Hi @kaggleteam,
\n\nIn Liverpool Ion Switching competition, we were having the same number of digits on LB, as we have in Alaska2 Steganalysis. Because of this, it was hard to track what was happening in the competition. Same is the case here as well.
\n\nSo, kindly can we have more digits on LB here while the competition is going on. In Liverpool Ion switching, you guys gave more digits after competition end which was not useful info to make a better decision. Here is the link to that LB.
\n\nCurrently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. One more point is that private LB will be based on another 80% of the data, hence knowing how much our submission scores on public LB with some more digits will help in deciding which submissions to choose for final scores.
\n\nAny thoughts?
\n\nThanks.
", "messages": [ { "id": 906278, "postDate": "2020-06-29T07:19:29.807Z", "content": "> Currently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there.
\n\nAFAIK, If you sort My Submissions by Public Score, they are listed best to worst, even for those with the same 3 digit score
", "rawMarkdown": "> Currently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. \n\nAFAIK, If you sort My Submissions by Public Score, they are listed best to worst, even for those with the same 3 digit score ", "votes": 5, "replies": [ { "id": 906337, "postDate": "2020-06-29T07:56:47.077Z", "content": "Did not know that LOL. But, still, I feel there can be one more or two-digit on the LB for a better understanding of scores.
", "rawMarkdown": "Did not know that LOL. But, still, I feel there can be one more or two-digit on the LB for a better understanding of scores. " }, { "id": 906418, "postDate": "2020-06-29T09:30:44.257Z", "content": "The kaggle policy appears to be only open it up near competition end.\n> We intentionally restrict the number of decimal places to (1) discourage leaderboard probing, and (2) encourage local validation.
", "rawMarkdown": "The [kaggle policy](https://www.kaggle.com/c/understanding_cloud_organization/discussion/115834#667720) appears to be only open it up near competition end.\n> We intentionally restrict the number of decimal places to (1) discourage leaderboard probing, and (2) encourage local validation.\n", "votes": 1 }, { "id": 906431, "postDate": "2020-06-29T09:38:38.263Z", "content": "Got it. So probably, there is no way we will see more digits before the end.
", "rawMarkdown": "Got it. So probably, there is no way we will see more digits before the end. " }, { "id": 906437, "postDate": "2020-06-29T09:41:37.147Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 927186, "postDate": "2020-07-13T08:45:30.193Z", "content": "@addisonhoward Time to add a couple more digits? 😺
", "rawMarkdown": "@addisonhoward Time to add a couple more digits? 😺 ", "votes": 1 }, { "id": 906336, "postDate": "2020-06-29T07:55:03.753Z", "content": "@robga is correct, however, I guess nobody would mind it if another digit is added to the LB
", "rawMarkdown": "@robga is correct, however, I guess nobody would mind it if another digit is added to the LB ", "votes": 1, "replies": [ { "id": 906340, "postDate": "2020-06-29T07:58:25.890Z", "content": "yes, it will be better to understand the scores I think
", "rawMarkdown": "yes, it will be better to understand the scores I think " }, { "id": 906382, "postDate": "2020-06-29T08:47:07.677Z", "content": "Nice jump, btw! Try to go under 50!
", "rawMarkdown": "Nice jump, btw! Try to go under 50!", "votes": 1 }, { "id": 906425, "postDate": "2020-06-29T09:35:11.823Z", "content": "Thanks, I will try my best.
", "rawMarkdown": "Thanks, I will try my best. ", "votes": 1 } ] }, { "id": 906257, "postDate": "2020-06-29T06:50:39.620Z", "content": "Hi @kaggleteam,
\n\nIn Liverpool Ion Switching competition, we were having the same number of digits on LB, as we have in Alaska2 Steganalysis. Because of this, it was hard to track what was happening in the competition. Same is the case here as well.
\n\nSo, kindly can we have more digits on LB here while the competition is going on. In Liverpool Ion switching, you guys gave more digits after competition end which was not useful info to make a better decision. Here is the link to that LB.
\n\nCurrently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. One more point is that private LB will be based on another 80% of the data, hence knowing how much our submission scores on public LB with some more digits will help in deciding which submissions to choose for final scores.
\n\nAny thoughts?
\n\nThanks.
