{
  "id": 16623,
  "title": "This competition is not popular at all",
  "url": "/competitions/dato-native/discussion/16623",
  "author_name": "",
  "post_date": "2015-09-22T21:52:03.417Z",
  "votes": 1,
  "comment_count": 17,
  "views": 3441,
  "content": "<p>Hmm. I wonder if it can hit 100 teams at the end.</p>",
  "messages": [
    {
      "id": "93214",
      "postDate": "09/22/2015 21:52:03",
      "content": "<p>Hmm. I wonder if it can hit 100 teams at the end.</p>",
      "rawMarkdown": "Hmm. I wonder if it can hit 100 teams at the end.",
      "votes": null
    },
    {
      "id": "93215",
      "postDate": "09/22/2015 21:54:49",
      "content": "<p>large data set with no scripts and a late competition reset to boot. setting 100 as a O/U looks pretty good to me. </p>",
      "rawMarkdown": "large data set with no scripts and a late competition reset to boot. setting 100 as a O/U looks pretty good to me.",
      "votes": null
    },
    {
      "id": "93220",
      "postDate": "09/22/2015 22:47:20",
      "content": "<p>Anyone have a simple script which can get to above 0.7 or 0.8 ?</p>\n\n<p>Post it in the forum and the competition will become popular...</p>",
      "rawMarkdown": "Anyone have a simple script which can get to above 0.7 or 0.8 ?\r\n\r\nPost it in the forum and the competition will become popular...",
      "votes": null
    },
    {
      "id": "93225",
      "postDate": "09/22/2015 23:04:33",
      "content": "<p>Hmm. I am not sure if that kind of popularity (with 'scripting' teams) will add any fun. That being said, just with count features in links and images will give you ~0.8 AUC.</p>",
      "rawMarkdown": "Hmm. I am not sure if that kind of popularity (with 'scripting' teams) will add any fun. That being said, just with count features in links and images will give you ~0.8 AUC.",
      "votes": null
    },
    {
      "id": "93226",
      "postDate": "09/22/2015 23:09:08",
      "content": "<p>[quote=SkyLibrary;93220]\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\nPost it in the forum and the competition will become popular...\n[/quote]\nThere is one! The Dato script already achieves around that.</p>",
      "rawMarkdown": "[quote=SkyLibrary;93220]\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\nPost it in the forum and the competition will become popular...\r\n[/quote]\r\nThere is one! The Dato script already achieves around that.",
      "votes": null
    },
    {
      "id": "93227",
      "postDate": "09/22/2015 23:24:28",
      "content": "<p>Forgot about that one, but I mean some script that can generate result quickly, maybe only look at title of page, type of page or what Little Boat suggested.</p>\n\n<p>[quote=NxGTR;93226]</p>\n\n<p>[quote=SkyLibrary;93220]\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\nPost it in the forum and the competition will become popular...\n[/quote]\nThere is one! The Dato script already achieves around that.</p>\n\n<p>[/quote]</p>",
      "rawMarkdown": "Forgot about that one, but I mean some script that can generate result quickly, maybe only look at title of page, type of page or what Little Boat suggested.\r\n\r\n[quote=NxGTR;93226]\r\n\r\n[quote=SkyLibrary;93220]\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\nPost it in the forum and the competition will become popular...\r\n[/quote]\r\nThere is one! The Dato script already achieves around that.\r\n\r\n\r\n[/quote]",
      "votes": null
    },
    {
      "id": "93240",
      "postDate": "09/23/2015 03:36:16",
      "content": "<p>[quote=SkyLibrary;93220]</p>\n\n<p>Anyone have a simple script which can get to above 0.7 or 0.8 ?</p>\n\n<p>Post it in the forum and the competition will become popular...</p>\n\n<p>[/quote]</p>\n\n<p>Ask and ye shall receive: <a href=\"https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model\">https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model</a></p>",
      "rawMarkdown": "[quote=SkyLibrary;93220]\r\n\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\n\r\nPost it in the forum and the competition will become popular...\r\n\r\n[/quote]\r\n\r\nAsk and ye shall receive: https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model",
      "votes": null
    },
    {
      "id": "93271",
      "postDate": "09/23/2015 10:54:45",
      "content": "<p>From a lot of the posts on here, I think a lot of people were focusing on the script provided (by the competition) and trying to build on that, vs doing anything on their own. Once the reboot, it seems the bulk of those who &quot;came back&quot; are those trying other things.</p>\n\n<p>As David noted in his post, maybe that will help bring people in, pointing them in new directions, etc.</p>",
