{
  "id": 670659,
  "title": "Will anyone release a higher-scoring strategy in the last 5 days of the competition?",
  "url": "/competitions/cafa-6-protein-function-prediction/discussion/670659",
  "author_name": "",
  "post_date": "2026-01-29T02:40:13.329131700Z",
  "votes": -7,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Will anyone release a higher-scoring strategy in the last 5 days of the competition?</p>",
  "messages": [
    {
      "id": "3398351",
      "postDate": "01/29/2026 02:40:13",
      "content": "<p>Will anyone release a higher-scoring strategy in the last 5 days of the competition?</p>",
      "rawMarkdown": "Will anyone release a higher-scoring strategy in the last 5 days of the competition?",
      "votes": null
    },
    {
      "id": "3398672",
      "postDate": "01/29/2026 14:15:31",
      "content": "<p>You hope that？</p>",
      "rawMarkdown": "You hope that？",
      "votes": null
    },
    {
      "id": "3398941",
      "postDate": "01/29/2026 23:59:20",
      "content": "<p>These last days are usually for finishing methods and preparing final submissions; maybe as a side effect the public LB goes up a bit, but I think we’re already at the limit.</p>\n<p>Regarding using the best Kaggle code/datasets, I’m not sure it’s a good idea to rely on that signal… it could very well be a clear case of overfitting. It also makes me a bit uneasy to work with things I can’t fully trace or understand, even though it’s true that incorporating it does improve my public rank.</p>",
      "rawMarkdown": "These last days are usually for finishing methods and preparing final submissions; maybe as a side effect the public LB goes up a bit, but I think we’re already at the limit.\n\nRegarding using the best Kaggle code/datasets, I’m not sure it’s a good idea to rely on that signal… it could very well be a clear case of overfitting. It also makes me a bit uneasy to work with things I can’t fully trace or understand, even though it’s true that incorporating it does improve my public rank.",
      "votes": null
    },
    {
      "id": "3398980",
      "postDate": "01/30/2026 02:01:29",
      "content": "<p>I just wanted to see what people would do.</p>",
      "rawMarkdown": "I just wanted to see what people would do.",
      "votes": null
    },
    {
      "id": "3398981",
      "postDate": "01/30/2026 02:01:51",
      "content": "<p>That makes sense.</p>",
      "rawMarkdown": "That makes sense.",
      "votes": null
    },
    {
      "id": "3400045",
      "postDate": "01/31/2026 19:43:40",
      "content": "<p>How is the private score calculated? Is it recall?</p>",
      "rawMarkdown": "How is the private score calculated? Is it recall?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3398672,
      "author_name": "zejunfool",
      "author_url": "",
      "post_date": "01/29/2026 14:15:31",
      "content": "<p>You hope that？</p>",
      "votes": null,
      "replies": [
        {
          "id": 3398980,
          "author_name": "jinchengzhang1",
          "author_url": "",
          "post_date": "01/30/2026 02:01:29",
          "content": "<p>I just wanted to see what people would do.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3398941,
      "author_name": "xaxipirulazo",
      "author_url": "",
      "post_date": "01/29/2026 23:59:20",
      "content": "<p>These last days are usually for finishing methods and preparing final submissions; maybe as a side effect the public LB goes up a bit, but I think we’re already at the limit.</p>\n<p>Regarding using the best Kaggle code/datasets, I’m not sure it’s a good idea to rely on that signal… it could very well be a clear case of overfitting. It also makes me a bit uneasy to work with things I can’t fully trace or understand, even though it’s true that incorporating it does improve my public rank.</p>",
      "votes": null,
      "replies": [
        {
          "id": 3398981,
          "author_name": "jinchengzhang1",
          "author_url": "",
          "post_date": "01/30/2026 02:01:51",
          "content": "<p>That makes sense.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3400045,
      "author_name": "yashrajprasad",
      "author_url": "",
      "post_date": "01/31/2026 19:43:40",
      "content": "<p>How is the private score calculated? Is it recall?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3398351": "Will anyone release a higher-scoring strategy in the last 5 days of the competition?",
    "3398672": "You hope that？",
    "3398941": "These last days are usually for finishing methods and preparing final submissions; maybe as a side effect the public LB goes up a bit, but I think we’re already at the limit.\n\nRegarding using the best Kaggle code/datasets, I’m not sure it’s a good idea to rely on that signal… it could very well be a clear case of overfitting. It also makes me a bit uneasy to work with things I can’t fully trace or understand, even though it’s true that incorporating it does improve my public rank.",
    "3398980": "I just wanted to see what people would do.",
    "3398981": "That makes sense.",
    "3400045": "How is the private score calculated? Is it recall?"
  },
  "source": "meta"
}