{
  "id": 319650,
  "title": "Miracle of Coinscedence ",
  "url": "/competitions/happy-whale-and-dolphin/discussion/319650",
  "author_name": "Gleb",
  "post_date": "2022-04-18T11:10:34.120000",
  "votes": 136,
  "comment_count": 77,
  "views": 0,
  "content": "<p>While I fail at deep learning, i still can parse html's!  Shall we?</p>\n<p>Lets look at top100 teams from HappyWhale competition. Out of those lets ignore teams with submission counts &gt; 90, why not. Short list of 44 teams with 159 users to have a closer look at.<br>\nLets count users in those teams: Counter({1: 11,  2: 3,  3: 1,  4: 6, 5: 23}). Hmm, doesnt look intuitive, 23 teams of 5 users at top100 with under 90 submissions. Quite a performance!</p>\n<p>To the main part: for some reason can we find out when users at top100 registered at kaggle? Yes we can!  Here is a short list of dates:</p>\n<pre><code>  2 user(s) registered at 2021-08-22\n  3 user(s) registered at 2022-03-23\n  3 user(s) registered at 2021-07-26\n  3 user(s) registered at 2022-04-05\n  3 user(s) registered at 2022-04-11\n  4 user(s) registered at 2021-09-16\n  4 user(s) registered at 2022-02-28\n  4 user(s) registered at 2022-03-01\n  6 user(s) registered at 2022-03-20\n  6 user(s) registered at 2022-04-03\n  7 user(s) registered at 2022-01-09\n  7 user(s) registered at 2022-01-11\n  8 user(s) registered at 2022-01-24\n 10 user(s) registered at 2021-09-18\n 13 user(s) registered at 2021-09-19\n</code></pre>\n<p>27 users registred on the same date of 16-19.09.21 !  another 14 at 09-11.01.22 !  And all of them in the same competition, in top100, with same amount of submissions made, within a team of 5?  Im lost! here are some of them, with a lot of dates matching. For sure some teams are less interesting, keep in mind, all of this is just stats. </p>\n<p>(Highlighted only dates through 16-19 09 2021)</p>\n<table>\n<thead>\n<tr>\n<th>team</th>\n<th>user0</th>\n<th>user1</th>\n<th>user2</th>\n<th>user3</th>\n<th>user4</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>10sarahcla</td>\n<td>2021-03-22</td>\n<td>2021-07-15</td>\n<td><strong>2021-09-16</strong></td>\n<td>2022-01-22</td>\n<td>2022-03-23</td>\n</tr>\n<tr>\n<td>13Lindapan</td>\n<td>2021-04-23</td>\n<td>2021-04-30</td>\n<td>2021-07-21</td>\n<td>2021-07-22</td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>15Wesley</td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-11</td>\n</tr>\n<tr>\n<td>24OtisBing</td>\n<td><strong>2021-09-16</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n<td>2022-01-24</td>\n<td></td>\n</tr>\n<tr>\n<td>25JSPOW</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-11</td>\n</tr>\n<tr>\n<td>31simba</td>\n<td>2021-06-28</td>\n<td>2021-10-31</td>\n<td>2021-11-29</td>\n<td>2022-02-26</td>\n<td>2022-03-20</td>\n</tr>\n<tr>\n<td>41Jesus</td>\n<td><strong>2021-09-16</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>42Kingfish</td>\n<td>2020-08-04</td>\n<td>2021-04-30</td>\n<td>2022-01-13</td>\n<td>2022-02-04</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>45BobHUN</td>\n<td>2022-01-24</td>\n<td>2022-02-22</td>\n<td>2022-02-28</td>\n<td>2022-02-28</td>\n<td></td>\n</tr>\n<tr>\n<td>47Darryl</td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-23</td>\n<td>2022-01-24</td>\n</tr>\n<tr>\n<td>52MilkMilk</td>\n<td><strong>2021-09-16</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n</tr>\n<tr>\n<td>59V V</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-02</td>\n</tr>\n<tr>\n<td>61PM</td>\n<td>2022-03-20</td>\n<td>2022-03-23</td>\n<td>2022-03-23</td>\n<td>2022-04-03</td>\n<td></td>\n</tr>\n<tr>\n<td>64SpencerW</td>\n<td>2022-03-20</td>\n<td>2022-03-20</td>\n<td>2022-03-20</td>\n<td>2022-04-03</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>69RapJOHN</td>\n<td>2021-07-26</td>\n<td>2021-07-26</td>\n<td>2021-07-26</td>\n<td>2022-01-24</td>\n<td></td>\n</tr>\n<tr>\n<td>72Again</td>\n<td>2022-01-09</td>\n<td>2022-01-11</td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n<td>2022-03-20</td>\n</tr>\n<tr>\n<td>73gilber</td>\n<td>2021-08-22</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>75FredPlan</td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-23</td>\n<td>2022-02-22</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>76Frederic</td>\n<td><strong>2021-09-18</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-24</td>\n<td>2022-02-28</td>\n</tr>\n<tr>\n<td>77World</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-02-28</td>\n<td>2022-03-02</td>\n</tr>\n<tr>\n<td>86Hykdt</td>\n<td>2020-02-19</td>\n<td>2021-10-21</td>\n<td>2022-01-16</td>\n<td>2022-04-03</td>\n<td>2022-04-05</td>\n</tr>\n<tr>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n</tr>\n</tbody>\n</table>",
  "messages": [
    {
      "id": 1759132,
      "postDate": "2022-04-18T11:10:34.120Z",
      "content": "<p>While I fail at deep learning, i still can parse html's!  Shall we?</p>\n<p>Lets look at top100 teams from HappyWhale competition. Out of those lets ignore teams with submission counts &gt; 90, why not. Short list of 44 teams with 159 users to have a closer look at.<br>\nLets count users in those teams: Counter({1: 11,  2: 3,  3: 1,  4: 6, 5: 23}). Hmm, doesnt look intuitive, 23 teams of 5 users at top100 with under 90 submissions. Quite a performance!</p>\n<p>To the main part: for some reason can we find out when users at top100 registered at kaggle? Yes we can!  Here is a short list of dates:</p>\n<pre><code>  2 user(s) registered at 2021-08-22\n  3 user(s) registered at 2022-03-23\n  3 user(s) registered at 2021-07-26\n  3 user(s) registered at 2022-04-05\n  3 user(s) registered at 2022-04-11\n  4 user(s) registered at 2021-09-16\n  4 user(s) registered at 2022-02-28\n  4 user(s) registered at 2022-03-01\n  6 user(s) registered at 2022-03-20\n  6 user(s) registered at 2022-04-03\n  7 user(s) registered at 2022-01-09\n  7 user(s) registered at 2022-01-11\n  8 user(s) registered at 2022-01-24\n 10 user(s) registered at 2021-09-18\n 13 user(s) registered at 2021-09-19\n</code></pre>\n<p>27 users registred on the same date of 16-19.09.21 !  another 14 at 09-11.01.22 !  And all of them in the same competition, in top100, with same amount of submissions made, within a team of 5?  Im lost! here are some of them, with a lot of dates matching. For sure some teams are less interesting, keep in mind, all of this is just stats. </p>\n<p>(Highlighted only dates through 16-19 09 2021)</p>\n<table>\n<thead>\n<tr>\n<th>team</th>\n<th>user0</th>\n<th>user1</th>\n<th>user2</th>\n<th>user3</th>\n<th>user4</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>10sarahcla</td>\n<td>2021-03-22</td>\n<td>2021-07-15</td>\n<td><strong>2021-09-16</strong></td>\n<td>2022-01-22</td>\n<td>2022-03-23</td>\n</tr>\n<tr>\n<td>13Lindapan</td>\n<td>2021-04-23</td>\n<td>2021-04-30</td>\n<td>2021-07-21</td>\n<td>2021-07-22</td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>15Wesley</td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-11</td>\n</tr>\n<tr>\n<td>24OtisBing</td>\n<td><strong>2021-09-16</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n<td>2022-01-24</td>\n<td></td>\n</tr>\n<tr>\n<td>25JSPOW</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-11</td>\n</tr>\n<tr>\n<td>31simba</td>\n<td>2021-06-28</td>\n<td>2021-10-31</td>\n<td>2021-11-29</td>\n<td>2022-02-26</td>\n<td>2022-03-20</td>\n</tr>\n<tr>\n<td>41Jesus</td>\n<td><strong>2021-09-16</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>42Kingfish</td>\n<td>2020-08-04</td>\n<td>2021-04-30</td>\n<td>2022-01-13</td>\n<td>2022-02-04</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>45BobHUN</td>\n<td>2022-01-24</td>\n<td>2022-02-22</td>\n<td>2022-02-28</td>\n<td>2022-02-28</td>\n<td></td>\n</tr>\n<tr>\n<td>47Darryl</td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-23</td>\n<td>2022-01-24</td>\n</tr>\n<tr>\n<td>52MilkMilk</td>\n<td><strong>2021-09-16</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n</tr>\n<tr>\n<td>59V V</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-01</td>\n<td>2022-03-02</td>\n</tr>\n<tr>\n<td>61PM</td>\n<td>2022-03-20</td>\n<td>2022-03-23</td>\n<td>2022-03-23</td>\n<td>2022-04-03</td>\n<td></td>\n</tr>\n<tr>\n<td>64SpencerW</td>\n<td>2022-03-20</td>\n<td>2022-03-20</td>\n<td>2022-03-20</td>\n<td>2022-04-03</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>69RapJOHN</td>\n<td>2021-07-26</td>\n<td>2021-07-26</td>\n<td>2021-07-26</td>\n<td>2022-01-24</td>\n<td></td>\n</tr>\n<tr>\n<td>72Again</td>\n<td>2022-01-09</td>\n<td>2022-01-11</td>\n<td>2022-01-11</td>\n<td>2022-01-24</td>\n<td>2022-03-20</td>\n</tr>\n<tr>\n<td>73gilber</td>\n<td>2021-08-22</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n</tr>\n<tr>\n<td>75FredPlan</td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-01-11</td>\n<td>2022-01-23</td>\n<td>2022-02-22</td>\n<td>2022-04-03</td>\n</tr>\n<tr>\n<td>76Frederic</td>\n<td><strong>2021-09-18</strong></td>\n<td>2022-01-09</td>\n<td>2022-01-09</td>\n<td>2022-01-24</td>\n<td>2022-02-28</td>\n</tr>\n<tr>\n<td>77World</td>\n<td><strong>2021-09-18</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td><strong>2021-09-19</strong></td>\n<td>2022-02-28</td>\n<td>2022-03-02</td>\n</tr>\n<tr>\n<td>86Hykdt</td>\n<td>2020-02-19</td>\n<td>2021-10-21</td>\n<td>2022-01-16</td>\n<td>2022-04-03</td>\n<td>2022-04-05</td>\n</tr>\n<tr>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n<td></td>\n</tr>\n</tbody>\n</table>",
