{
  "id": 107998,
  "title": "Is this game wrong or am I wrong?",
  "url": "/competitions/aptos2019-blindness-detection/discussion/107998",
  "author_name": "",
  "post_date": "2019-09-08T11:28:53.495000",
  "votes": 13,
  "comment_count": 2,
  "views": 0,
  "content": "<p>First of all, Congratulations to the winners and golds!I am very honored to be able to compete with you on the same stage.\n@wowfattie @firenero @aispiriants @artemis874 @lanjunyelan @dxchen @huangzihan @garybios @haqishen @bloodaxe @selimsef</p>\n\n<p>I am not complaining because I have learned a lot from this competition.</p>\n\n<p>BUT I am very confused now and hope to gain some experience from the failures.\nI am a novice player, this is my first cv game, I have worked hard for 1 month. After the results of the competition are announced, I jumped from 2% to 38%.\nI admit that I may have over fitted public testset. But I think the gap between public testset and private should not be too large, otherwise what is the significance of the existence of public testset.\nI noticed that many of the teams in the front row were not affected by the earthquake. I think it are not random but because of strength. May  I  ask this question what should I do, or I am lacking what is causing the current situation. I don't believe in the public testset, what should I believe in this game?Is it because I am not lucky?I would be grateful if you could enlighten me.</p>",
  "messages": [
    {
      "id": 621251,
      "postDate": "2019-09-08T11:28:53.497Z",
      "content": "<p>First of all, Congratulations to the winners and golds!I am very honored to be able to compete with you on the same stage.\n@wowfattie @firenero @aispiriants @artemis874 @lanjunyelan @dxchen @huangzihan @garybios @haqishen @bloodaxe @selimsef</p>\n\n<p>I am not complaining because I have learned a lot from this competition.</p>\n\n<p>BUT I am very confused now and hope to gain some experience from the failures.\nI am a novice player, this is my first cv game, I have worked hard for 1 month. After the results of the competition are announced, I jumped from 2% to 38%.\nI admit that I may have over fitted public testset. But I think the gap between public testset and private should not be too large, otherwise what is the significance of the existence of public testset.\nI noticed that many of the teams in the front row were not affected by the earthquake. I think it are not random but because of strength. May  I  ask this question what should I do, or I am lacking what is causing the current situation. I don't believe in the public testset, what should I believe in this game?Is it because I am not lucky?I would be grateful if you could enlighten me.</p>",
      "rawMarkdown": "\n\nFirst of all, Congratulations to the winners and golds!I am very honored to be able to compete with you on the same stage.\n@wowfattie @firenero @aispiriants @artemis874 @lanjunyelan @dxchen @huangzihan @garybios @haqishen @bloodaxe @selimsef\n\nI am not complaining because I have learned a lot from this competition.\n\nBUT I am very confused now and hope to gain some experience from the failures.\nI am a novice player, this is my first cv game, I have worked hard for 1 month. After the results of the competition are announced, I jumped from 2% to 38%.\nI admit that I may have over fitted public testset. But I think the gap between public testset and private should not be too large, otherwise what is the significance of the existence of public testset.\nI noticed that many of the teams in the front row were not affected by the earthquake. I think it are not random but because of strength. May  I  ask this question what should I do, or I am lacking what is causing the current situation. I don't believe in the public testset, what should I believe in this game?Is it because I am not lucky?I would be grateful if you could enlighten me.\n",
      "votes": 13
    },
    {
      "id": 624295,
      "postDate": "2019-09-12T00:02:43.420Z",
      "content": "<p>I fully agree.\nMany people devised pre-processing and tried to make a robust model.However, it seems that the effort was not very important in this competition.</p>\n\n<p>I thought it would be difficult to work on the Synchronous KernelOnly competition, where we can't see private tests.</p>",
      "rawMarkdown": "I fully agree.\nMany people devised pre-processing and tried to make a robust model.However, it seems that the effort was not very important in this competition.\n\nI thought it would be difficult to work on the Synchronous KernelOnly competition, where we can't see private tests."
    },
    {
      "id": 621317,
      "postDate": "2019-09-08T12:36:23.477Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 624295,
      "author_name": "currypurin",
      "author_url": "",
      "post_date": "2019-09-12T00:02:43.420000",
      "content": "<p>I fully agree.\nMany people devised pre-processing and tried to make a robust model.However, it seems that the effort was not very important in this competition.</p>\n\n<p>I thought it would be difficult to work on the Synchronous KernelOnly competition, where we can't see private tests.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 621317,
      "author_name": "",
      "author_url": "",
      "post_date": "2019-09-08T12:36:23.477000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "621251": "\n\nFirst of all, Congratulations to the winners and golds!I am very honored to be able to compete with you on the same stage.\n@wowfattie @firenero @aispiriants @artemis874 @lanjunyelan @dxchen @huangzihan @garybios @haqishen @bloodaxe @selimsef\n\nI am not complaining because I have learned a lot from this competition.\n\nBUT I am very confused now and hope to gain some experience from the failures.\nI am a novice player, this is my first cv game, I have worked hard for 1 month. After the results of the competition are announced, I jumped from 2% to 38%.\nI admit that I may have over fitted public testset. But I think the gap between public testset and private should not be too large, otherwise what is the significance of the existence of public testset.\nI noticed that many of the teams in the front row were not affected by the earthquake. I think it are not random but because of strength. May  I  ask this question what should I do, or I am lacking what is causing the current situation. I don't believe in the public testset, what should I believe in this game?Is it because I am not lucky?I would be grateful if you could enlighten me.\n",
    "624295": "I fully agree.\nMany people devised pre-processing and tried to make a robust model.However, it seems that the effort was not very important in this competition.\n\nI thought it would be difficult to work on the Synchronous KernelOnly competition, where we can't see private tests.",
    "621317": ""
  }
}