{
  "id": 98110,
  "title": "Please remove kaggle points and medals from this competition",
  "url": "/competitions/generative-dog-images/discussion/98110",
  "author_name": "",
  "post_date": "2019-07-01T10:31:01.654436100Z",
  "votes": 15,
  "comment_count": 12,
  "views": 0,
  "content": "<p>Title says it all. Reasons are everywhere!</p>",
  "messages": [
    {
      "id": "565750",
      "postDate": "07/01/2019 10:31:01",
      "content": "<p>Title says it all. Reasons are everywhere!</p>",
      "rawMarkdown": "Title says it all. Reasons are everywhere!",
      "votes": null
    },
    {
      "id": "565760",
      "postDate": "07/01/2019 10:43:30",
      "content": "<p>There were far worse competitions than this one. For example google analytics from the last year...</p>",
      "rawMarkdown": "There were far worse competitions than this one. For example google analytics from the last year...",
      "votes": null
    },
    {
      "id": "565782",
      "postDate": "07/01/2019 11:27:49",
      "content": "<p>Resolving the problems due to fake submissions is fair but it would be a shame to remove kaggle points from the competition I think</p>",
      "rawMarkdown": "Resolving the problems due to fake submissions is fair but it would be a shame to remove kaggle points from the competition I think",
      "votes": null
    },
    {
      "id": "565789",
      "postDate": "07/01/2019 11:36:11",
      "content": "<p>That's reasonable but will significantly decrease the incentives for stress-testing of GAN competition design</p>",
      "rawMarkdown": "That's reasonable but will significantly decrease the incentives for stress-testing of GAN competition design",
      "votes": null
    },
    {
      "id": "565891",
      "postDate": "07/01/2019 14:01:15",
      "content": "<p>Close to this one, here and there noise is cornerstone</p>\n\n<p>Here you generate pictures from the noise</p>\n\n<p>In Google analytics you could submit noise and end up in silver </p>",
      "rawMarkdown": "Close to this one, here and there noise is cornerstone\n\nHere you generate pictures from the noise\n\nIn Google analytics you could submit noise and end up in silver",
      "votes": null
    },
    {
      "id": "566007",
      "postDate": "07/01/2019 17:31:07",
      "content": "<p>An alternative approach could be: to drop all the rules except for kernel limitations and to encourage solution sharing after the competition, for example, by giving the money prize for the most oustanding one (or several). Thus, Kaggle will collect all the possible loopholes and will come up with countermeasures for the next competition. Doing it on the fly seems to be painful and ineffective.</p>",
      "rawMarkdown": "An alternative approach could be: to drop all the rules except for kernel limitations and to encourage solution sharing after the competition, for example, by giving the money prize for the most oustanding one (or several). Thus, Kaggle will collect all the possible loopholes and will come up with countermeasures for the next competition. Doing it on the fly seems to be painful and ineffective.",
      "votes": null
    },
    {
      "id": "566588",
      "postDate": "07/02/2019 10:53:54",
      "content": "<p>Give competition medals and points only for publicly available kernels, it could solve the problem of cheating. </p>",
      "rawMarkdown": "Give competition medals and points only for publicly available kernels, it could solve the problem of cheating.",
      "votes": null
    },
    {
      "id": "566606",
      "postDate": "07/02/2019 11:05:44",
      "content": "<p>Do you mean made public after the end of the competition? It would be good.</p>",
      "rawMarkdown": "Do you mean made public after the end of the competition? It would be good.",
      "votes": null
    },
    {
      "id": "566630",
      "postDate": "07/02/2019 11:44:12",
      "content": "<p>Yup, all \"medals\" solutions must be published in order to receive points and medals. I think it is in consistent with the very idea of research competition (i.e. many new ideas could be found even in not top scores solutions). Kernel section would be flooded, but I don't think it is a big problem for such small competition. </p>",
      "rawMarkdown": "Yup, all \"medals\" solutions must be published in order to receive points and medals. I think it is in consistent with the very idea of research competition (i.e. many new ideas could be found even in not top scores solutions). Kernel section would be flooded, but I don't think it is a big problem for such small competition.",
      "votes": null
    },
    {
      "id": "566632",
      "postDate": "07/02/2019 11:45:42",
      "content": "<p>Agreed. We may repeat the success of \"hackathons\" competitions such as \"Data Science for Good\".</p>",
      "rawMarkdown": "Agreed. We may repeat the success of \"hackathons\" competitions such as \"Data Science for Good\".",
      "votes": null
    },
    {
      "id": "566850",
      "postDate": "07/02/2019 16:51:27",
      "content": "<p>What i am feeling that competition is competition. And rules will be also defined by the host of competition. Here the prize money and medal will motivate many to create something new in GAN or it might be some new method or new algorithm.  </p>",
      "rawMarkdown": "What i am feeling that competition is competition. And rules will be also defined by the host of competition. Here the prize money and medal will motivate many to create something new in GAN or it might be some new method or new algorithm.",
      "votes": null
    },
    {
      "id": "568679",
      "postDate": "07/05/2019 09:53:22",
      "content": "<p>Lol, we really want X5 to dominate this competition! :-p</p>",
      "rawMarkdown": "Lol, we really want X5 to dominate this competition! :-p",
      "votes": null
    },
    {
      "id": "568733",
      "postDate": "07/05/2019 10:50:55",
      "content": "<p>Appreciate your desire</p>",
      "rawMarkdown": "Appreciate your desire",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 565760,
      "author_name": "artgor",
      "author_url": "",
      "post_date": "07/01/2019 10:43:30",