      "rawMarkdown": "While I fail at deep learning, i still can parse html's!  Shall we?\n\nLets look at top100 teams from HappyWhale competition. Out of those lets ignore teams with submission counts > 90, why not. Short list of 44 teams with 159 users to have a closer look at.\nLets count users in those teams: Counter({1: 11,  2: 3,  3: 1,  4: 6, 5: 23}). Hmm, doesnt look intuitive, 23 teams of 5 users at top100 with under 90 submissions. Quite a performance!\n\nTo the main part: for some reason can we find out when users at top100 registered at kaggle? Yes we can!  Here is a short list of dates:\n```\n  2 user(s) registered at 2021-08-22\n  3 user(s) registered at 2022-03-23\n  3 user(s) registered at 2021-07-26\n  3 user(s) registered at 2022-04-05\n  3 user(s) registered at 2022-04-11\n  4 user(s) registered at 2021-09-16\n  4 user(s) registered at 2022-02-28\n  4 user(s) registered at 2022-03-01\n  6 user(s) registered at 2022-03-20\n  6 user(s) registered at 2022-04-03\n  7 user(s) registered at 2022-01-09\n  7 user(s) registered at 2022-01-11\n  8 user(s) registered at 2022-01-24\n 10 user(s) registered at 2021-09-18\n 13 user(s) registered at 2021-09-19\n```\n27 users registred on the same date of 16-19.09.21 !  another 14 at 09-11.01.22 !  And all of them in the same competition, in top100, with same amount of submissions made, within a team of 5?  Im lost! here are some of them, with a lot of dates matching. For sure some teams are less interesting, keep in mind, all of this is just stats. \n\n(Highlighted only dates through 16-19 09 2021)\n\n| team | user0 | user1 | user2  | user3 | user4 |\n| --- | --- | --- | --- | --- | --- |\n|10sarahcla|2021-03-22 | 2021-07-15 | **2021-09-16** | 2022-01-22 | 2022-03-23|\n|13Lindapan|2021-04-23 | 2021-04-30 | 2021-07-21 | 2021-07-22 | **2021-09-19**|\n|15Wesley  |**2021-09-19** | 2022-01-09 | 2022-01-09 | 2022-01-09 | 2022-01-11|\n|24OtisBing|**2021-09-16** | 2022-01-11 | 2022-01-24 | 2022-01-24|\n|25JSPOW   |**2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19** | 2022-01-11|\n|31simba   |2021-06-28 | 2021-10-31 | 2021-11-29 | 2022-02-26 | 2022-03-20|\n|41Jesus   |**2021-09-16** | **2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19**|\n|42Kingfish|2020-08-04 | 2021-04-30 | 2022-01-13 | 2022-02-04 | 2022-04-03|\n|45BobHUN  |2022-01-24 | 2022-02-22 | 2022-02-28 | 2022-02-28|\n|47Darryl  |**2021-09-19** | **2021-09-19** | 2022-01-09 | 2022-01-23 | 2022-01-24|\n|52MilkMilk|**2021-09-16** | **2021-09-18** | **2021-09-18** | 2022-01-11 | 2022-01-24|\n|59V V     |2022-03-01 | 2022-03-01 | 2022-03-01 | 2022-03-01 | 2022-03-02|\n|61PM      |2022-03-20 | 2022-03-23 | 2022-03-23 | 2022-04-03|\n|64SpencerW|2022-03-20 | 2022-03-20 | 2022-03-20 | 2022-04-03 | 2022-04-03|\n|69RapJOHN |2021-07-26 | 2021-07-26 | 2021-07-26 | 2022-01-24|\n|72Again   |2022-01-09 | 2022-01-11 | 2022-01-11 | 2022-01-24 | 2022-03-20|\n|73gilber  |2021-08-22 | **2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19**|\n|75FredPlan|**2021-09-19** | 2022-01-11 | 2022-01-23 | 2022-02-22 | 2022-04-03|\n|76Frederic|**2021-09-18** | 2022-01-09 | 2022-01-09 | 2022-01-24 | 2022-02-28|\n|77World   |**2021-09-18** | **2021-09-19** | **2021-09-19** | 2022-02-28 | 2022-03-02|\n|86Hykdt   |2020-02-19 | 2021-10-21 | 2022-01-16 | 2022-04-03 | 2022-04-05|\n|  |  |  |  |  |  | \n",
      "votes": 135
    },
    {
      "id": 1759238,
      "postDate": "2022-04-18T13:17:36.997Z",
      "content": "<p>One of the many reasons why I believe the time of csv competitions should be finally over.</p>",
      "rawMarkdown": "One of the many reasons why I believe the time of csv competitions should be finally over.",
      "votes": 32,
      "replies": [
        {
          "id": 1759255,
          "postDate": "2022-04-18T13:41:05.760Z",
          "content": "<p>Sad but true, csv-like competitions way too easy to break :(</p>",
          "rawMarkdown": "Sad but true, csv-like competitions way too easy to break :(\n",
          "votes": 3
        },
        {
          "id": 1759283,
          "postDate": "2022-04-18T13:56:58.470Z",
          "content": "<p>There is only hope that such cheaters quite easy to identify and drop from final LB.<br>\nOtherwise the value of kaggle medals and ranks is greatly depreciated.</p>",
          "rawMarkdown": "There is only hope that such cheaters quite easy to identify and drop from final LB.\nOtherwise the value of kaggle medals and ranks is greatly depreciated.",
          "votes": 9
        },
        {
          "id": 1759383,
          "postDate": "2022-04-18T15:27:02.740Z",
          "content": "<p>CSV competitions just make it easier for them to cheat and hide it. I saw some really suspicious accounts in one team, all made 7 months ago played in similar competitions with other accounts made 7 months ago and all got similar rankings (bronze~silver). It is clear that this is an ongoing problem :( and it is very prominent in CSV competitions. </p>",
          "rawMarkdown": "CSV competitions just make it easier for them to cheat and hide it. I saw some really suspicious accounts in one team, all made 7 months ago played in similar competitions with other accounts made 7 months ago and all got similar rankings (bronze~silver). It is clear that this is an ongoing problem :( and it is very prominent in CSV competitions. ",
          "votes": 7
        },
        {
          "id": 1759740,
          "postDate": "2022-04-18T21:08:07.713Z",
          "content": "<p><a href=\"https://www.kaggle.com/psi\" target=\"_blank\">@psi</a> not sure this is the explanation. You just reported a similar issue in feedback competition which was a code competition.</p>\n<p>This said I am now in favor of code comps for other reasons:</p>\n<ol>\n<li>a bit more fair to those with less compute power than others.</li>\n<li>harder to exploit test data distribution to game the metric.</li>\n</ol>",
          "rawMarkdown": "@psi not sure this is the explanation. You just reported a similar issue in feedback competition which was a code competition.\n\nThis said I am now in favor of code comps for other reasons:\n1. a bit more fair to those with less compute power than others.\n2. harder to exploit test data distribution to game the metric.",
          "votes": 6
        },
        {
          "id": 1759794,
          "postDate": "2022-04-18T22:39:52.853Z",
          "content": "<p>Yes you are right, I only noticed the Feedback stuff afterwards. And again, just observance, no claim.<br>\nIt is still easier in csv competitions.</p>\n<p>Totally agree with your other points. </p>",
          "rawMarkdown": "Yes you are right, I only noticed the Feedback stuff afterwards. And again, just observance, no claim.\nIt is still easier in csv competitions.\n\nTotally agree with your other points. ",
          "votes": 3
        },
        {
          "id": 1759817,
          "postDate": "2022-04-18T23:48:01.467Z",
          "content": "<p>impressive last minute progression!</p>",
          "rawMarkdown": "impressive last minute progression!",
          "votes": 1
        },
        {
          "id": 1760906,
          "postDate": "2022-04-19T15:30:19.927Z",
          "content": "<p>Hi Psi, new to Kaggle!</p>\n<p>I understand the lack of transparency when it comes to participation in competitions can lead to oddities and fly in the face of why we're all here, to grow and learn. That said, could you please expand a little bit more on why csv competitions should be discoured beyond the given reason for this topic?</p>\n<p>Thanks!</p>",
          "rawMarkdown": "Hi Psi, new to Kaggle!\n\nI understand the lack of transparency when it comes to participation in competitions can lead to oddities and fly in the face of why we're all here, to grow and learn. That said, could you please expand a little bit more on why csv competitions should be discoured beyond the given reason for this topic?\n\nThanks!"
