{
  "id": 285150,
  "title": "Why not hiding / disabling the public leaderbord ? ",
  "url": "/competitions/g-research-crypto-forecasting/discussion/285150",
  "author_name": "",
  "post_date": "2021-11-03T14:22:13.921829400Z",
  "votes": 7,
  "comment_count": 8,
  "views": 0,
  "content": "<p>Since it is not meaningfull, why show / maintain a public LB ?</p>",
  "messages": [
    {
      "id": "1569537",
      "postDate": "11/03/2021 14:22:13",
      "content": "<p>Since it is not meaningfull, why show / maintain a public LB ?</p>",
      "rawMarkdown": "Since it is not meaningfull, why show / maintain a public LB ?",
      "votes": null
    },
    {
      "id": "1569554",
      "postDate": "11/03/2021 14:36:14",
      "content": "<p>I don't think it's a bad idea to have it there so we can do quick tests.</p>",
      "rawMarkdown": "I don't think it's a bad idea to have it there so we can do quick tests.",
      "votes": null
    },
    {
      "id": "1569558",
      "postDate": "11/03/2021 14:39:16",
      "content": "<p>We can get our score from our submission tab. What is the point of maintening the LB tab ?</p>",
      "rawMarkdown": "We can get our score from our submission tab. What is the point of maintening the LB tab ?",
      "votes": null
    },
    {
      "id": "1569560",
      "postDate": "11/03/2021 14:41:39",
      "content": "<p><a href=\"https://www.kaggle.com/lucasmorin\" target=\"_blank\">@lucasmorin</a> True🤔</p>",
      "rawMarkdown": "lucasmorin True🤔",
      "votes": null
    },
    {
      "id": "1569684",
      "postDate": "11/03/2021 16:45:03",
      "content": "<p>Dear <a href=\"https://www.kaggle.com/lucasmorin\" target=\"_blank\">@lucasmorin</a> </p>\n<p>¿I guess in an effort to make it feel more like a standard kaggle competition? <br>\nI am not sure about this comp: if the data is public, and there is no real leaderboard, why not (either virtually, or put some <a href=\"https://www.investopedia.com/terms/s/skininthegame.asp\" target=\"_blank\">skin in the game</a> if you are confident) invest $1000 in one (or all of the) of the cryptocurrencies on February 1, 2022, perform trades based on your model predictions, and three months later on May 3, 2022 see how much money you have, and that is your final (ersatz) score! 😃 or ☹️<br>\n…Of course if one did that, use could then be made of external data, such as Elon Musk tweets <em>etc</em> during the final phase…<br>\nIn other words; I am not sure what this competition brings to the table that we couldn't do  for ourselves off-line (other than some prize money for the Top 10 notebooks, which if they work could be worth a small fortune)</p>\n<p>All the best,<br>\ncarl</p>",
      "rawMarkdown": "Dear @lucasmorin \n\n¿I guess in an effort to make it feel more like a standard kaggle competition? \nI am not sure about this comp: if the data is public, and there is no real leaderboard, why not (either virtually, or put some [skin in the game](https://www.investopedia.com/terms/s/skininthegame.asp) if you are confident) invest $1000 in one (or all of the) of the cryptocurrencies on February 1, 2022, perform trades based on your model predictions, and three months later on May 3, 2022 see how much money you have, and that is your final (ersatz) score! 😃 or ☹️\n...Of course if one did that, use could then be made of external data, such as Elon Musk tweets *etc* during the final phase...\nIn other words; I am not sure what this competition brings to the table that we couldn't do  for ourselves off-line (other than some prize money for the Top 10 notebooks, which if they work could be worth a small fortune)\n\nAll the best,\ncarl",
      "votes": null
    },
    {
      "id": "1569694",
      "postDate": "11/03/2021 16:55:48",
      "content": "<p>As long as people are not spamming with full scores, the public LB is a good thing to have. And public LB score is good to know anyway. You can treat it as your validation set!</p>",
      "rawMarkdown": "As long as people are not spamming with full scores, the public LB is a good thing to have. And public LB score is good to know anyway. You can treat it as your validation set!",
      "votes": null
    },
    {
      "id": "1571904",
      "postDate": "11/05/2021 09:12:22",
      "content": "<p><a href=\"https://www.kaggle.com/maggiemd\" target=\"_blank\">@maggiemd</a>, <a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a>, <a href=\"https://www.kaggle.com/andrewscherer\" target=\"_blank\">@andrewscherer</a> now that we have perfect scores and people use their LB position for (questionnable) advertising, would you consider this option ? </p>",
      "rawMarkdown": "maggiemd, @sohier, @andrewscherer now that we have perfect scores and people use their LB position for (questionnable) advertising, would you consider this option ?",
      "votes": null
    },
    {
      "id": "1571915",
      "postDate": "11/05/2021 09:20:14",
      "content": "<p>What now ?</p>",
      "rawMarkdown": "What now ?",
      "votes": null
    },
    {
      "id": "1573082",
      "postDate": "11/06/2021 08:52:20",
      "content": "<p>haha LB seems spammed :p when i wrote the comment there was none</p>",
      "rawMarkdown": "haha LB seems spammed :p when i wrote the comment there was none",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1569554,
