{
  "id": 297865,
  "title": "Notebook score ... my thoughts about voting and forking",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/297865",
  "author_name": "",
  "post_date": "2021-12-30T07:59:24.792379400Z",
  "votes": 18,
  "comment_count": 14,
  "views": 0,
  "content": "<p>This is a totally loose idea to the Kaggle staff. <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a> </p>\n<p>Have you considered the idea of ​​rewarding notebooks not only for likes but also forks?  Are votes only a good indicator of rewarding notebooks? There is a lot of talk in AI now about cognitive bias.</p>\n<p>Two examples:</p>\n<ul>\n<li>\"I like, know the person or the notebook is of a famous person (GM) so it must be good\" - number of forks ZERO, number of votes GOLD.</li>\n<li>\"The notebook is very useful …\" - the number of forks is 100, the number of votes is only …  20 (bronze or … 49 silver 😂 ).</li>\n</ul>\n<p>Which of these notebook is more useful? Which brings more in people opinion (votes/forks)?</p>\n<p>Maybe you should consider changing the ranking system? Fork and vote? So, for example, a notebook that has over 50 forks is also gold …</p>",
  "messages": [
    {
      "id": "1632903",
      "postDate": "12/30/2021 07:59:24",
      "content": "<p>This is a totally loose idea to the Kaggle staff. <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a> </p>\n<p>Have you considered the idea of ​​rewarding notebooks not only for likes but also forks?  Are votes only a good indicator of rewarding notebooks? There is a lot of talk in AI now about cognitive bias.</p>\n<p>Two examples:</p>\n<ul>\n<li>\"I like, know the person or the notebook is of a famous person (GM) so it must be good\" - number of forks ZERO, number of votes GOLD.</li>\n<li>\"The notebook is very useful …\" - the number of forks is 100, the number of votes is only …  20 (bronze or … 49 silver 😂 ).</li>\n</ul>\n<p>Which of these notebook is more useful? Which brings more in people opinion (votes/forks)?</p>\n<p>Maybe you should consider changing the ranking system? Fork and vote? So, for example, a notebook that has over 50 forks is also gold …</p>",
      "rawMarkdown": "This is a totally loose idea to the Kaggle staff. @addisonhoward \n\nHave you considered the idea of ​​rewarding notebooks not only for likes but also forks?  Are votes only a good indicator of rewarding notebooks? There is a lot of talk in AI now about cognitive bias.\n\nTwo examples:\n- \"I like, know the person or the notebook is of a famous person (GM) so it must be good\" - number of forks ZERO, number of votes GOLD.\n- \"The notebook is very useful ...\" - the number of forks is 100, the number of votes is only ...  20 (bronze or ... 49 silver 😂 ).\n\nWhich of these notebook is more useful? Which brings more in people opinion (votes/forks)?\n\nMaybe you should consider changing the ranking system? Fork and vote? So, for example, a notebook that has over 50 forks is also gold ...",
      "votes": null
    },
    {
      "id": "1632907",
      "postDate": "12/30/2021 08:01:25",
      "content": "<p>One potential issue. One can fork a notebook <code>inf</code> time, Hence it creates a problem </p>",
      "rawMarkdown": "One potential issue. One can fork a notebook `inf` time, Hence it creates a problem",
      "votes": null
    },
    {
      "id": "1632911",
      "postDate": "12/30/2021 08:04:33",
      "content": "<p>I am sure people from Kaggle knows how to solve this problem :) </p>",
      "rawMarkdown": "I am sure people from Kaggle knows how to solve this problem :)",
      "votes": null
    },
    {
      "id": "1632913",
      "postDate": "12/30/2021 08:05:43",
      "content": "<p>An you know … it will solve one more problem … everyone will make sure that others do not copy + paste.</p>",
      "rawMarkdown": "An you know ... it will solve one more problem ... everyone will make sure that others do not copy + paste.",
      "votes": null
    },
    {
      "id": "1632932",
      "postDate": "12/30/2021 08:25:54",
      "content": "<p>I think downvote should also be introduced for some plagiarism or low-quality notebooks.</p>",
      "rawMarkdown": "I think downvote should also be introduced for some plagiarism or low-quality notebooks.",
      "votes": null
    },
    {
      "id": "1632936",
      "postDate": "12/30/2021 08:34:47",
      "content": "<p>👍👍👍👍 I agree … My notebooks should be better as well so … this will motivate me to create better notebooks. </p>",