      "rawMarkdown": "From a lot of the posts on here, I think a lot of people were focusing on the script provided (by the competition) and trying to build on that, vs doing anything on their own. Once the reboot, it seems the bulk of those who \"came back\" are those trying other things.\r\n\r\nAs David noted in his post, maybe that will help bring people in, pointing them in new directions, etc.",
      "votes": null
    },
    {
      "id": "93272",
      "postDate": "09/23/2015 11:25:09",
      "content": "<p>I would have liked to participate but there is so much to do and nearly no time left... But maybe I will give a try on this after the end.</p>",
      "rawMarkdown": "I would have liked to participate but there is so much to do and nearly no time left... But maybe I will give a try on this after the end.",
      "votes": null
    },
    {
      "id": "93282",
      "postDate": "09/23/2015 14:13:43",
      "content": "<p>With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a <a href=\"https://www.kaggle.com/wiki/UserRankingAndTierSystem\">Kaggle Master</a>).</p>\n\n<p>Some competitions (including <a href=\"https://www.kaggle.com/c/deloitte-western-australia-rental-prices\">Western Australia Rental Prices</a>) and an upcoming one I'm excited about require Kaggle Masters status to participate.</p>",
      "rawMarkdown": "With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a [Kaggle Master](https://www.kaggle.com/wiki/UserRankingAndTierSystem)).\r\n\r\nSome competitions (including [Western Australia Rental Prices](https://www.kaggle.com/c/deloitte-western-australia-rental-prices)) and an upcoming one I'm excited about require Kaggle Masters status to participate.",
      "votes": null
    },
    {
      "id": "93289",
      "postDate": "09/23/2015 15:55:49",
      "content": "<p>The point system is really a charm, if there are only 86 teams, and if you just make a all zero submission, which put you at rank 86, you can still earn 1650 points ...</p>",
      "rawMarkdown": "The point system is really a charm, if there are only 86 teams, and if you just make a all zero submission, which put you at rank 86, you can still earn 1650 points ...",
      "votes": null
    },
    {
      "id": "93466",
      "postDate": "09/26/2015 07:42:34",
      "content": "<p>Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, <em>easy</em> top 10 is probably a bit harder than it looks.</p>",
      "rawMarkdown": "Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, *easy* top 10 is probably a bit harder than it looks.",
      "votes": null
    },
    {
      "id": "93473",
      "postDate": "09/26/2015 12:02:23",
      "content": "<p>[quote=Mike Kim;93466]</p>\n\n<p>Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, <em>easy</em> top 10 is probably a bit harder than it looks.</p>\n\n<p>[/quote]</p>\n\n<p>However, ANY finish in this competition, even an all zeros predictions, will still grant a lot of Kaggle points.</p>",
      "rawMarkdown": "[quote=Mike Kim;93466]\r\n\r\nSure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, *easy* top 10 is probably a bit harder than it looks.\r\n\r\n[/quote]\r\n\r\nHowever, ANY finish in this competition, even an all zeros predictions, will still grant a lot of Kaggle points.",
      "votes": null
    },
    {
      "id": "93571",
      "postDate": "09/28/2015 09:16:41",
      "content": "<p>Make top 10 is still hard, lol.\nIt looks NxGTR can make it.</p>",
      "rawMarkdown": "Make top 10 is still hard, lol.\r\nIt looks NxGTR can make it.",
      "votes": null
    },
    {
      "id": "93657",
      "postDate": "09/29/2015 20:29:39",
      "content": "<p>Confirmed! This post is a native ads sponsored by stumbleupon and kaggle. </p>",
      "rawMarkdown": "Confirmed! This post is a native ads sponsored by stumbleupon and kaggle.",
      "votes": null
    },
    {
      "id": "93669",
      "postDate": "09/29/2015 23:14:43",
      "content": "<p>@rcarson, did you make this prediction by saving raw html of this thread and feeding it into your classifier? :)</p>",
      "rawMarkdown": "rcarson, did you make this prediction by saving raw html of this thread and feeding it into your classifier? :)",
      "votes": null
    },
    {
      "id": "95799",
      "postDate": "10/12/2015 04:05:06",
