{
  "id": 169242,
  "title": "New here on Kaggle...general Query",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/169242",
  "author_name": "",
  "post_date": "2020-07-23T10:19:54.075998800Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I  am wondering why is this a competition at all ,if teams can easily run similar models with almost similar accuracies ?</p>",
  "messages": [
    {
      "id": "941611",
      "postDate": "07/23/2020 10:19:54",
      "content": "<p>I  am wondering why is this a competition at all ,if teams can easily run similar models with almost similar accuracies ?</p>",
      "rawMarkdown": "I  am wondering why is this a competition at all ,if teams can easily run similar models with almost similar accuracies ?",
      "votes": null
    },
    {
      "id": "941694",
      "postDate": "07/23/2020 11:02:23",
      "content": "<p>Just means it's a close competition at the moment. In the handful of competitions I've followed from start to finish (and my impression reading discussions on already finished competitions), the scores tend to move in waves (both at the top and in the middle). The top ones it's probably because better solutions take time to develop, the middle ones it's often because someone shares a public kernel with a new idea - that's why towards the end of competitions we are discouraged from posting very high scoring kernels because people can jump in and run the high scoring kernel for a score they have 'deserved' less than someone who has put time into developing a slightly worse solution.</p>",
      "rawMarkdown": "Just means it's a close competition at the moment. In the handful of competitions I've followed from start to finish (and my impression reading discussions on already finished competitions), the scores tend to move in waves (both at the top and in the middle). The top ones it's probably because better solutions take time to develop, the middle ones it's often because someone shares a public kernel with a new idea - that's why towards the end of competitions we are discouraged from posting very high scoring kernels because people can jump in and run the high scoring kernel for a score they have 'deserved' less than someone who has put time into developing a slightly worse solution.",
      "votes": null
    },
    {
      "id": "941697",
      "postDate": "07/23/2020 11:03:51",
      "content": "<p>The cool thing is that the competition isn't solely about the scores! We also get votes for sharing kernels and having discussions.</p>\n\n<p>I've always found it amazing that some of the people with the best scores are the biggest sharers of ideas and code. </p>",
      "rawMarkdown": "The cool thing is that the competition isn't solely about the scores! We also get votes for sharing kernels and having discussions.\n\nI've always found it amazing that some of the people with the best scores are the biggest sharers of ideas and code.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 941694,
      "author_name": "jameschapman19",
      "author_url": "",
      "post_date": "07/23/2020 11:02:23",
      "content": "<p>Just means it's a close competition at the moment. In the handful of competitions I've followed from start to finish (and my impression reading discussions on already finished competitions), the scores tend to move in waves (both at the top and in the middle). The top ones it's probably because better solutions take time to develop, the middle ones it's often because someone shares a public kernel with a new idea - that's why towards the end of competitions we are discouraged from posting very high scoring kernels because people can jump in and run the high scoring kernel for a score they have 'deserved' less than someone who has put time into developing a slightly worse solution.</p>",
      "votes": null,
      "replies": [
        {
          "id": 941697,
          "author_name": "jameschapman19",
          "author_url": "",
          "post_date": "07/23/2020 11:03:51",
          "content": "<p>The cool thing is that the competition isn't solely about the scores! We also get votes for sharing kernels and having discussions.</p>\n\n<p>I've always found it amazing that some of the people with the best scores are the biggest sharers of ideas and code. </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "941611": "I  am wondering why is this a competition at all ,if teams can easily run similar models with almost similar accuracies ?",
    "941694": "Just means it's a close competition at the moment. In the handful of competitions I've followed from start to finish (and my impression reading discussions on already finished competitions), the scores tend to move in waves (both at the top and in the middle). The top ones it's probably because better solutions take time to develop, the middle ones it's often because someone shares a public kernel with a new idea - that's why towards the end of competitions we are discouraged from posting very high scoring kernels because people can jump in and run the high scoring kernel for a score they have 'deserved' less than someone who has put time into developing a slightly worse solution.",
    "941697": "The cool thing is that the competition isn't solely about the scores! We also get votes for sharing kernels and having discussions.\n\nI've always found it amazing that some of the people with the best scores are the biggest sharers of ideas and code."
  },
  "source": "meta"
}