", "rawMarkdown": "Hi @kaggleteam, \n\nIn [Liverpool Ion Switching competition](https://www.kaggle.com/c/liverpool-ion-switching), we were having the same number of digits on LB, as we have in Alaska2 Steganalysis. Because of this, it was hard to track what was happening in the competition. Same is the case here as well. \n\nSo, kindly can we have more digits on LB here while the competition is going on. In Liverpool Ion switching, you guys gave more digits after competition end which was not useful info to make a better decision. Here is the [link](https://www.kaggle.com/c/liverpool-ion-switching/leaderboard) to that LB.\n\nCurrently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. One more point is that private LB will be based on another 80% of the data, hence knowing how much our submission scores on public LB with some more digits will help in deciding which submissions to choose for final scores. \n\nAny thoughts? \n\nThanks." } ], "comments": [ { "id": 906278, "author_name": "robga", "author_url": "", "post_date": "2020-06-29T07:19:29.807000", "content": "> Currently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there.
\n\nAFAIK, If you sort My Submissions by Public Score, they are listed best to worst, even for those with the same 3 digit score
", "votes": 5, "replies": [ { "id": 906337, "author_name": "Urvish", "author_url": "", "post_date": "2020-06-29T07:56:47.077000", "content": "Did not know that LOL. But, still, I feel there can be one more or two-digit on the LB for a better understanding of scores.
", "votes": 0, "replies": [] }, { "id": 906418, "author_name": "robga", "author_url": "", "post_date": "2020-06-29T09:30:44.257000", "content": "The kaggle policy appears to be only open it up near competition end.\n> We intentionally restrict the number of decimal places to (1) discourage leaderboard probing, and (2) encourage local validation.
", "votes": 1, "replies": [] }, { "id": 906431, "author_name": "Urvish", "author_url": "", "post_date": "2020-06-29T09:38:38.263000", "content": "Got it. So probably, there is no way we will see more digits before the end.
", "votes": 0, "replies": [] }, { "id": 906437, "author_name": "", "author_url": "", "post_date": "2020-06-29T09:41:37.147000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 927186, "author_name": "robga", "author_url": "", "post_date": "2020-07-13T08:45:30.193000", "content": "@addisonhoward Time to add a couple more digits? 😺
", "votes": 1, "replies": [] }, { "id": 906336, "author_name": "Mighty Rains", "author_url": "", "post_date": "2020-06-29T07:55:03.753000", "content": "@robga is correct, however, I guess nobody would mind it if another digit is added to the LB
", "votes": 1, "replies": [ { "id": 906340, "author_name": "Urvish", "author_url": "", "post_date": "2020-06-29T07:58:25.890000", "content": "yes, it will be better to understand the scores I think
", "votes": 0, "replies": [] }, { "id": 906382, "author_name": "Mighty Rains", "author_url": "", "post_date": "2020-06-29T08:47:07.677000", "content": "Nice jump, btw! Try to go under 50!
", "votes": 1, "replies": [] }, { "id": 906425, "author_name": "Urvish", "author_url": "", "post_date": "2020-06-29T09:35:11.823000", "content": "Thanks, I will try my best.
", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "906278": "> Currently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. \n\nAFAIK, If you sort My Submissions by Public Score, they are listed best to worst, even for those with the same 3 digit score ", "927186": "@addisonhoward Time to add a couple more digits? 😺 ", "906336": "@robga is correct, however, I guess nobody would mind it if another digit is added to the LB ", "906257": "Hi @kaggleteam, \n\nIn [Liverpool Ion Switching competition](https://www.kaggle.com/c/liverpool-ion-switching), we were having the same number of digits on LB, as we have in Alaska2 Steganalysis. Because of this, it was hard to track what was happening in the competition. Same is the case here as well. \n\nSo, kindly can we have more digits on LB here while the competition is going on. In Liverpool Ion switching, you guys gave more digits after competition end which was not useful info to make a better decision. Here is the [link](https://www.kaggle.com/c/liverpool-ion-switching/leaderboard) to that LB.\n\nCurrently, with the same score, I got LB boost from 300 to 126 and I do not know which submission got me there. One more point is that private LB will be based on another 80% of the data, hence knowing how much our submission scores on public LB with some more digits will help in deciding which submissions to choose for final scores. \n\nAny thoughts? \n\nThanks." } }