        },
        {
          "id": 1762358,
          "postDate": "2022-04-20T16:17:32.927Z",
          "content": "<p>Hi <a href=\"https://www.kaggle.com/jordancarlen\" target=\"_blank\">@jordancarlen</a>  - I think <a href=\"https://www.kaggle.com/cpmpml\" target=\"_blank\">@cpmpml</a> already nicely elaborated on a few of them. </p>\n<p>Some other benefits of code competitions are:</p>\n<ul>\n<li>It restricts mindless blending, which has no practicality for any production scenario and can also be quite useless to the host.  </li>\n<li>You are learning some engineering aspects that can be very valuable in your job as a data scientist</li>\n<li>It makes replicability of your solution easier</li>\n</ul>",
          "rawMarkdown": "Hi @jordancarlen  - I think @cpmpml already nicely elaborated on a few of them. \n\nSome other benefits of code competitions are:\n\n- It restricts mindless blending, which has no practicality for any production scenario and can also be quite useless to the host.  \n- You are learning some engineering aspects that can be very valuable in your job as a data scientist\n- It makes replicability of your solution easier",
          "votes": 3
        }
      ]
    },
    {
      "id": 1759154,
      "postDate": "2022-04-18T11:36:40.407Z",
      "content": "<p>They seem to mostly hide their countries, which is unusual on kaggle.</p>\n<p>Also, I found some paid tutoring services (experienced players with newbies who pay for medals), which may be the source of cheating: <br>\n<a href=\"https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156\" target=\"_blank\">https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156</a><br>\n<img src=\"https://i.ibb.co/wLC0kJj/043cb377ec04c713bea60f0c94084da.png\" alt=\"1\"><br>\n<img src=\"https://i.ibb.co/tLrQFJW/ed328472e89ee56c87f68e4730b2db4.png\" alt=\"2\"></p>",
      "rawMarkdown": "They seem to mostly hide their countries, which is unusual on kaggle.\n\nAlso, I found some paid tutoring services (experienced players with newbies who pay for medals), which may be the source of cheating: \nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156\n![1](https://i.ibb.co/wLC0kJj/043cb377ec04c713bea60f0c94084da.png)\n![2](https://i.ibb.co/tLrQFJW/ed328472e89ee56c87f68e4730b2db4.png)",
      "votes": 25,
      "replies": [
        {
          "id": 1759165,
          "postDate": "2022-04-18T11:47:48.533Z",
          "content": "<p><a href=\"https://www.linstitute.net/archives/513144\" target=\"_blank\">https://www.linstitute.net/archives/513144</a><br>\n<a href=\"https://www.xiaohongshu.com/discovery/item/624ab6550000000001024cca\" target=\"_blank\">https://www.xiaohongshu.com/discovery/item/624ab6550000000001024cca</a></p>\n<p>I also found some links.</p>",
          "rawMarkdown": "https://www.linstitute.net/archives/513144\nhttps://www.xiaohongshu.com/discovery/item/624ab6550000000001024cca\n\nI also found some links.\n",
          "votes": 8
        },
        {
          "id": 1759207,
          "postDate": "2022-04-18T12:38:24.080Z",
          "content": "<p>Not bad :) My notebook mentioned and shared :) <br>\nI have to join this school … to improve my skills :) </p>\n<blockquote>\n  <p>对CV感兴趣的同学不要错过了</p>\n</blockquote>\n<p>I will join next time :)</p>",
          "rawMarkdown": "Not bad :) My notebook mentioned and shared :) \nI have to join this school ... to improve my skills :) \n\n\n> 对CV感兴趣的同学不要错过了\n\nI will join next time :)",
          "votes": 16
        },
        {
          "id": 1759272,
          "postDate": "2022-04-18T13:49:27.977Z",
          "content": "<p>That cracks me up. 😂</p>",
          "rawMarkdown": "That cracks me up. 😂",
          "votes": 4
        },
        {
          "id": 1760191,
          "postDate": "2022-04-19T05:57:42.503Z",
          "content": "<p>…….😂😂</p>",
          "rawMarkdown": ".......😂😂"
        }
      ]
    },
    {
      "id": 1759781,
      "postDate": "2022-04-18T22:25:27.153Z",
      "content": "<p>Kaggle should organize a fraud detection competition for kaggle!<br>\nThis competition will provide a lot of data :)</p>",
      "rawMarkdown": "Kaggle should organize a fraud detection competition for kaggle!\nThis competition will provide a lot of data :)",
      "votes": 19,
      "replies": [
        {
          "id": 1759802,
          "postDate": "2022-04-18T22:56:16.727Z",
          "content": "<p>and maybe we'll see a record high level of \"geese teams\" there in top)</p>",
          "rawMarkdown": "and maybe we'll see a record high level of \"geese teams\" there in top)",
          "votes": 3
        }
      ]
    },
    {
      "id": 1759843,
      "postDate": "2022-04-19T00:46:17.723Z",
      "content": "<p>🙏🙏🙏🙏🙏🙏😭</p>",
      "rawMarkdown": "🙏🙏🙏🙏🙏🙏😭",
      "votes": 17,
      "replies": [
        {
          "id": 1762182,
          "postDate": "2022-04-20T13:42:06.163Z",
          "content": "<p>congratulation!!</p>",
          "rawMarkdown": "congratulation!!"
        }
      ]
    },
    {
      "id": 1759660,
      "postDate": "2022-04-18T19:35:16.660Z",
      "content": "<p>Just add some \"Miracle\" here.  <br>\nLook at <strong>Lindapanpan</strong> team and when they submitted.  Until April 9, all team members were inactive at least a month. Then, on April 9, they all decided to make sabmitions, April 10 - silence, April 11 again a lot of sabmitions from all team members. And then team up….  </p>\n<p>May be it is \"Miracle of Coinscedence\"<br>\nBut look at <strong>Wesley</strong> team - almost the same situation, even the same dates …</p>",
      "rawMarkdown": "Just add some \"Miracle\" here.  \nLook at **Lindapanpan** team and when they submitted.  Until April 9, all team members were inactive at least a month. Then, on April 9, they all decided to make sabmitions, April 10 - silence, April 11 again a lot of sabmitions from all team members. And then team up....  \n\nMay be it is \"Miracle of Coinscedence\"\nBut look at **Wesley** team - almost the same situation, even the same dates ...",
      "votes": 12,
      "replies": [
        {
          "id": 1759708,
          "postDate": "2022-04-18T20:23:52.823Z",
          "content": "<p>Some of them also participated in Feedback competition, and placed right next to each other at around rank ~100. Now they are in the same team :)</p>",
          "rawMarkdown": "Some of them also participated in Feedback competition, and placed right next to each other at around rank ~100. Now they are in the same team :)",
          "votes": 14
        }
      ]
    },
    {
      "id": 1761005,
      "postDate": "2022-04-19T16:23:08.677Z",
      "content": "<blockquote>\n  <p>While I fail at deep learning, </p>\n</blockquote>\n<p>well, if being 19th in a highly competitive challenge is failing then lots of people are failing with you!</p>",
      "rawMarkdown": "> While I fail at deep learning, \n\nwell, if being 19th in a highly competitive challenge is failing then lots of people are failing with you!",
      "votes": 9,
      "replies": [
        {
          "id": 1762429,
          "postDate": "2022-04-20T17:16:06.507Z",
          "content": "<p>thank you cpmp, from all failing people. we will get better, learning from the leaders</p>",
          "rawMarkdown": "thank you cpmp, from all failing people. we will get better, learning from the leaders",
          "votes": 2
        }
      ]
    },
    {
      "id": 1759163,
      "postDate": "2022-04-18T11:44:21.823Z",
      "content": "<p>Real shame, there's obviously some private sharing going on at the very least. Hopefully Kaggle can whittle it down, but there will always be some remaining who broke the rules undetected. Regardless, it's been a fun competition: learnt a lot of cool new stuff and can't wait to see the winning write ups.</p>",
      "rawMarkdown": "Real shame, there's obviously some private sharing going on at the very least. Hopefully Kaggle can whittle it down, but there will always be some remaining who broke the rules undetected. Regardless, it's been a fun competition: learnt a lot of cool new stuff and can't wait to see the winning write ups.",
      "votes": 10
    },
    {
      "id": 1760006,
      "postDate": "2022-04-19T02:59:41.210Z",
      "content": "<p>I guess those students cannot get their money back if they were removed from the LB.</p>\n<p><code>0 risk</code> for those institution 😂</p>",
      "rawMarkdown": "I guess those students cannot get their money back if they were removed from the LB.\n\n`0 risk` for those institution 😂",
      "votes": 9
    },
    {
      "id": 1760120,
      "postDate": "2022-04-19T04:37:53.300Z",
      "content": "<p>I agree with other comments about making the analysis of cheating a part of the competition, with the end result being a more authentic competition. It's a shame what folks will do with/for money.</p>",
      "rawMarkdown": "I agree with other comments about making the analysis of cheating a part of the competition, with the end result being a more authentic competition. It's a shame what folks will do with/for money.",
      "votes": 7
    },
    {
      "id": 1759146,
      "postDate": "2022-04-18T11:26:29.103Z",
      "content": "<p><a href=\"https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635\" target=\"_blank\">https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635</a><br>\nMelanoma competition V2.0?</p>",
      "rawMarkdown": "https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635\nMelanoma competition V2.0?",
      "votes": 7,
      "replies": [
        {
          "id": 1759150,
          "postDate": "2022-04-18T11:29:37.043Z",
          "content": "<p>Seems like group behavior.<br>\nWhat is the result of Melanoma competition， will these suspicious accounts be removed from this competition?</p>",
          "rawMarkdown": "Seems like group behavior.\nWhat is the result of Melanoma competition， will these suspicious accounts be removed from this competition?",
          "votes": 3
        },
        {
          "id": 1759155,
          "postDate": "2022-04-18T11:37:42.130Z",
          "content": "<p>If I could recall correctly they got removed when the private leaderboard was finalized, although the shakeup was huge so most of them placed worse on the leaderboard. </p>",
          "rawMarkdown": "If I could recall correctly they got removed when the private leaderboard was finalized, although the shakeup was huge so most of them placed worse on the leaderboard. ",
          "votes": 1
        },
        {
          "id": 1759160,
          "postDate": "2022-04-18T11:40:45.927Z",