      "author_name": "maherelouahabi",
      "author_url": "",
      "post_date": "11/03/2021 14:36:14",
      "content": "<p>I don't think it's a bad idea to have it there so we can do quick tests.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1569558,
          "author_name": "lucasmorin",
          "author_url": "",
          "post_date": "11/03/2021 14:39:16",
          "content": "<p>We can get our score from our submission tab. What is the point of maintening the LB tab ?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1569560,
          "author_name": "maherelouahabi",
          "author_url": "",
          "post_date": "11/03/2021 14:41:39",
          "content": "<p><a href=\"https://www.kaggle.com/lucasmorin\" target=\"_blank\">@lucasmorin</a> True🤔</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1569684,
      "author_name": "carlmcbrideellis",
      "author_url": "",
      "post_date": "11/03/2021 16:45:03",
      "content": "<p>Dear <a href=\"https://www.kaggle.com/lucasmorin\" target=\"_blank\">@lucasmorin</a> </p>\n<p>¿I guess in an effort to make it feel more like a standard kaggle competition? <br>\nI am not sure about this comp: if the data is public, and there is no real leaderboard, why not (either virtually, or put some <a href=\"https://www.investopedia.com/terms/s/skininthegame.asp\" target=\"_blank\">skin in the game</a> if you are confident) invest $1000 in one (or all of the) of the cryptocurrencies on February 1, 2022, perform trades based on your model predictions, and three months later on May 3, 2022 see how much money you have, and that is your final (ersatz) score! 😃 or ☹️<br>\n…Of course if one did that, use could then be made of external data, such as Elon Musk tweets <em>etc</em> during the final phase…<br>\nIn other words; I am not sure what this competition brings to the table that we couldn't do  for ourselves off-line (other than some prize money for the Top 10 notebooks, which if they work could be worth a small fortune)</p>\n<p>All the best,<br>\ncarl</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1569694,
      "author_name": "zarif98sjs",
      "author_url": "",
      "post_date": "11/03/2021 16:55:48",
      "content": "<p>As long as people are not spamming with full scores, the public LB is a good thing to have. And public LB score is good to know anyway. You can treat it as your validation set!</p>",
      "votes": null,
      "replies": [
        {
          "id": 1571915,
          "author_name": "lucasmorin",
          "author_url": "",
          "post_date": "11/05/2021 09:20:14",
          "content": "<p>What now ?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1573082,
          "author_name": "zarif98sjs",
          "author_url": "",
          "post_date": "11/06/2021 08:52:20",
          "content": "<p>haha LB seems spammed :p when i wrote the comment there was none</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1571904,
      "author_name": "lucasmorin",
      "author_url": "",
      "post_date": "11/05/2021 09:12:22",
      "content": "<p><a href=\"https://www.kaggle.com/maggiemd\" target=\"_blank\">@maggiemd</a>, <a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a>, <a href=\"https://www.kaggle.com/andrewscherer\" target=\"_blank\">@andrewscherer</a> now that we have perfect scores and people use their LB position for (questionnable) advertising, would you consider this option ? </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1569537": "Since it is not meaningfull, why show / maintain a public LB ?",
    "1569554": "I don't think it's a bad idea to have it there so we can do quick tests.",
    "1569558": "We can get our score from our submission tab. What is the point of maintening the LB tab ?",
    "1569560": "lucasmorin True🤔",
    "1569684": "Dear @lucasmorin \n\n¿I guess in an effort to make it feel more like a standard kaggle competition? \nI am not sure about this comp: if the data is public, and there is no real leaderboard, why not (either virtually, or put some [skin in the game](https://www.investopedia.com/terms/s/skininthegame.asp) if you are confident) invest $1000 in one (or all of the) of the cryptocurrencies on February 1, 2022, perform trades based on your model predictions, and three months later on May 3, 2022 see how much money you have, and that is your final (ersatz) score! 😃 or ☹️\n...Of course if one did that, use could then be made of external data, such as Elon Musk tweets *etc* during the final phase...\nIn other words; I am not sure what this competition brings to the table that we couldn't do  for ourselves off-line (other than some prize money for the Top 10 notebooks, which if they work could be worth a small fortune)\n\nAll the best,\ncarl",
    "1569694": "As long as people are not spamming with full scores, the public LB is a good thing to have. And public LB score is good to know anyway. You can treat it as your validation set!",
    "1571904": "maggiemd, @sohier, @andrewscherer now that we have perfect scores and people use their LB position for (questionnable) advertising, would you consider this option ?",
    "1571915": "What now ?",
    "1573082": "haha LB seems spammed :p when i wrote the comment there was none"
  },
  "source": "meta"
}