future data? I understand you have to add extra columns to both train and test data frames, but I can't imagine on what criteria.", "votes": null }, { "id": "3242163", "postDate": "07/05/2025 16:38:48", "content": "

All datasets we used are publicly available for download online, and there is no future peeking. For the rest (data/methods), we may consider making them public after the competition ends.

", "rawMarkdown": "All datasets we used are publicly available for download online, and there is no future peeking. For the rest (data/methods), we may consider making them public after the competition ends.", "votes": null }, { "id": "3242714", "postDate": "07/06/2025 09:19:02", "content": "

I wonder how you recover the timestamp of the test set and how to make sure that your recovering is 100% correct. If your recovering is not 100% correct then there may be forward-looking in the test set.

", "rawMarkdown": "I wonder how you recover the timestamp of the test set and how to make sure that your recovering is 100% correct. If your recovering is not 100% correct then there may be forward-looking in the test set.", "votes": null }, { "id": "3243193", "postDate": "07/06/2025 21:14:54", "content": "

I would think the code/solution needs to be shown, otherwise there are many people that could reverse engineer the timestamp (ID) values in the test dataset or engage in other 'label hacking' methodologies.

", "rawMarkdown": "I would think the code/solution needs to be shown, otherwise there are many people that could reverse engineer the timestamp (ID) values in the test dataset or engage in other 'label hacking' methodologies.", "votes": null }, { "id": "3243195", "postDate": "07/06/2025 21:15:48", "content": "

Does your solution actually train on the train data at all? Or it exclusively uses public data?

", "rawMarkdown": "Does your solution actually train on the train data at all? Or it exclusively uses public data?", "votes": null }, { "id": "3243290", "postDate": "07/07/2025 01:47:28", "content": "

Our team is not obligated to address such unfounded questions.

", "rawMarkdown": "Our team is not obligated to address such unfounded questions.", "votes": null }, { "id": "3243292", "postDate": "07/07/2025 01:58:00", "content": "

If adding timestamps to the test cover was your own initiative/attempt, feel free to continue. We have never performed any operation to timestamp tests

", "rawMarkdown": "If adding timestamps to the test cover was your own initiative/attempt, feel free to continue. We have never performed any operation to timestamp tests", "votes": null }, { "id": "3243293", "postDate": "07/07/2025 01:58:24", "content": "

You are correct. No need to get so defensive about it, that might raise its own question.

", "rawMarkdown": "You are correct. No need to get so defensive about it, that might raise its own question.", "votes": null }, { "id": "3243401", "postDate": "07/07/2025 05:04:41", "content": "

Based on the host's utter silence on anything related to those top performing LB scores so far, I don't think they give a shift of the fairness of the game. Likely this one will disappoint everyone that adhered to the rules and invested their time and efforts. Hopefully they'll prove me wrong lol

", "rawMarkdown": "Based on the host's utter silence on anything related to those top performing LB scores so far, I don't think they give a shift of the fairness of the game. Likely this one will disappoint everyone that adhered to the rules and invested their time and efforts. Hopefully they'll prove me wrong lol", "votes": null }, { "id": "3243402", "postDate": "07/07/2025 05:10:30", "content": "

Couldn’t agree more.

", "rawMarkdown": "Couldn’t agree more.", "votes": null }, { "id": "3243517", "postDate": "07/07/2025 08:57:14", "content": "

I do think the host will review the notebook, but it will not be rigorous. If the notebook has some pretrained model that trainied on the reversed test set, they'll likely pass it off. For now, I have shift focus to another competition as this competition is likely disappointing.

", "rawMarkdown": "I do think the host will review the notebook, but it will not be rigorous. If the notebook has some pretrained model that trainied on the reversed test set, they'll likely pass it off. For now, I have shift focus to another competition as this competition is likely disappointing.", "votes": null }, { "id": "3243774", "postDate": "07/07/2025 14:09:15", "content": "

Yes. If a top-ranked team is disqualified for any reason, including failing to submit their code for validation, the prize will typically pass down to the next highest-ranked team that successfully submits a valid and reproducible solution.

", "rawMarkdown": "## **Yes. If a top-ranked team is disqualified for any reason, including failing to submit their code for validation, the prize will typically pass down to the next highest-ranked team that successfully submits a valid and reproducible solution.**", "votes": null }, { "id": "3243794", "postDate": "07/07/2025 14:25:30", "content": "

The host is probably looking for code that will help them make money in trading. That is probably why it is a regression instead of a time series competition as well. How that plays out with the awards, I'm not sure.

