{
  "id": 65303,
  "title": "Simple model removed from the leaderboard?",
  "url": "/competitions/rsna-pneumonia-detection-challenge/discussion/65303",
  "author_name": "",
  "post_date": "2018-09-09T01:12:53.658502100Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Also, why are the current submissions achieving such a relatively low LB score? Is there a potential data leak?</p>",
  "messages": [
    {
      "id": "383563",
      "postDate": "09/09/2018 01:12:53",
      "content": "<p>Also, why are the current submissions achieving such a relatively low LB score? Is there a potential data leak?</p>",
      "rawMarkdown": "Also, why are the current submissions achieving such a relatively low LB score? Is there a potential data leak?",
      "votes": null
    },
    {
      "id": "383602",
      "postDate": "09/09/2018 04:04:28",
      "content": "<p>I thinking, if there is a data leak, our submition score will be relatevly hight</p>",
      "rawMarkdown": "I thinking, if there is a data leak, our submition score will be relatevly hight",
      "votes": null
    },
    {
      "id": "385307",
      "postDate": "09/10/2018 18:47:38",
      "content": "<p>If there were a leak, the LB scores would probably be better.</p>\n\n<p>This is a difficult problem.  It's possible to do well.  I removed the simple model benchmark because it was built using an incorrect portion of the data set.  We'll be replacing the benchmark with another shortly.</p>",
      "rawMarkdown": "If there were a leak, the LB scores would probably be better.\n\nThis is a difficult problem.  It's possible to do well.  I removed the simple model benchmark because it was built using an incorrect portion of the data set.  We'll be replacing the benchmark with another shortly.",
      "votes": null
    },
    {
      "id": "385350",
      "postDate": "09/10/2018 20:21:59",
      "content": "<p>Thanks for the update.</p>",
      "rawMarkdown": "Thanks for the update.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 383602,
      "author_name": "koza4ukdmitrij",
      "author_url": "",
      "post_date": "09/09/2018 04:04:28",
      "content": "<p>I thinking, if there is a data leak, our submition score will be relatevly hight</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 385307,
      "author_name": "philculliton",
      "author_url": "",
      "post_date": "09/10/2018 18:47:38",
      "content": "<p>If there were a leak, the LB scores would probably be better.</p>\n\n<p>This is a difficult problem.  It's possible to do well.  I removed the simple model benchmark because it was built using an incorrect portion of the data set.  We'll be replacing the benchmark with another shortly.</p>",
      "votes": null,
      "replies": [
        {
          "id": 385350,
          "author_name": "taindow",
          "author_url": "",
          "post_date": "09/10/2018 20:21:59",
          "content": "<p>Thanks for the update.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "383563": "Also, why are the current submissions achieving such a relatively low LB score? Is there a potential data leak?",
    "383602": "I thinking, if there is a data leak, our submition score will be relatevly hight",
    "385307": "If there were a leak, the LB scores would probably be better.\n\nThis is a difficult problem.  It's possible to do well.  I removed the simple model benchmark because it was built using an incorrect portion of the data set.  We'll be replacing the benchmark with another shortly.",
    "385350": "Thanks for the update."
  },
  "source": "meta"
}