      "rawMarkdown": "👍👍👍👍 I agree ... My notebooks should be better as well so ... this will motivate me to create better notebooks.",
      "votes": null
    },
    {
      "id": "1632945",
      "postDate": "12/30/2021 08:49:21",
      "content": "<p>But …. sometimes it is better to publish something of average quality than publish nothing …. 😂😲😊</p>",
      "rawMarkdown": "But .... sometimes it is better to publish something of average quality than publish nothing .... 😂😲😊",
      "votes": null
    },
    {
      "id": "1633208",
      "postDate": "12/30/2021 14:27:05",
      "content": "<p>weighted metric= (vote_by_unique_id+ forked_by_unique_id)/2</p>",
      "rawMarkdown": "weighted metric= (vote_by_unique_id+ forked_by_unique_id)/2",
      "votes": null
    },
    {
      "id": "1633211",
      "postDate": "12/30/2021 14:29:36",
      "content": "<p>:) as I said … maybe this is not good idea … but … looks good for me. People do not fork useless notebooks :) I think :)</p>",
      "rawMarkdown": ":) as I said … maybe this is not good idea … but … looks good for me. People do not fork useless notebooks :) I think :)",
      "votes": null
    },
    {
      "id": "1633233",
      "postDate": "12/30/2021 15:01:57",
      "content": "<p>then how about penalty weight by number of forks from same ID? </p>",
      "rawMarkdown": "then how about penalty weight by number of forks from same ID?",
      "votes": null
    },
    {
      "id": "1633758",
      "postDate": "12/31/2021 03:12:12",
      "content": "<p>Before downloading, show a box to upvote before forking/downloading to support the author. I think most people don't vote because they forget/just want to try notebook out. This might help without being a big change. Just an idea. I hope Kaggle can figure out a solution.</p>",
      "rawMarkdown": "Before downloading, show a box to upvote before forking/downloading to support the author. I think most people don't vote because they forget/just want to try notebook out. This might help without being a big change. Just an idea. I hope Kaggle can figure out a solution.",
      "votes": null
    },
    {
      "id": "1634082",
      "postDate": "12/31/2021 10:49:06",
      "content": "<p>Thank you for your comment. I agree. For me votes are important, forks as well. Because this is feedback of my work and some kind of appreciation. I always have a choice …. write notebook and do not share or share …. I think … sharing is benefitial for everybody … for me - I can improve myself and for those who are just starting out..</p>",
      "rawMarkdown": "Thank you for your comment. I agree. For me votes are important, forks as well. Because this is feedback of my work and some kind of appreciation. I always have a choice .... write notebook and do not share or share .... I think ... sharing is benefitial for everybody ... for me - I can improve myself and for those who are just starting out..",
      "votes": null
    },
    {
      "id": "1634292",
      "postDate": "12/31/2021 14:40:13",
      "content": "<p>You could limit the amount of forks counted per user. So maybe only the first or even the first two forks from a single user actually count. The others would not count towards medals. </p>",
      "rawMarkdown": "You could limit the amount of forks counted per user. So maybe only the first or even the first two forks from a single user actually count. The others would not count towards medals.",
      "votes": null
    },
    {
      "id": "1634350",
      "postDate": "12/31/2021 16:07:01",
      "content": "<p>obviously the metric is not only based on upvotes which can be observed.</p>",
      "rawMarkdown": "obviously the metric is not only based on upvotes which can be observed.",
      "votes": null
    },
    {
      "id": "1635420",
      "postDate": "01/01/2022 18:25:06",
      "content": "<p>While there is/and (unfortunately) will be unbalanced fork/vote ratio in kaggle notebooks people share, the recently observed forking frenzy is an <strong>unusual</strong> one. There seem to be <strong>bots</strong> involved. It has been noted/discussed in the following forums. </p>\n<p><a href=\"https://www.kaggle.com/general/296418#1628897\" target=\"_blank\">https://www.kaggle.com/general/296418#1628897</a><br>\n<a href=\"https://www.kaggle.com/general/296872\" target=\"_blank\">https://www.kaggle.com/general/296872</a><br>\n<a href=\"https://www.kaggle.com/general/296240\" target=\"_blank\">https://www.kaggle.com/general/296240</a></p>\n<p>So do not feel too bad about it if you have a high fork/vote ratio. Notebooks without any valuable contents have been forked so many times.</p>",