", "rawMarkdown": "The host is probably looking for code that will help them make money in trading. That is probably why it is a regression instead of a time series competition as well. How that plays out with the awards, I'm not sure.", "votes": null }, { "id": "3244606", "postDate": "07/08/2025 09:59:32", "content": "

@drwtrading waiting for your reply！

", "rawMarkdown": "drwtrading waiting for your reply！", "votes": null }, { "id": "3245355", "postDate": "07/09/2025 08:40:47", "content": "

最78难绷的第一是个新号 上来刷到0.95了 这下没招了 都上手法吧 都在这ridge这一块/.

", "rawMarkdown": "最78难绷的第一是个新号 上来刷到0.95了 这下没招了 都上手法吧 都在这ridge这一块/.", "votes": null } ], "comments": [ { "id": 3241905, "author_name": "jankowalski2000", "author_url": "", "post_date": "07/05/2025 10:52:10", "content": "

Why do you consider this scenario? It's too early to talk about prizes. Do you assume / know your solution will be in top?

", "votes": null, "replies": [ { "id": 3241916, "author_name": "z1493916656", "author_url": "", "post_date": "07/05/2025 11:06:57", "content": "

This is just one consideration. If the top-ranked teams using leaked data are not disqualified (and rewards are not passed down), then there’s no point for us to continue the competition.

", "votes": null, "replies": [ { "id": 3241924, "author_name": "jankowalski2000", "author_url": "", "post_date": "07/05/2025 11:12:13", "content": "

In my opinion, the private leaderboard will be cleared and verified. To claim prize, you should provide a solution. Do you score 14th place without any external dataset?

", "votes": null, "replies": [ { "id": 3241933, "author_name": "z1493916656", "author_url": "", "post_date": "07/05/2025 11:17:25", "content": "

We only used external datasets permitted by the competition rules, and we can provide both the solution and the code for verification

", "votes": null, "replies": [ { "id": 3241943, "author_name": "jankowalski2000", "author_url": "", "post_date": "07/05/2025 11:32:08", "content": "

Interesting, how do you know an external dataset is permitted? We don't know timestamps, so how to determine what is future data? I understand you have to add extra columns to both train and test data frames, but I can't imagine on what criteria.

", "votes": null, "replies": [ { "id": 3242163, "author_name": "z1493916656", "author_url": "", "post_date": "07/05/2025 16:38:48", "content": "

All datasets we used are publicly available for download online, and there is no future peeking. For the rest (data/methods), we may consider making them public after the competition ends.

", "votes": null, "replies": [ { "id": 3242714, "author_name": "peterzhoubot", "author_url": "", "post_date": "07/06/2025 09:19:02", "content": "

I wonder how you recover the timestamp of the test set and how to make sure that your recovering is 100% correct. If your recovering is not 100% correct then there may be forward-looking in the test set.

", "votes": null, "replies": [ { "id": 3243292, "author_name": "z1493916656", "author_url": "", "post_date": "07/07/2025 01:58:00", "content": "

If adding timestamps to the test cover was your own initiative/attempt, feel free to continue. We have never performed any operation to timestamp tests

", "votes": null, "replies": [] } ] }, { "id": 3243195, "author_name": "taylorsamarel", "author_url": "", "post_date": "07/06/2025 21:15:48", "content": "

Does your solution actually train on the train data at all? Or it exclusively uses public data?

", "votes": null, "replies": [ { "id": 3243290, "author_name": "z1493916656", "author_url": "", "post_date": "07/07/2025 01:47:28", "content": "

Our team is not obligated to address such unfounded questions.

", "votes": null, "replies": [ { "id": 3243293, "author_name": "taylorsamarel", "author_url": "", "post_date": "07/07/2025 01:58:24", "content": "

You are correct. No need to get so defensive about it, that might raise its own question.

", "votes": null, "replies": [] } ] } ] } ] } ] } ] } ] }, { "id": 3241934, "author_name": "jankowalski2000", "author_url": "", "post_date": "07/05/2025 11:17:52", "content": "

If you believe in your solution, you should continue.

", "votes": null, "replies": [] } ] } ] }, { "id": 3243193, "author_name": "taylorsamarel", "author_url": "", "post_date": "07/06/2025 21:14:54", "content": "

I would think the code/solution needs to be shown, otherwise there are many people that could reverse engineer the timestamp (ID) values in the test dataset or engage in other 'label hacking' methodologies.

", "votes": null, "replies": [] }, { "id": 3243401, "author_name": "tony271ynot", "author_url": "", "post_date": "07/07/2025 05:04:41", "content": "

Based on the host's utter silence on anything related to those top performing LB scores so far, I don't think they give a shift of the fairness of the game. Likely this one will disappoint everyone that adhered to the rules and invested their time and efforts. Hopefully they'll prove me wrong lol

", "votes": null, "replies": [ { "id": 3243402, "author_name": "z1493916656", "author_url": "", "post_date": "07/07/2025 05:10:30", "content": "

Couldn’t agree more.

