{
  "id": 2023,
  "title": "Is it always better to predict 10 member based on the evaluation method?",
  "url": "/competitions/FacebookRecruiting/discussion/2023",
  "author_name": "",
  "post_date": "2012-06-10T13:50:58.477Z",
  "votes": null,
  "comment_count": 1,
  "views": 2049,
  "content": "<p>Am I right? The denominator n is the true number of follows not the number of follows I predicted?</p>",
  "messages": [
    {
      "id": "11509",
      "postDate": "06/10/2012 13:50:58",
      "content": "<p>Am I right? The denominator n is the true number of follows not the number of follows I predicted?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "11510",
      "postDate": "06/10/2012 16:53:43",
      "content": "<p>The denominator is the lesser of the true number of follows and ten.<br>\r\nhttp://www.kaggle.com/c/FacebookRecruiting/forums/t/2002/alternate-explanation-of-mean-average-precision<br>\r\nThere is no disadvantage to predicting a full set of ten, unless it means you put bad predictions before good predictions in that list.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 11510,
      "author_name": "ajschumacher",
      "author_url": "",
      "post_date": "06/10/2012 16:53:43",
      "content": "<p>The denominator is the lesser of the true number of follows and ten.<br>\r\nhttp://www.kaggle.com/c/FacebookRecruiting/forums/t/2002/alternate-explanation-of-mean-average-precision<br>\r\nThere is no disadvantage to predicting a full set of ten, unless it means you put bad predictions before good predictions in that list.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "11509": "",
    "11510": ""
  },
  "source": "meta"
}