      "content": "<p>There were far worse competitions than this one. For example google analytics from the last year...</p>",
      "votes": null,
      "replies": [
        {
          "id": 565891,
          "author_name": "venheads",
          "author_url": "",
          "post_date": "07/01/2019 14:01:15",
          "content": "<p>Close to this one, here and there noise is cornerstone</p>\n\n<p>Here you generate pictures from the noise</p>\n\n<p>In Google analytics you could submit noise and end up in silver </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 568679,
          "author_name": "projdev",
          "author_url": "",
          "post_date": "07/05/2019 09:53:22",
          "content": "<p>Lol, we really want X5 to dominate this competition! :-p</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 568733,
          "author_name": "venheads",
          "author_url": "",
          "post_date": "07/05/2019 10:50:55",
          "content": "<p>Appreciate your desire</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 565782,
      "author_name": "francoisdubois",
      "author_url": "",
      "post_date": "07/01/2019 11:27:49",
      "content": "<p>Resolving the problems due to fake submissions is fair but it would be a shame to remove kaggle points from the competition I think</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 565789,
      "author_name": "",
      "author_url": "",
      "post_date": "07/01/2019 11:36:11",
      "content": "<p>That's reasonable but will significantly decrease the incentives for stress-testing of GAN competition design</p>",
      "votes": null,
      "replies": [
        {
          "id": 566007,
          "author_name": "",
          "author_url": "",
          "post_date": "07/01/2019 17:31:07",
          "content": "<p>An alternative approach could be: to drop all the rules except for kernel limitations and to encourage solution sharing after the competition, for example, by giving the money prize for the most oustanding one (or several). Thus, Kaggle will collect all the possible loopholes and will come up with countermeasures for the next competition. Doing it on the fly seems to be painful and ineffective.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 566632,
          "author_name": "",
          "author_url": "",
          "post_date": "07/02/2019 11:45:42",
          "content": "<p>Agreed. We may repeat the success of \"hackathons\" competitions such as \"Data Science for Good\".</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 566588,
      "author_name": "",
      "author_url": "",
      "post_date": "07/02/2019 10:53:54",
      "content": "<p>Give competition medals and points only for publicly available kernels, it could solve the problem of cheating. </p>",
      "votes": null,
      "replies": [
        {
          "id": 566606,
          "author_name": "francoisdubois",
          "author_url": "",
          "post_date": "07/02/2019 11:05:44",
          "content": "<p>Do you mean made public after the end of the competition? It would be good.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 566630,
          "author_name": "",
          "author_url": "",
          "post_date": "07/02/2019 11:44:12",
          "content": "<p>Yup, all \"medals\" solutions must be published in order to receive points and medals. I think it is in consistent with the very idea of research competition (i.e. many new ideas could be found even in not top scores solutions). Kernel section would be flooded, but I don't think it is a big problem for such small competition. </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 566850,
      "author_name": "dataraj",
      "author_url": "",
      "post_date": "07/02/2019 16:51:27",
      "content": "<p>What i am feeling that competition is competition. And rules will be also defined by the host of competition. Here the prize money and medal will motivate many to create something new in GAN or it might be some new method or new algorithm.  </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "565750": "Title says it all. Reasons are everywhere!",
    "565760": "There were far worse competitions than this one. For example google analytics from the last year...",
    "565782": "Resolving the problems due to fake submissions is fair but it would be a shame to remove kaggle points from the competition I think",
    "565789": "That's reasonable but will significantly decrease the incentives for stress-testing of GAN competition design",
    "565891": "Close to this one, here and there noise is cornerstone\n\nHere you generate pictures from the noise\n\nIn Google analytics you could submit noise and end up in silver",
    "566007": "An alternative approach could be: to drop all the rules except for kernel limitations and to encourage solution sharing after the competition, for example, by giving the money prize for the most oustanding one (or several). Thus, Kaggle will collect all the possible loopholes and will come up with countermeasures for the next competition. Doing it on the fly seems to be painful and ineffective.",
    "566588": "Give competition medals and points only for publicly available kernels, it could solve the problem of cheating.",
    "566606": "Do you mean made public after the end of the competition? It would be good.",
    "566630": "Yup, all \"medals\" solutions must be published in order to receive points and medals. I think it is in consistent with the very idea of research competition (i.e. many new ideas could be found even in not top scores solutions). Kernel section would be flooded, but I don't think it is a big problem for such small competition.",
    "566632": "Agreed. We may repeat the success of \"hackathons\" competitions such as \"Data Science for Good\".",
    "566850": "What i am feeling that competition is competition. And rules will be also defined by the host of competition. Here the prize money and medal will motivate many to create something new in GAN or it might be some new method or new algorithm.",
    "568679": "Lol, we really want X5 to dominate this competition! :-p",
    "568733": "Appreciate your desire"
  },
  "source": "meta"
}