          "content": "<p><br>\n <a href=\"https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/175156#973662\" target=\"_blank\">https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/175156#973662</a></p>\n<p>Edit: My bad, the kaggle master who got called out got his account deleted but not removed from the leaderboard?? I guess the comment on the discussion was made before the account was deleted. </p>",
          "rawMarkdown": "~~The one that got called out didn't get removed surprisingly.~~\n https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/175156#973662\n\nEdit: My bad, the kaggle master who got called out got his account deleted but not removed from the leaderboard?? I guess the comment on the discussion was made before the account was deleted. ",
          "votes": 1
        }
      ]
    },
    {
      "id": 1759353,
      "postDate": "2022-04-18T14:59:12.423Z",
      "content": "<p>I know for sure there is  paid tutoring business.  There are rich persons who like to buy medals for \"academic honor\".</p>",
      "rawMarkdown": "I know for sure there is  paid tutoring business.  There are rich persons who like to buy medals for \"academic honor\".\n\n",
      "votes": 8,
      "replies": [
        {
          "id": 1759846,
          "postDate": "2022-04-19T00:48:40.930Z",
          "content": "<p>They can easily use medals for resumes/jobs so there's a big incentive. Clear breaking of the rules either way.</p>",
          "rawMarkdown": "They can easily use medals for resumes/jobs so there's a big incentive. Clear breaking of the rules either way.",
          "votes": 2
        }
      ]
    },
    {
      "id": 1759180,
      "postDate": "2022-04-18T12:04:28.570Z",
      "content": "<p>You should write to <a>compliance@kaggle.com</a> if you have suspicion of cheating.</p>",
      "rawMarkdown": "You should write to compliance@kaggle.com if you have suspicion of cheating.",
      "votes": 7
    },
    {
      "id": 1760359,
      "postDate": "2022-04-19T08:30:31.123Z",
      "content": "<p>Strange situation with the competition. Each comes with different purposes. I'm here because of my interest in communicating with current professionals.</p>",
      "rawMarkdown": "Strange situation with the competition. Each comes with different purposes. I'm here because of my interest in communicating with current professionals.",
      "votes": 5
    },
    {
      "id": 1762063,
      "postDate": "2022-04-20T12:11:06.100Z",
      "content": "<p>My huge congratulations to those who finally got their DESERVED medals after the LB was filtered!🙌</p>",
      "rawMarkdown": "My huge congratulations to those who finally got their DESERVED medals after the LB was filtered!🙌",
      "votes": 6
    },
    {
      "id": 1760571,
      "postDate": "2022-04-19T11:45:35.473Z",
      "content": "<p>Good work <a href=\"https://www.kaggle.com/bakeryproducts\" target=\"_blank\">@bakeryproducts</a><br>\nGaining knowledge and experience matters the most at the end of the day.</p>\n<p>I request <a href=\"https://www.kaggle.com/tedcheese\" target=\"_blank\">@tedcheese</a>, <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a> to kindly look into this and do the needful</p>",
      "rawMarkdown": "Good work @bakeryproducts\nGaining knowledge and experience matters the most at the end of the day.\n\nI request @tedcheese, @addisonhoward to kindly look into this and do the needful",
      "votes": 6
    },
    {
      "id": 1760115,
      "postDate": "2022-04-19T04:35:35.997Z",
      "content": "<p>Identify  <em>geese</em> by unique characteristics</p>",
      "rawMarkdown": "Identify ~~whales and dolphins~~ *geese* by unique characteristics",
      "votes": 6
    },
    {
      "id": 1759839,
      "postDate": "2022-04-19T00:38:38.097Z",
      "content": "<p>Dream Team<br>\n<img src=\"https://i.ibb.co/9yp2YXZ/19-04-2022-021116.jpg\" alt=\"\"></p>",
      "rawMarkdown": "Dream Team\n![](https://i.ibb.co/9yp2YXZ/19-04-2022-021116.jpg)",
      "votes": 6
    },
    {
      "id": 1759439,
      "postDate": "2022-04-18T16:02:05.123Z",
      "content": "<p>There is one more interesting thing. Many of these newbie accounts in top were online many days ago! That's really strange as well. Check it out.</p>",
      "rawMarkdown": "There is one more interesting thing. Many of these newbie accounts in top were online many days ago! That's really strange as well. Check it out.",
      "votes": 6,
      "replies": [
        {
          "id": 1759835,
          "postDate": "2022-04-19T00:35:06.230Z",
          "content": "<p>For example, take a look at current 12 place (there are 5 geese now), 4 of 5 of them last seen online 7 days ago. And the same situation with many other teams, people in which were not only online a lot of days ago, but also were online at the same time</p>",
          "rawMarkdown": "For example, take a look at current 12 place (there are 5 geese now), 4 of 5 of them last seen online 7 days ago. And the same situation with many other teams, people in which were not only online a lot of days ago, but also were online at the same time",
          "votes": 2
        }
      ]
    },
    {
      "id": 1762040,
      "postDate": "2022-04-20T12:04:42.190Z",
      "content": "<p>Wow, Your data analysis played a role, and it seems like a lot of cheating teams were eliminated.  Now our team rank improve 3 place.  &lt;25 -&gt; 22&gt; </p>",
      "rawMarkdown": "Wow, Your data analysis played a role, and it seems like a lot of cheating teams were eliminated.  Now our team rank improve 3 place.  <25 -> 22> ",
      "votes": 3
    },
    {
      "id": 1760510,
      "postDate": "2022-04-19T10:47:12.160Z",
      "content": "<p>Seems You are initiating a new paradigm where we will use Miracle Learning ….🎉🎉 </p>",
      "rawMarkdown": "Seems You are initiating a new paradigm where we will use Miracle Learning ....🎉🎉 ",
      "votes": 3
    },
    {
      "id": 1762173,
      "postDate": "2022-04-20T13:33:41.190Z",
      "content": "<p>25 teams got removed. 2 gold teams among them!<br>\ncongrats to all.</p>",
      "rawMarkdown": "25 teams got removed. 2 gold teams among them!\ncongrats to all.",
      "votes": 4,
      "replies": [
        {
          "id": 1762428,
          "postDate": "2022-04-20T17:14:07.910Z",
          "content": "<p>let's say thanks to kaggle team. and congratulations to Isamu.</p>",
          "rawMarkdown": "let's say thanks to kaggle team. and congratulations to Isamu.",
          "votes": 6
        }
      ]
    },
    {
      "id": 1759898,
      "postDate": "2022-04-19T01:25:57.110Z",
      "content": "<p>I believe Kaggle usually removes fraudulent teams when validating the final results; I'm sure they do analysis similar to this.</p>\n<p>My position usually changes about 2-3 places from competition closes to announcing final results, which I assume means they cull that many teams.</p>",
      "rawMarkdown": "I believe Kaggle usually removes fraudulent teams when validating the final results; I'm sure they do analysis similar to this.\n\nMy position usually changes about 2-3 places from competition closes to announcing final results, which I assume means they cull that many teams.",
      "votes": 4
    },
    {
      "id": 1759636,
      "postDate": "2022-04-18T19:15:15.783Z",
      "content": "<p>So many geese together!</p>",
      "rawMarkdown": "So many geese together!",
      "votes": 4,
      "replies": [
        {
          "id": 1759652,
          "postDate": "2022-04-18T19:27:51.127Z",
          "content": "<p>I think most of such accounts are created only to have concurrent access to GPU/TPU. In this competition blending plays significant role - wide range of models are beneficial here.</p>",
          "rawMarkdown": "I think most of such accounts are created only to have concurrent access to GPU/TPU. In this competition blending plays significant role - wide range of models are beneficial here.",
          "votes": 3
        }
      ]
    },
    {
      "id": 1759233,
      "postDate": "2022-04-18T13:12:35.923Z",
      "content": "<p>Great statistics, Gleb!<br>\nWhen I went through some accounts of those strange teams I also found interesting data about their former activity.  These findings give evidence that similar practices didn't appear so suddenly and aren't limited only on current novices/contributors.<br>\nOne team for all: <br>\nLeeHann, a member of the <strong>sarahclarks</strong> team (joined 3 months ago) already has 2 medals with only 4 days of activity (counted before this competition)!<br>\nCao,Xianzhi this year 2 medals with 7 days of activity!</p>\n<p>And it is similar for all the \"experts\" in those strange teams, e.g. team Jesus, team MilkMilk, team Gilber…</p>",
      "rawMarkdown": "Great statistics, Gleb!\nWhen I went through some accounts of those strange teams I also found interesting data about their former activity.  These findings give evidence that similar practices didn't appear so suddenly and aren't limited only on current novices/contributors.\nOne team for all: \nLeeHann, a member of the **sarahclarks** team (joined 3 months ago) already has 2 medals with only 4 days of activity (counted before this competition)!\nCao,Xianzhi this year 2 medals with 7 days of activity!\n\nAnd it is similar for all the \"experts\" in those strange teams, e.g. team Jesus, team MilkMilk, team Gilber...",
      "votes": 3,
      "replies": [
        {
          "id": 1760408,
          "postDate": "2022-04-19T09:07:54.687Z",
          "content": "<p>I am a computer vision engineer and the organizer of our team. This is my first Kaggle experience and I met my current teammates on Twitter and they helped us very much and got a very good ranking.</p>\n<p>I don't know the past experience of my teammates, <strong>but I am hundred percent sure that we have no cheating in this competition and did not receive any help from commercial institutions.</strong></p>\n<p>I hate all cheating and deception as much as you do, and I take responsibility for what I say.</p>",
          "rawMarkdown": "I am a computer vision engineer and the organizer of our team. This is my first Kaggle experience and I met my current teammates on Twitter and they helped us very much and got a very good ranking.\n\nI don't know the past experience of my teammates, **but I am hundred percent sure that we have no cheating in this competition and did not receive any help from commercial institutions.**\n\nI hate all cheating and deception as much as you do, and I take responsibility for what I say."