      "content": "<p>[quote=Ben Hamner;93282]</p>\n\n<p>With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a <a href=\"https://www.kaggle.com/wiki/UserRankingAndTierSystem\">Kaggle Master</a>).</p>\n\n<p>Some competitions (including <a href=\"https://www.kaggle.com/c/deloitte-western-australia-rental-prices\">Western Australia Rental Prices</a>) and an upcoming one I'm excited about require Kaggle Masters status to participate.</p>\n\n<p>[/quote]\nIt's not that easy at all ==</p>",
      "rawMarkdown": "[quote=Ben Hamner;93282]\r\n\r\nWith only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a [Kaggle Master](https://www.kaggle.com/wiki/UserRankingAndTierSystem)).\r\n\r\nSome competitions (including [Western Australia Rental Prices](https://www.kaggle.com/c/deloitte-western-australia-rental-prices)) and an upcoming one I'm excited about require Kaggle Masters status to participate.\r\n\r\n\r\n[/quote]\r\nIt's not that easy at all ==",
      "votes": null
    },
    {
      "id": "95802",
      "postDate": "10/12/2015 06:35:36",
      "content": "<p>I won't have time to work on it, but I thought this was a great competition.</p>\n\n<p>I agree. My impression was that it was pretty darn hard to move up, or stay at the same level, on a leaderboard in this competition. It could be because of some sort of self-selection of the most persistent contestants :) Let's see, few (only 2?) scripts, large raw data, competition reset - only 274 teams.</p>",
      "rawMarkdown": "I won't have time to work on it, but I thought this was a great competition.\r\n\r\nI agree. My impression was that it was pretty darn hard to move up, or stay at the same level, on a leaderboard in this competition. It could be because of some sort of self-selection of the most persistent contestants :) Let's see, few (only 2?) scripts, large raw data, competition reset - only 274 teams.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 93215,
      "author_name": "dmcgarry",
      "author_url": "",
      "post_date": "09/22/2015 21:54:49",
      "content": "<p>large data set with no scripts and a late competition reset to boot. setting 100 as a O/U looks pretty good to me. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93220,
      "author_name": "skylibrary",
      "author_url": "",
      "post_date": "09/22/2015 22:47:20",
      "content": "<p>Anyone have a simple script which can get to above 0.7 or 0.8 ?</p>\n\n<p>Post it in the forum and the competition will become popular...</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93225,
      "author_name": "xiaozhouwang",
      "author_url": "",
      "post_date": "09/22/2015 23:04:33",
      "content": "<p>Hmm. I am not sure if that kind of popularity (with 'scripting' teams) will add any fun. That being said, just with count features in links and images will give you ~0.8 AUC.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93226,
      "author_name": "carloshuertas",
      "author_url": "",
      "post_date": "09/22/2015 23:09:08",
      "content": "<p>[quote=SkyLibrary;93220]\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\nPost it in the forum and the competition will become popular...\n[/quote]\nThere is one! The Dato script already achieves around that.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93227,
      "author_name": "skylibrary",
      "author_url": "",
      "post_date": "09/22/2015 23:24:28",
      "content": "<p>Forgot about that one, but I mean some script that can generate result quickly, maybe only look at title of page, type of page or what Little Boat suggested.</p>\n\n<p>[quote=NxGTR;93226]</p>\n\n<p>[quote=SkyLibrary;93220]\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\nPost it in the forum and the competition will become popular...\n[/quote]\nThere is one! The Dato script already achieves around that.</p>\n\n<p>[/quote]</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93240,
      "author_name": "davidshinn",
      "author_url": "",
      "post_date": "09/23/2015 03:36:16",
      "content": "<p>[quote=SkyLibrary;93220]</p>\n\n<p>Anyone have a simple script which can get to above 0.7 or 0.8 ?</p>\n\n<p>Post it in the forum and the competition will become popular...</p>\n\n<p>[/quote]</p>\n\n<p>Ask and ye shall receive: <a href=\"https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model\">https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93271,
      "author_name": "omgponies",
      "author_url": "",
      "post_date": "09/23/2015 10:54:45",