", "votes": null, "replies": [] }, { "id": 3243794, "author_name": "taylorsamarel", "author_url": "", "post_date": "07/07/2025 14:25:30", "content": "

The host is probably looking for code that will help them make money in trading. That is probably why it is a regression instead of a time series competition as well. How that plays out with the awards, I'm not sure.

", "votes": null, "replies": [] } ] }, { "id": 3243517, "author_name": "buphmquc", "author_url": "", "post_date": "07/07/2025 08:57:14", "content": "

I do think the host will review the notebook, but it will not be rigorous. If the notebook has some pretrained model that trainied on the reversed test set, they'll likely pass it off. For now, I have shift focus to another competition as this competition is likely disappointing.

", "votes": null, "replies": [] }, { "id": 3243774, "author_name": "", "author_url": "", "post_date": "07/07/2025 14:09:15", "content": "

Yes. If a top-ranked team is disqualified for any reason, including failing to submit their code for validation, the prize will typically pass down to the next highest-ranked team that successfully submits a valid and reproducible solution.

", "votes": null, "replies": [] }, { "id": 3244606, "author_name": "z1493916656", "author_url": "", "post_date": "07/08/2025 09:59:32", "content": "

@drwtrading waiting for your reply！

", "votes": null, "replies": [] }, { "id": 3245355, "author_name": "liujianingcfec", "author_url": "", "post_date": "07/09/2025 08:40:47", "content": "

最78难绷的第一是个新号 上来刷到0.95了 这下没招了 都上手法吧 都在这ridge这一块/.

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3241890": "drwtrading : If a top 5 team doesn't submit their code/solution (no offense intended), will the prize pass to the next highest-ranked team that does submit?\nOur team wants to understand how this affects prize eligibility. Thanks!", "3241905": "Why do you consider this scenario? It's too early to talk about prizes. Do you assume / know your solution will be in top?", "3241916": "This is just one consideration. If the top-ranked teams using leaked data are not disqualified (and rewards are not passed down), then there’s no point for us to continue the competition.", "3241924": "In my opinion, the private leaderboard will be cleared and verified. To claim prize, you should provide a solution. Do you score 14th place without any external dataset?", "3241933": "We only used external datasets permitted by the competition rules, and we can provide both the solution and the code for verification", "3241934": "If you believe in your solution, you should continue.", "3241943": "Interesting, how do you know an external dataset is permitted? We don't know timestamps, so how to determine what is future data? I understand you have to add extra columns to both train and test data frames, but I can't imagine on what criteria.", "3242163": "All datasets we used are publicly available for download online, and there is no future peeking. For the rest (data/methods), we may consider making them public after the competition ends.", "3242714": "I wonder how you recover the timestamp of the test set and how to make sure that your recovering is 100% correct. If your recovering is not 100% correct then there may be forward-looking in the test set.", "3243193": "I would think the code/solution needs to be shown, otherwise there are many people that could reverse engineer the timestamp (ID) values in the test dataset or engage in other 'label hacking' methodologies.", "3243195": "Does your solution actually train on the train data at all? Or it exclusively uses public data?", "3243290": "Our team is not obligated to address such unfounded questions.", "3243292": "If adding timestamps to the test cover was your own initiative/attempt, feel free to continue. We have never performed any operation to timestamp tests", "3243293": "You are correct. No need to get so defensive about it, that might raise its own question.", "3243401": "Based on the host's utter silence on anything related to those top performing LB scores so far, I don't think they give a shift of the fairness of the game. Likely this one will disappoint everyone that adhered to the rules and invested their time and efforts. Hopefully they'll prove me wrong lol", "3243402": "Couldn’t agree more.", "3243517": "I do think the host will review the notebook, but it will not be rigorous. If the notebook has some pretrained model that trainied on the reversed test set, they'll likely pass it off. For now, I have shift focus to another competition as this competition is likely disappointing.", "3243774": "## **Yes. If a top-ranked team is disqualified for any reason, including failing to submit their code for validation, the prize will typically pass down to the next highest-ranked team that successfully submits a valid and reproducible solution.**", "3243794": "The host is probably looking for code that will help them make money in trading. That is probably why it is a regression instead of a time series competition as well. How that plays out with the awards, I'm not sure.", "3244606": "drwtrading waiting for your reply！", "3245355": "最78难绷的第一是个新号 上来刷到0.95了 这下没招了 都上手法吧 都在这ridge这一块/." }, "source": "meta" }