        },
        {
          "id": 1760462,
          "postDate": "2022-04-19T10:08:02.883Z",
          "content": "<p><a href=\"https://www.kaggle.com/sarahclarks\" target=\"_blank\">@sarahclarks</a> Can you share with everyone your team submission history? </p>",
          "rawMarkdown": "@sarahclarks Can you share with everyone your team submission history? ",
          "votes": 2
        },
        {
          "id": 1760511,
          "postDate": "2022-04-19T10:48:17.927Z",
          "content": "<p>If kaggle officials need us to prove anything, we will provide it in the form of emails, including screenshots of private messages from twiiter when we team up and chat logs from slack during the competition.🙂</p>",
          "rawMarkdown": "If kaggle officials need us to prove anything, we will provide it in the form of emails, including screenshots of private messages from twiiter when we team up and chat logs from slack during the competition.🙂"
        },
        {
          "id": 1760516,
          "postDate": "2022-04-19T10:56:36.830Z",
          "content": "<p>I doubt kaggle officials speak chinese, but we all speak submissionhistorese!</p>",
          "rawMarkdown": "I doubt kaggle officials speak chinese, but we all speak submissionhistorese!",
          "votes": 1
        },
        {
          "id": 1760518,
          "postDate": "2022-04-19T11:03:00.633Z",
          "content": "<p>Good note again, Gleb.<br>\nWe may have accidentally <strong>found the root</strong> of all the miraculous Easter teams' improvements. The activity statistics can really be the key. <br>\n\"sarahclarks\" probably still solo made the first 4 submissions on April 2-nd.</p>",
          "rawMarkdown": "Good note again, Gleb.\nWe may have accidentally **found the root** of all the miraculous Easter teams' improvements. The activity statistics can really be the key. \n\"sarahclarks\" probably still solo made the first 4 submissions on April 2-nd."
        },
        {
          "id": 1760557,
          "postDate": "2022-04-19T11:34:15.217Z",
          "content": "<p><a href=\"https://www.kaggle.com/sarahclarks\" target=\"_blank\">@sarahclarks</a>,<br>\nwhy so much anger and the coward downvotes, I only cite the publicly available data.<br>\nIf you really are who you claim to be, in my opinion it can be a case of abused novice's naivety.</p>\n<blockquote>\n  <p>I take responsibility for what I say.</p>\n</blockquote>\n<p>First you should take responsibility for who you team up with. Basic search of former activity is highly recommended.<br>\nAnd if you don't know your teammates well you can never say who else they share your solutions behind your back with. There were many similar cases discussed in the past.</p>",
          "rawMarkdown": "@sarahclarks,\nwhy so much anger and the coward downvotes, I only cite the publicly available data.\nIf you really are who you claim to be, in my opinion it can be a case of abused novice's naivety.\n\n> I take responsibility for what I say.\n\nFirst you should take responsibility for who you team up with. Basic search of former activity is highly recommended.\nAnd if you don't know your teammates well you can never say who else they share your solutions behind your back with. There were many similar cases discussed in the past.",
          "votes": 2
        },
        {
          "id": 1761196,
          "postDate": "2022-04-19T18:09:01.393Z",
          "content": "<p>Thank you my downvoters to give the best evidence that we probably hit the target.</p>\n<p><a href=\"https://www.kaggle.com/wcukierski\" target=\"_blank\">@wcukierski</a>  , <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a>  I hope that you read all this discussion thread (and also <a href=\"url\" target=\"_blank\">https://www.kaggle.com/competitions/happy-whale-and-dolphin/discussion/319273</a>) to get some inspiration where to find the roots and solutions of the current unsustainable situation. </p>",
          "rawMarkdown": "Thank you my downvoters to give the best evidence that we probably hit the target.\n\n@wcukierski  , @addisonhoward  I hope that you read all this discussion thread (and also [https://www.kaggle.com/competitions/happy-whale-and-dolphin/discussion/319273](url)) to get some inspiration where to find the roots and solutions of the current unsustainable situation. ",
          "votes": 2
        },
        {
          "id": 1762042,
          "postDate": "2022-04-20T12:05:21.047Z",
          "content": "<p>Well I guess they read the thread :). There were quite a few teams that cheated this competition. Our team jumped over 20 places from private lb to finalized private lb.</p>",
          "rawMarkdown": "Well I guess they read the thread :). There were quite a few teams that cheated this competition. Our team jumped over 20 places from private lb to finalized private lb."
        }
      ]
    },
    {
      "id": 1762048,
      "postDate": "2022-04-20T12:07:21.783Z",
      "content": "<p>Gold zone also cancelled a team.</p>",
      "rawMarkdown": "Gold zone also cancelled a team.",
      "votes": 1
    },
    {
      "id": 1759760,
      "postDate": "2022-04-18T21:41:08.787Z",
      "content": "<p>Reading all these comments and looking at the original description…. I think Kaggle will have a few busy days with cleaning out all the non-compliant users….</p>",
      "rawMarkdown": "Reading all these comments and looking at the original description.... I think Kaggle will have a few busy days with cleaning out all the non-compliant users....",
      "votes": 2,
      "replies": [
        {
          "id": 1759797,
          "postDate": "2022-04-18T22:48:17.020Z",
          "content": "<p>I have serious doubts that anyone would really dig deep into it, much less investigate</p>",
          "rawMarkdown": "I have serious doubts that anyone would really dig deep into it, much less investigate",
          "votes": 1
        }
      ]
    },
    {
      "id": 1767219,
      "postDate": "2022-04-25T07:16:28.107Z",
      "content": "<p>Now, this is a real data scientist at work. Neat piece of data investigation/EDA. Kaggle staff need to take cognizance of the observations</p>",
      "rawMarkdown": "Now, this is a real data scientist at work. Neat piece of data investigation/EDA. Kaggle staff need to take cognizance of the observations"
    },
    {
      "id": 1764471,
      "postDate": "2022-04-22T14:08:02.700Z",
      "content": "<p>yes, great work</p>",
      "rawMarkdown": "yes, great work"
    },
    {
      "id": 1764088,
      "postDate": "2022-04-22T06:20:44.797Z",
      "content": "<p>Wow, Interesting!</p>",
      "rawMarkdown": "Wow, Interesting!"
    },
    {
      "id": 1763529,
      "postDate": "2022-04-21T15:34:51.737Z",
      "content": "<p>really good</p>",
      "rawMarkdown": "really good"
    },
    {
      "id": 1763376,
      "postDate": "2022-04-21T13:20:54.263Z",
      "content": "<p>Hello World\\</p>",
      "rawMarkdown": "Hello World\\\n"
    },
    {
      "id": 1763183,
      "postDate": "2022-04-21T10:21:58.403Z",
      "content": "<p>eeeeeeeeeee</p>",
      "rawMarkdown": "eeeeeeeeeee"
    },
    {
      "id": 1762679,
      "postDate": "2022-04-20T22:52:06.800Z",
      "content": "<p>Wow, that is very surprising to see. I wonder how often this strategy is used?</p>",
      "rawMarkdown": "Wow, that is very surprising to see. I wonder how often this strategy is used?"
    },
    {
      "id": 1762424,
      "postDate": "2022-04-20T17:08:57.040Z",
      "content": "<p>Need to replace these lines with a blank space?</p>",
      "rawMarkdown": "Need to replace these lines with a blank space?"
    },
    {
      "id": 1761554,
      "postDate": "2022-04-20T02:08:12.590Z",
      "content": "<p>great work!!</p>",
      "rawMarkdown": "great work!!"