      "content": "<p>From a lot of the posts on here, I think a lot of people were focusing on the script provided (by the competition) and trying to build on that, vs doing anything on their own. Once the reboot, it seems the bulk of those who &quot;came back&quot; are those trying other things.</p>\n\n<p>As David noted in his post, maybe that will help bring people in, pointing them in new directions, etc.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93272,
      "author_name": "wildwizard",
      "author_url": "",
      "post_date": "09/23/2015 11:25:09",
      "content": "<p>I would have liked to participate but there is so much to do and nearly no time left... But maybe I will give a try on this after the end.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93282,
      "author_name": "benhamner",
      "author_url": "",
      "post_date": "09/23/2015 14:13:43",
      "content": "<p>With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a <a href=\"https://www.kaggle.com/wiki/UserRankingAndTierSystem\">Kaggle Master</a>).</p>\n\n<p>Some competitions (including <a href=\"https://www.kaggle.com/c/deloitte-western-australia-rental-prices\">Western Australia Rental Prices</a>) and an upcoming one I'm excited about require Kaggle Masters status to participate.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93289,
      "author_name": "skylibrary",
      "author_url": "",
      "post_date": "09/23/2015 15:55:49",
      "content": "<p>The point system is really a charm, if there are only 86 teams, and if you just make a all zero submission, which put you at rank 86, you can still earn 1650 points ...</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93466,
      "author_name": "mikeskim",
      "author_url": "",
      "post_date": "09/26/2015 07:42:34",
      "content": "<p>Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, <em>easy</em> top 10 is probably a bit harder than it looks.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93473,
      "author_name": "zachmayer",
      "author_url": "",
      "post_date": "09/26/2015 12:02:23",
      "content": "<p>[quote=Mike Kim;93466]</p>\n\n<p>Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, <em>easy</em> top 10 is probably a bit harder than it looks.</p>\n\n<p>[/quote]</p>\n\n<p>However, ANY finish in this competition, even an all zeros predictions, will still grant a lot of Kaggle points.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93571,
      "author_name": "yejiming",
      "author_url": "",
      "post_date": "09/28/2015 09:16:41",
      "content": "<p>Make top 10 is still hard, lol.\nIt looks NxGTR can make it.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93657,
      "author_name": "jiweiliu",
      "author_url": "",
      "post_date": "09/29/2015 20:29:39",
      "content": "<p>Confirmed! This post is a native ads sponsored by stumbleupon and kaggle. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 93669,
      "author_name": "bekbolatov",
      "author_url": "",
      "post_date": "09/29/2015 23:14:43",
      "content": "<p>@rcarson, did you make this prediction by saving raw html of this thread and feeding it into your classifier? :)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 95799,
      "author_name": "jiweiliu",
      "author_url": "",
      "post_date": "10/12/2015 04:05:06",
      "content": "<p>[quote=Ben Hamner;93282]</p>\n\n<p>With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a <a href=\"https://www.kaggle.com/wiki/UserRankingAndTierSystem\">Kaggle Master</a>).</p>\n\n<p>Some competitions (including <a href=\"https://www.kaggle.com/c/deloitte-western-australia-rental-prices\">Western Australia Rental Prices</a>) and an upcoming one I'm excited about require Kaggle Masters status to participate.</p>\n\n<p>[/quote]\nIt's not that easy at all ==</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 95802,
      "author_name": "bekbolatov",
      "author_url": "",
      "post_date": "10/12/2015 06:35:36",
      "content": "<p>I won't have time to work on it, but I thought this was a great competition.</p>\n\n<p>I agree. My impression was that it was pretty darn hard to move up, or stay at the same level, on a leaderboard in this competition. It could be because of some sort of self-selection of the most persistent contestants :) Let's see, few (only 2?) scripts, large raw data, competition reset - only 274 teams.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "93214": "Hmm. I wonder if it can hit 100 teams at the end.",
    "93215": "large data set with no scripts and a late competition reset to boot. setting 100 as a O/U looks pretty good to me.",