      "rawMarkdown": "While there is/and (unfortunately) will be unbalanced fork/vote ratio in kaggle notebooks people share, the recently observed forking frenzy is an **unusual** one. There seem to be **bots** involved. It has been noted/discussed in the following forums. \n\nhttps://www.kaggle.com/general/296418#1628897\nhttps://www.kaggle.com/general/296872\nhttps://www.kaggle.com/general/296240\n\nSo do not feel too bad about it if you have a high fork/vote ratio. Notebooks without any valuable contents have been forked so many times.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1632907,
      "author_name": "awsaf49",
      "author_url": "",
      "post_date": "12/30/2021 08:01:25",
      "content": "<p>One potential issue. One can fork a notebook <code>inf</code> time, Hence it creates a problem </p>",
      "votes": null,
      "replies": [
        {
          "id": 1632911,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/30/2021 08:04:33",
          "content": "<p>I am sure people from Kaggle knows how to solve this problem :) </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1632913,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/30/2021 08:05:43",
          "content": "<p>An you know … it will solve one more problem … everyone will make sure that others do not copy + paste.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1634292,
          "author_name": "vexxingbanana",
          "author_url": "",
          "post_date": "12/31/2021 14:40:13",
          "content": "<p>You could limit the amount of forks counted per user. So maybe only the first or even the first two forks from a single user actually count. The others would not count towards medals. </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1632932,
      "author_name": "zzy990106",
      "author_url": "",
      "post_date": "12/30/2021 08:25:54",
      "content": "<p>I think downvote should also be introduced for some plagiarism or low-quality notebooks.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1632936,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/30/2021 08:34:47",
          "content": "<p>👍👍👍👍 I agree … My notebooks should be better as well so … this will motivate me to create better notebooks. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1632945,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/30/2021 08:49:21",
          "content": "<p>But …. sometimes it is better to publish something of average quality than publish nothing …. 😂😲😊</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1633208,
      "author_name": "dragonzhang",
      "author_url": "",
      "post_date": "12/30/2021 14:27:05",
      "content": "<p>weighted metric= (vote_by_unique_id+ forked_by_unique_id)/2</p>",
      "votes": null,
      "replies": [
        {
          "id": 1633211,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/30/2021 14:29:36",
          "content": "<p>:) as I said … maybe this is not good idea … but … looks good for me. People do not fork useless notebooks :) I think :)</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1633233,
          "author_name": "dragonzhang",
          "author_url": "",
          "post_date": "12/30/2021 15:01:57",
          "content": "<p>then how about penalty weight by number of forks from same ID? </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1634350,
          "author_name": "dragonzhang",
          "author_url": "",
          "post_date": "12/31/2021 16:07:01",
          "content": "<p>obviously the metric is not only based on upvotes which can be observed.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1633758,
      "author_name": "outwrest",
      "author_url": "",
      "post_date": "12/31/2021 03:12:12",
      "content": "<p>Before downloading, show a box to upvote before forking/downloading to support the author. I think most people don't vote because they forget/just want to try notebook out. This might help without being a big change. Just an idea. I hope Kaggle can figure out a solution.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1634082,
          "author_name": "remekkinas",
          "author_url": "",
          "post_date": "12/31/2021 10:49:06",