    },
    {
      "id": 1760281,
      "postDate": "2022-04-19T07:21:32.367Z",
      "content": "<p>good finding</p>",
      "rawMarkdown": "good finding"
    },
    {
      "id": 1759700,
      "postDate": "2022-04-18T20:17:27.273Z",
      "content": "<p>Seems Interesting. </p>",
      "rawMarkdown": "Seems Interesting. "
    },
    {
      "id": 1759358,
      "postDate": "2022-04-18T15:04:09.697Z",
      "content": "<p>A lot of coincidence</p>",
      "rawMarkdown": "A lot of coincidence"
    },
    {
      "id": 2934588,
      "postDate": "2024-07-24T15:01:57.507Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1766196,
      "postDate": "2022-04-24T09:34:02.993Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1765507,
      "postDate": "2022-04-23T15:24:59.457Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1759262,
      "postDate": "2022-04-18T13:43:59.280Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1759237,
      "postDate": "2022-04-18T13:14:29.460Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1765163,
      "postDate": "2022-04-23T08:27:30.100Z",
      "content": "<p>thanks for the kaggleteam</p>",
      "rawMarkdown": "thanks for the kaggleteam"
    },
    {
      "id": 1761472,
      "postDate": "2022-04-20T00:55:47.257Z",
      "content": "<p>Thank You for your sharing</p>",
      "rawMarkdown": "Thank You for your sharing"
    }
  ],
  "comments": [
    {
      "id": 1759238,
      "author_name": "Psi",
      "author_url": "",
      "post_date": "2022-04-18T13:17:36.997000",
      "content": "<p>One of the many reasons why I believe the time of csv competitions should be finally over.</p>",
      "votes": 32,
      "replies": [
        {
          "id": 1759255,
          "author_name": "Ilya Dobrynin",
          "author_url": "",
          "post_date": "2022-04-18T13:41:05.760000",
          "content": "<p>Sad but true, csv-like competitions way too easy to break :(</p>",
          "votes": 3,
          "replies": []
        },
        {
          "id": 1759283,
          "author_name": "Ilya Dobrynin",
          "author_url": "",
          "post_date": "2022-04-18T13:56:58.470000",
          "content": "<p>There is only hope that such cheaters quite easy to identify and drop from final LB.<br>\nOtherwise the value of kaggle medals and ranks is greatly depreciated.</p>",
          "votes": 9,
          "replies": []
        },
        {
          "id": 1759383,
          "author_name": "outwrest",
          "author_url": "",
          "post_date": "2022-04-18T15:27:02.740000",
          "content": "<p>CSV competitions just make it easier for them to cheat and hide it. I saw some really suspicious accounts in one team, all made 7 months ago played in similar competitions with other accounts made 7 months ago and all got similar rankings (bronze~silver). It is clear that this is an ongoing problem :( and it is very prominent in CSV competitions. </p>",
          "votes": 7,
          "replies": []
        },
        {
          "id": 1759740,
          "author_name": "CPMP",
          "author_url": "",
          "post_date": "2022-04-18T21:08:07.713000",
          "content": "<p><a href=\"https://www.kaggle.com/psi\" target=\"_blank\">@psi</a> not sure this is the explanation. You just reported a similar issue in feedback competition which was a code competition.</p>\n<p>This said I am now in favor of code comps for other reasons:</p>\n<ol>\n<li>a bit more fair to those with less compute power than others.</li>\n<li>harder to exploit test data distribution to game the metric.</li>\n</ol>",
          "votes": 6,
          "replies": []
        },
        {
          "id": 1759794,
          "author_name": "Psi",
          "author_url": "",
          "post_date": "2022-04-18T22:39:52.853000",
          "content": "<p>Yes you are right, I only noticed the Feedback stuff afterwards. And again, just observance, no claim.<br>\nIt is still easier in csv competitions.</p>\n<p>Totally agree with your other points. </p>",
          "votes": 3,
          "replies": []
        },
        {
          "id": 1759817,
          "author_name": "CPMP",
          "author_url": "",
          "post_date": "2022-04-18T23:48:01.467000",
          "content": "<p>impressive last minute progression!</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 1760906,
          "author_name": "Jordan Carlen",
          "author_url": "",
          "post_date": "2022-04-19T15:30:19.927000",
          "content": "<p>Hi Psi, new to Kaggle!</p>\n<p>I understand the lack of transparency when it comes to participation in competitions can lead to oddities and fly in the face of why we're all here, to grow and learn. That said, could you please expand a little bit more on why csv competitions should be discoured beyond the given reason for this topic?</p>\n<p>Thanks!</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1762358,
          "author_name": "Psi",
          "author_url": "",
          "post_date": "2022-04-20T16:17:32.927000",
          "content": "<p>Hi <a href=\"https://www.kaggle.com/jordancarlen\" target=\"_blank\">@jordancarlen</a>  - I think <a href=\"https://www.kaggle.com/cpmpml\" target=\"_blank\">@cpmpml</a> already nicely elaborated on a few of them. </p>\n<p>Some other benefits of code competitions are:</p>\n<ul>\n<li>It restricts mindless blending, which has no practicality for any production scenario and can also be quite useless to the host.  </li>\n<li>You are learning some engineering aspects that can be very valuable in your job as a data scientist</li>\n<li>It makes replicability of your solution easier</li>\n</ul>",
          "votes": 3,
          "replies": []
        }
      ]
    },
    {
      "id": 1759154,
      "author_name": "Leon",
      "author_url": "",
      "post_date": "2022-04-18T11:36:40.407000",
      "content": "<p>They seem to mostly hide their countries, which is unusual on kaggle.</p>\n<p>Also, I found some paid tutoring services (experienced players with newbies who pay for medals), which may be the source of cheating: <br>\n<a href=\"https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156\" target=\"_blank\">https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156</a><br>\n<img src=\"https://i.ibb.co/wLC0kJj/043cb377ec04c713bea60f0c94084da.png\" alt=\"1\"><br>\n<img src=\"https://i.ibb.co/tLrQFJW/ed328472e89ee56c87f68e4730b2db4.png\" alt=\"2\"></p>",
      "votes": 25,
      "replies": [
        {
          "id": 1759165,
          "author_name": "老肥",
          "author_url": "",
          "post_date": "2022-04-18T11:47:48.533000",
          "content": "<p><a href=\"https://www.linstitute.net/archives/513144\" target=\"_blank\">https://www.linstitute.net/archives/513144</a><br>\n<a href=\"https://www.xiaohongshu.com/discovery/item/624ab6550000000001024cca\" target=\"_blank\">https://www.xiaohongshu.com/discovery/item/624ab6550000000001024cca</a></p>\n<p>I also found some links.</p>",
          "votes": 8,
          "replies": []
        },
        {
          "id": 1759207,
          "author_name": "Remek Kinas",
          "author_url": "",
          "post_date": "2022-04-18T12:38:24.080000",
          "content": "<p>Not bad :) My notebook mentioned and shared :) <br>\nI have to join this school … to improve my skills :) </p>\n<blockquote>\n  <p>对CV感兴趣的同学不要错过了</p>\n</blockquote>\n<p>I will join next time :)</p>",
          "votes": 16,
          "replies": []
        },
        {
          "id": 1759272,
          "author_name": "Yangranran",
          "author_url": "",
          "post_date": "2022-04-18T13:49:27.977000",
          "content": "<p>That cracks me up. 😂</p>",
          "votes": 4,
          "replies": []
        },
        {
          "id": 1760191,
          "author_name": "Time Master",
          "author_url": "",
          "post_date": "2022-04-19T05:57:42.503000",
          "content": "<p>…….😂😂</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1759781,
      "author_name": "Aleksandr Snorkin",
      "author_url": "",
      "post_date": "2022-04-18T22:25:27.153000",
      "content": "<p>Kaggle should organize a fraud detection competition for kaggle!<br>\nThis competition will provide a lot of data :)</p>",
      "votes": 19,
      "replies": [
        {
          "id": 1759802,
          "author_name": "alexkfd",
          "author_url": "",
          "post_date": "2022-04-18T22:56:16.727000",
          "content": "<p>and maybe we'll see a record high level of \"geese teams\" there in top)</p>",
          "votes": 3,
          "replies": []
        }
      ]
    },
    {
      "id": 1759843,
      "author_name": "Isamu",
      "author_url": "",
      "post_date": "2022-04-19T00:46:17.723000",
      "content": "<p>🙏🙏🙏🙏🙏🙏😭</p>",
      "votes": 17,
      "replies": [
        {
          "id": 1762182,
          "author_name": "SonChangYup",
          "author_url": "",
          "post_date": "2022-04-20T13:42:06.163000",
          "content": "<p>congratulation!!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1759660,
      "author_name": "Evgenii Kononenko",
      "author_url": "",
      "post_date": "2022-04-18T19:35:16.660000",
      "content": "<p>Just add some \"Miracle\" here.  <br>\nLook at <strong>Lindapanpan</strong> team and when they submitted.  Until April 9, all team members were inactive at least a month. Then, on April 9, they all decided to make sabmitions, April 10 - silence, April 11 again a lot of sabmitions from all team members. And then team up….  </p>\n<p>May be it is \"Miracle of Coinscedence\"<br>\nBut look at <strong>Wesley</strong> team - almost the same situation, even the same dates …</p>",
      "votes": 12,
      "replies": [
        {
          "id": 1759708,
          "author_name": "Psi",
          "author_url": "",
          "post_date": "2022-04-18T20:23:52.823000",
          "content": "<p>Some of them also participated in Feedback competition, and placed right next to each other at around rank ~100. Now they are in the same team :)</p>",
          "votes": 14,
          "replies": []
        }
      ]
    },
    {
      "id": 1761005,
      "author_name": "CPMP",
      "author_url": "",
      "post_date": "2022-04-19T16:23:08.677000",
      "content": "<blockquote>\n  <p>While I fail at deep learning, </p>\n</blockquote>\n<p>well, if being 19th in a highly competitive challenge is failing then lots of people are failing with you!</p>",
      "votes": 9,
      "replies": [
        {
          "id": 1762429,
          "author_name": "Gleb",
          "author_url": "",