    "93220": "Anyone have a simple script which can get to above 0.7 or 0.8 ?\r\n\r\nPost it in the forum and the competition will become popular...",
    "93225": "Hmm. I am not sure if that kind of popularity (with 'scripting' teams) will add any fun. That being said, just with count features in links and images will give you ~0.8 AUC.",
    "93226": "[quote=SkyLibrary;93220]\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\nPost it in the forum and the competition will become popular...\r\n[/quote]\r\nThere is one! The Dato script already achieves around that.",
    "93227": "Forgot about that one, but I mean some script that can generate result quickly, maybe only look at title of page, type of page or what Little Boat suggested.\r\n\r\n[quote=NxGTR;93226]\r\n\r\n[quote=SkyLibrary;93220]\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\nPost it in the forum and the competition will become popular...\r\n[/quote]\r\nThere is one! The Dato script already achieves around that.\r\n\r\n\r\n[/quote]",
    "93240": "[quote=SkyLibrary;93220]\r\n\r\nAnyone have a simple script which can get to above 0.7 or 0.8 ?\r\n\r\nPost it in the forum and the competition will become popular...\r\n\r\n[/quote]\r\n\r\nAsk and ye shall receive: https://www.kaggle.com/c/dato-native/forums/t/16626/beat-the-benchmark-0-90388-with-simple-model",
    "93271": "From a lot of the posts on here, I think a lot of people were focusing on the script provided (by the competition) and trying to build on that, vs doing anything on their own. Once the reboot, it seems the bulk of those who \"came back\" are those trying other things.\r\n\r\nAs David noted in his post, maybe that will help bring people in, pointing them in new directions, etc.",
    "93272": "I would have liked to participate but there is so much to do and nearly no time left... But maybe I will give a try on this after the end.",
    "93282": "With only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a [Kaggle Master](https://www.kaggle.com/wiki/UserRankingAndTierSystem)).\r\n\r\nSome competitions (including [Western Australia Rental Prices](https://www.kaggle.com/c/deloitte-western-australia-rental-prices)) and an upcoming one I'm excited about require Kaggle Masters status to participate.",
    "93289": "The point system is really a charm, if there are only 86 teams, and if you just make a all zero submission, which put you at rank 86, you can still earn 1650 points ...",
    "93466": "Sure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, *easy* top 10 is probably a bit harder than it looks.",
    "93473": "[quote=Mike Kim;93466]\r\n\r\nSure it looks easy to get top 10, but you have to realize the people who stayed in the competition after the reset are not a random sample from the original competitors. Most of the Kagglers in this current one (the first 50 or so) are the top of the LB of the prior one and/or are the most committed. I believe you had to actively re-engage to enter this reset competition. That means many of the less persistent Kagglers dropped out. And it's usually the case you have to grind to the end to have a shot at winning. So yeah, *easy* top 10 is probably a bit harder than it looks.\r\n\r\n[/quote]\r\n\r\nHowever, ANY finish in this competition, even an all zeros predictions, will still grant a lot of Kaggle points.",
    "93571": "Make top 10 is still hard, lol.\r\nIt looks NxGTR can make it.",
    "93657": "Confirmed! This post is a native ads sponsored by stumbleupon and kaggle.",
    "93669": "rcarson, did you make this prediction by saving raw html of this thread and feeding it into your classifier? :)",
    "95799": "[quote=Ben Hamner;93282]\r\n\r\nWith only 86 teams and 3 weeks left, this is a great competition to earn a top 10 finish (one of the steps to becoming a [Kaggle Master](https://www.kaggle.com/wiki/UserRankingAndTierSystem)).\r\n\r\nSome competitions (including [Western Australia Rental Prices](https://www.kaggle.com/c/deloitte-western-australia-rental-prices)) and an upcoming one I'm excited about require Kaggle Masters status to participate.\r\n\r\n\r\n[/quote]\r\nIt's not that easy at all ==",
    "95802": "I won't have time to work on it, but I thought this was a great competition.\r\n\r\nI agree. My impression was that it was pretty darn hard to move up, or stay at the same level, on a leaderboard in this competition. It could be because of some sort of self-selection of the most persistent contestants :) Let's see, few (only 2?) scripts, large raw data, competition reset - only 274 teams."
  },
  "source": "meta"
}