          "content": "<p>Thank you for your comment. I agree. For me votes are important, forks as well. Because this is feedback of my work and some kind of appreciation. I always have a choice …. write notebook and do not share or share …. I think … sharing is benefitial for everybody … for me - I can improve myself and for those who are just starting out..</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1635420,
      "author_name": "desalegngeb",
      "author_url": "",
      "post_date": "01/01/2022 18:25:06",
      "content": "<p>While there is/and (unfortunately) will be unbalanced fork/vote ratio in kaggle notebooks people share, the recently observed forking frenzy is an <strong>unusual</strong> one. There seem to be <strong>bots</strong> involved. It has been noted/discussed in the following forums. </p>\n<p><a href=\"https://www.kaggle.com/general/296418#1628897\" target=\"_blank\">https://www.kaggle.com/general/296418#1628897</a><br>\n<a href=\"https://www.kaggle.com/general/296872\" target=\"_blank\">https://www.kaggle.com/general/296872</a><br>\n<a href=\"https://www.kaggle.com/general/296240\" target=\"_blank\">https://www.kaggle.com/general/296240</a></p>\n<p>So do not feel too bad about it if you have a high fork/vote ratio. Notebooks without any valuable contents have been forked so many times.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1632903": "This is a totally loose idea to the Kaggle staff. @addisonhoward \n\nHave you considered the idea of ​​rewarding notebooks not only for likes but also forks?  Are votes only a good indicator of rewarding notebooks? There is a lot of talk in AI now about cognitive bias.\n\nTwo examples:\n- \"I like, know the person or the notebook is of a famous person (GM) so it must be good\" - number of forks ZERO, number of votes GOLD.\n- \"The notebook is very useful ...\" - the number of forks is 100, the number of votes is only ...  20 (bronze or ... 49 silver 😂 ).\n\nWhich of these notebook is more useful? Which brings more in people opinion (votes/forks)?\n\nMaybe you should consider changing the ranking system? Fork and vote? So, for example, a notebook that has over 50 forks is also gold ...",
    "1632907": "One potential issue. One can fork a notebook `inf` time, Hence it creates a problem",
    "1632911": "I am sure people from Kaggle knows how to solve this problem :)",
    "1632913": "An you know ... it will solve one more problem ... everyone will make sure that others do not copy + paste.",
    "1632932": "I think downvote should also be introduced for some plagiarism or low-quality notebooks.",
    "1632936": "👍👍👍👍 I agree ... My notebooks should be better as well so ... this will motivate me to create better notebooks.",
    "1632945": "But .... sometimes it is better to publish something of average quality than publish nothing .... 😂😲😊",
    "1633208": "weighted metric= (vote_by_unique_id+ forked_by_unique_id)/2",
    "1633211": ":) as I said … maybe this is not good idea … but … looks good for me. People do not fork useless notebooks :) I think :)",
    "1633233": "then how about penalty weight by number of forks from same ID?",
    "1633758": "Before downloading, show a box to upvote before forking/downloading to support the author. I think most people don't vote because they forget/just want to try notebook out. This might help without being a big change. Just an idea. I hope Kaggle can figure out a solution.",
    "1634082": "Thank you for your comment. I agree. For me votes are important, forks as well. Because this is feedback of my work and some kind of appreciation. I always have a choice .... write notebook and do not share or share .... I think ... sharing is benefitial for everybody ... for me - I can improve myself and for those who are just starting out..",
    "1634292": "You could limit the amount of forks counted per user. So maybe only the first or even the first two forks from a single user actually count. The others would not count towards medals.",
    "1634350": "obviously the metric is not only based on upvotes which can be observed.",
    "1635420": "While there is/and (unfortunately) will be unbalanced fork/vote ratio in kaggle notebooks people share, the recently observed forking frenzy is an **unusual** one. There seem to be **bots** involved. It has been noted/discussed in the following forums. \n\nhttps://www.kaggle.com/general/296418#1628897\nhttps://www.kaggle.com/general/296872\nhttps://www.kaggle.com/general/296240\n\nSo do not feel too bad about it if you have a high fork/vote ratio. Notebooks without any valuable contents have been forked so many times."
  },
  "source": "meta"
}