          "post_date": "2022-04-20T17:16:06.507000",
          "content": "<p>thank you cpmp, from all failing people. we will get better, learning from the leaders</p>",
          "votes": 2,
          "replies": []
        }
      ]
    },
    {
      "id": 1759163,
      "author_name": "Tom Aindow",
      "author_url": "",
      "post_date": "2022-04-18T11:44:21.823000",
      "content": "<p>Real shame, there's obviously some private sharing going on at the very least. Hopefully Kaggle can whittle it down, but there will always be some remaining who broke the rules undetected. Regardless, it's been a fun competition: learnt a lot of cool new stuff and can't wait to see the winning write ups.</p>",
      "votes": 10,
      "replies": []
    },
    {
      "id": 1760006,
      "author_name": "Qishen Ha",
      "author_url": "",
      "post_date": "2022-04-19T02:59:41.210000",
      "content": "<p>I guess those students cannot get their money back if they were removed from the LB.</p>\n<p><code>0 risk</code> for those institution 😂</p>",
      "votes": 9,
      "replies": []
    },
    {
      "id": 1760120,
      "author_name": "Geoncs",
      "author_url": "",
      "post_date": "2022-04-19T04:37:53.300000",
      "content": "<p>I agree with other comments about making the analysis of cheating a part of the competition, with the end result being a more authentic competition. It's a shame what folks will do with/for money.</p>",
      "votes": 7,
      "replies": []
    },
    {
      "id": 1759146,
      "author_name": "Nischay Dhankhar",
      "author_url": "",
      "post_date": "2022-04-18T11:26:29.103000",
      "content": "<p><a href=\"https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635\" target=\"_blank\">https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635</a><br>\nMelanoma competition V2.0?</p>",
      "votes": 7,
      "replies": [
        {
          "id": 1759150,
          "author_name": "老肥",
          "author_url": "",
          "post_date": "2022-04-18T11:29:37.043000",
          "content": "<p>Seems like group behavior.<br>\nWhat is the result of Melanoma competition， will these suspicious accounts be removed from this competition?</p>",
          "votes": 3,
          "replies": []
        },
        {
          "id": 1759155,
          "author_name": "Nischay Dhankhar",
          "author_url": "",
          "post_date": "2022-04-18T11:37:42.130000",
          "content": "<p>If I could recall correctly they got removed when the private leaderboard was finalized, although the shakeup was huge so most of them placed worse on the leaderboard. </p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 1759160,
          "author_name": "Ari",
          "author_url": "",
          "post_date": "2022-04-18T11:40:45.927000",
          "content": "<p><br>\n <a href=\"https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/175156#973662\" target=\"_blank\">https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/175156#973662</a></p>\n<p>Edit: My bad, the kaggle master who got called out got his account deleted but not removed from the leaderboard?? I guess the comment on the discussion was made before the account was deleted. </p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 1759353,
      "author_name": "dragon zhang",
      "author_url": "",
      "post_date": "2022-04-18T14:59:12.423000",
      "content": "<p>I know for sure there is  paid tutoring business.  There are rich persons who like to buy medals for \"academic honor\".</p>",
      "votes": 8,
      "replies": [
        {
          "id": 1759846,
          "author_name": "outwrest",
          "author_url": "",
          "post_date": "2022-04-19T00:48:40.930000",
          "content": "<p>They can easily use medals for resumes/jobs so there's a big incentive. Clear breaking of the rules either way.</p>",
          "votes": 2,
          "replies": []
        }
      ]
    },
    {
      "id": 1759180,
      "author_name": "CPMP",
      "author_url": "",
      "post_date": "2022-04-18T12:04:28.570000",
      "content": "<p>You should write to <a>compliance@kaggle.com</a> if you have suspicion of cheating.</p>",
      "votes": 7,
      "replies": []
    },
    {
      "id": 1760359,
      "author_name": "Alex Parkhomenko",
      "author_url": "",
      "post_date": "2022-04-19T08:30:31.123000",
      "content": "<p>Strange situation with the competition. Each comes with different purposes. I'm here because of my interest in communicating with current professionals.</p>",
      "votes": 5,
      "replies": []
    },
    {
      "id": 1762063,
      "author_name": "alexkfd",
      "author_url": "",
      "post_date": "2022-04-20T12:11:06.100000",
      "content": "<p>My huge congratulations to those who finally got their DESERVED medals after the LB was filtered!🙌</p>",
      "votes": 6,
      "replies": []
    },
    {
      "id": 1760571,
      "author_name": "Balaji Selvaraj",
      "author_url": "",
      "post_date": "2022-04-19T11:45:35.473000",
      "content": "<p>Good work <a href=\"https://www.kaggle.com/bakeryproducts\" target=\"_blank\">@bakeryproducts</a><br>\nGaining knowledge and experience matters the most at the end of the day.</p>\n<p>I request <a href=\"https://www.kaggle.com/tedcheese\" target=\"_blank\">@tedcheese</a>, <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a> to kindly look into this and do the needful</p>",
      "votes": 6,
      "replies": []
    },
    {
      "id": 1760115,
      "author_name": "xhlulu",
      "author_url": "",
      "post_date": "2022-04-19T04:35:35.997000",
      "content": "<p>Identify  <em>geese</em> by unique characteristics</p>",
      "votes": 6,
      "replies": []
    },
    {
      "id": 1759839,
      "author_name": "alexkfd",
      "author_url": "",
      "post_date": "2022-04-19T00:38:38.097000",
      "content": "<p>Dream Team<br>\n<img src=\"https://i.ibb.co/9yp2YXZ/19-04-2022-021116.jpg\" alt=\"\"></p>",
      "votes": 6,
      "replies": []
    },
    {
      "id": 1759439,
      "author_name": "Vadim Timakin",
      "author_url": "",
      "post_date": "2022-04-18T16:02:05.123000",
      "content": "<p>There is one more interesting thing. Many of these newbie accounts in top were online many days ago! That's really strange as well. Check it out.</p>",
      "votes": 6,
      "replies": [
        {
          "id": 1759835,
          "author_name": "Vadim Timakin",
          "author_url": "",
          "post_date": "2022-04-19T00:35:06.230000",
          "content": "<p>For example, take a look at current 12 place (there are 5 geese now), 4 of 5 of them last seen online 7 days ago. And the same situation with many other teams, people in which were not only online a lot of days ago, but also were online at the same time</p>",
          "votes": 2,
          "replies": []
        }
      ]
    },
    {
      "id": 1762040,
      "author_name": "KKY",
      "author_url": "",
      "post_date": "2022-04-20T12:04:42.190000",
      "content": "<p>Wow, Your data analysis played a role, and it seems like a lot of cheating teams were eliminated.  Now our team rank improve 3 place.  &lt;25 -&gt; 22&gt; </p>",
      "votes": 3,
      "replies": []
    },
    {
      "id": 1760510,
      "author_name": "Iqbal Ali",
      "author_url": "",
      "post_date": "2022-04-19T10:47:12.160000",
      "content": "<p>Seems You are initiating a new paradigm where we will use Miracle Learning ….🎉🎉 </p>",
      "votes": 3,
      "replies": []
    },
    {
      "id": 1762173,
      "author_name": "Tian",
      "author_url": "",
      "post_date": "2022-04-20T13:33:41.190000",
      "content": "<p>25 teams got removed. 2 gold teams among them!<br>\ncongrats to all.</p>",
      "votes": 4,
      "replies": [
        {
          "id": 1762428,
          "author_name": "Gleb",
          "author_url": "",
          "post_date": "2022-04-20T17:14:07.910000",
          "content": "<p>let's say thanks to kaggle team. and congratulations to Isamu.</p>",
          "votes": 6,
          "replies": []
        }
      ]
    },
    {
      "id": 1759898,
      "author_name": "Lex Toumbourou",
      "author_url": "",
      "post_date": "2022-04-19T01:25:57.110000",
      "content": "<p>I believe Kaggle usually removes fraudulent teams when validating the final results; I'm sure they do analysis similar to this.</p>\n<p>My position usually changes about 2-3 places from competition closes to announcing final results, which I assume means they cull that many teams.</p>",
      "votes": 4,
      "replies": []
    },
    {
      "id": 1759636,
      "author_name": "Filemon",
      "author_url": "",
      "post_date": "2022-04-18T19:15:15.783000",
      "content": "<p>So many geese together!</p>",
      "votes": 4,
      "replies": [
        {
          "id": 1759652,
          "author_name": "Remek Kinas",
          "author_url": "",
          "post_date": "2022-04-18T19:27:51.127000",
          "content": "<p>I think most of such accounts are created only to have concurrent access to GPU/TPU. In this competition blending plays significant role - wide range of models are beneficial here.</p>",
          "votes": 3,
          "replies": []
        }
      ]
    },
    {
      "id": 1759233,
      "author_name": "Allie K.",
      "author_url": "",
      "post_date": "2022-04-18T13:12:35.923000",
      "content": "<p>Great statistics, Gleb!<br>\nWhen I went through some accounts of those strange teams I also found interesting data about their former activity.  These findings give evidence that similar practices didn't appear so suddenly and aren't limited only on current novices/contributors.<br>\nOne team for all: <br>\nLeeHann, a member of the <strong>sarahclarks</strong> team (joined 3 months ago) already has 2 medals with only 4 days of activity (counted before this competition)!<br>\nCao,Xianzhi this year 2 medals with 7 days of activity!</p>\n<p>And it is similar for all the \"experts\" in those strange teams, e.g. team Jesus, team MilkMilk, team Gilber…</p>",
      "votes": 3,
      "replies": [
        {
          "id": 1760408,
          "author_name": "sarahclarks",
          "author_url": "",
          "post_date": "2022-04-19T09:07:54.687000",
          "content": "<p>I am a computer vision engineer and the organizer of our team. This is my first Kaggle experience and I met my current teammates on Twitter and they helped us very much and got a very good ranking.</p>\n<p>I don't know the past experience of my teammates, <strong>but I am hundred percent sure that we have no cheating in this competition and did not receive any help from commercial institutions.</strong></p>\n<p>I hate all cheating and deception as much as you do, and I take responsibility for what I say.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1760462,
          "author_name": "Gleb",
          "author_url": "",
          "post_date": "2022-04-19T10:08:02.883000",
          "content": "<p><a href=\"https://www.kaggle.com/sarahclarks\" target=\"_blank\">@sarahclarks</a> Can you share with everyone your team submission history? </p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 1760511,
          "author_name": "sarahclarks",
          "author_url": "",
          "post_date": "2022-04-19T10:48:17.927000",
          "content": "<p>If kaggle officials need us to prove anything, we will provide it in the form of emails, including screenshots of private messages from twiiter when we team up and chat logs from slack during the competition.🙂</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1760516,
          "author_name": "Gleb",
          "author_url": "",
          "post_date": "2022-04-19T10:56:36.830000",
          "content": "<p>I doubt kaggle officials speak chinese, but we all speak submissionhistorese!</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 1760518,
          "author_name": "Allie K.",
          "author_url": "",
          "post_date": "2022-04-19T11:03:00.633000",
          "content": "<p>Good note again, Gleb.<br>\nWe may have accidentally <strong>found the root</strong> of all the miraculous Easter teams' improvements. The activity statistics can really be the key. <br>\n\"sarahclarks\" probably still solo made the first 4 submissions on April 2-nd.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1760557,
          "author_name": "Allie K.",
          "author_url": "",
          "post_date": "2022-04-19T11:34:15.217000",
          "content": "<p><a href=\"https://www.kaggle.com/sarahclarks\" target=\"_blank\">@sarahclarks</a>,<br>\nwhy so much anger and the coward downvotes, I only cite the publicly available data.<br>\nIf you really are who you claim to be, in my opinion it can be a case of abused novice's naivety.</p>\n<blockquote>\n  <p>I take responsibility for what I say.</p>\n</blockquote>\n<p>First you should take responsibility for who you team up with. Basic search of former activity is highly recommended.<br>\nAnd if you don't know your teammates well you can never say who else they share your solutions behind your back with. There were many similar cases discussed in the past.</p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 1761196,
          "author_name": "Allie K.",
          "author_url": "",
          "post_date": "2022-04-19T18:09:01.393000",
          "content": "<p>Thank you my downvoters to give the best evidence that we probably hit the target.</p>\n<p><a href=\"https://www.kaggle.com/wcukierski\" target=\"_blank\">@wcukierski</a>  , <a href=\"https://www.kaggle.com/addisonhoward\" target=\"_blank\">@addisonhoward</a>  I hope that you read all this discussion thread (and also <a href=\"url\" target=\"_blank\">https://www.kaggle.com/competitions/happy-whale-and-dolphin/discussion/319273</a>) to get some inspiration where to find the roots and solutions of the current unsustainable situation. </p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 1762042,
          "author_name": "Ari",
          "author_url": "",
          "post_date": "2022-04-20T12:05:21.047000",
          "content": "<p>Well I guess they read the thread :). There were quite a few teams that cheated this competition. Our team jumped over 20 places from private lb to finalized private lb.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1762048,
      "author_name": "lafe",
      "author_url": "",
      "post_date": "2022-04-20T12:07:21.783000",
      "content": "<p>Gold zone also cancelled a team.</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1759760,
      "author_name": "Robin Smits",
      "author_url": "",
      "post_date": "2022-04-18T21:41:08.787000",
      "content": "<p>Reading all these comments and looking at the original description…. I think Kaggle will have a few busy days with cleaning out all the non-compliant users….</p>",
      "votes": 2,
      "replies": [
        {
          "id": 1759797,
          "author_name": "alexkfd",
          "author_url": "",
          "post_date": "2022-04-18T22:48:17.020000",
          "content": "<p>I have serious doubts that anyone would really dig deep into it, much less investigate</p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 1767219,
      "author_name": "Allohvk",
      "author_url": "",
      "post_date": "2022-04-25T07:16:28.107000",
      "content": "<p>Now, this is a real data scientist at work. Neat piece of data investigation/EDA. Kaggle staff need to take cognizance of the observations</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1764471,
      "author_name": "BillyJ",
      "author_url": "",
      "post_date": "2022-04-22T14:08:02.700000",
      "content": "<p>yes, great work</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1764088,
      "author_name": "Srividya Mukundan",
      "author_url": "",
      "post_date": "2022-04-22T06:20:44.797000",
      "content": "<p>Wow, Interesting!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1763529,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-21T15:34:51.737000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1763376,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-21T13:20:54.263000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1763183,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-21T10:21:58.403000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1762679,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-20T22:52:06.800000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1762424,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-20T17:08:57.040000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1761554,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-20T02:08:12.590000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1760281,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-19T07:21:32.367000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1759700,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-18T20:17:27.273000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1759358,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-18T15:04:09.697000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2934588,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-07-24T15:01:57.507000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1766196,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-24T09:34:02.993000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1765507,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-23T15:24:59.457000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1759262,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-18T13:43:59.280000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1759237,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-18T13:14:29.460000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1765163,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-23T08:27:30.100000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1761472,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-04-20T00:55:47.257000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1759132": "While I fail at deep learning, i still can parse html's!  Shall we?\n\nLets look at top100 teams from HappyWhale competition. Out of those lets ignore teams with submission counts > 90, why not. Short list of 44 teams with 159 users to have a closer look at.\nLets count users in those teams: Counter({1: 11,  2: 3,  3: 1,  4: 6, 5: 23}). Hmm, doesnt look intuitive, 23 teams of 5 users at top100 with under 90 submissions. Quite a performance!\n\nTo the main part: for some reason can we find out when users at top100 registered at kaggle? Yes we can!  Here is a short list of dates:\n```\n  2 user(s) registered at 2021-08-22\n  3 user(s) registered at 2022-03-23\n  3 user(s) registered at 2021-07-26\n  3 user(s) registered at 2022-04-05\n  3 user(s) registered at 2022-04-11\n  4 user(s) registered at 2021-09-16\n  4 user(s) registered at 2022-02-28\n  4 user(s) registered at 2022-03-01\n  6 user(s) registered at 2022-03-20\n  6 user(s) registered at 2022-04-03\n  7 user(s) registered at 2022-01-09\n  7 user(s) registered at 2022-01-11\n  8 user(s) registered at 2022-01-24\n 10 user(s) registered at 2021-09-18\n 13 user(s) registered at 2021-09-19\n```\n27 users registred on the same date of 16-19.09.21 !  another 14 at 09-11.01.22 !  And all of them in the same competition, in top100, with same amount of submissions made, within a team of 5?  Im lost! here are some of them, with a lot of dates matching. For sure some teams are less interesting, keep in mind, all of this is just stats. \n\n(Highlighted only dates through 16-19 09 2021)\n\n| team | user0 | user1 | user2  | user3 | user4 |\n| --- | --- | --- | --- | --- | --- |\n|10sarahcla|2021-03-22 | 2021-07-15 | **2021-09-16** | 2022-01-22 | 2022-03-23|\n|13Lindapan|2021-04-23 | 2021-04-30 | 2021-07-21 | 2021-07-22 | **2021-09-19**|\n|15Wesley  |**2021-09-19** | 2022-01-09 | 2022-01-09 | 2022-01-09 | 2022-01-11|\n|24OtisBing|**2021-09-16** | 2022-01-11 | 2022-01-24 | 2022-01-24|\n|25JSPOW   |**2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19** | 2022-01-11|\n|31simba   |2021-06-28 | 2021-10-31 | 2021-11-29 | 2022-02-26 | 2022-03-20|\n|41Jesus   |**2021-09-16** | **2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19**|\n|42Kingfish|2020-08-04 | 2021-04-30 | 2022-01-13 | 2022-02-04 | 2022-04-03|\n|45BobHUN  |2022-01-24 | 2022-02-22 | 2022-02-28 | 2022-02-28|\n|47Darryl  |**2021-09-19** | **2021-09-19** | 2022-01-09 | 2022-01-23 | 2022-01-24|\n|52MilkMilk|**2021-09-16** | **2021-09-18** | **2021-09-18** | 2022-01-11 | 2022-01-24|\n|59V V     |2022-03-01 | 2022-03-01 | 2022-03-01 | 2022-03-01 | 2022-03-02|\n|61PM      |2022-03-20 | 2022-03-23 | 2022-03-23 | 2022-04-03|\n|64SpencerW|2022-03-20 | 2022-03-20 | 2022-03-20 | 2022-04-03 | 2022-04-03|\n|69RapJOHN |2021-07-26 | 2021-07-26 | 2021-07-26 | 2022-01-24|\n|72Again   |2022-01-09 | 2022-01-11 | 2022-01-11 | 2022-01-24 | 2022-03-20|\n|73gilber  |2021-08-22 | **2021-09-18** | **2021-09-18** | **2021-09-19** | **2021-09-19**|\n|75FredPlan|**2021-09-19** | 2022-01-11 | 2022-01-23 | 2022-02-22 | 2022-04-03|\n|76Frederic|**2021-09-18** | 2022-01-09 | 2022-01-09 | 2022-01-24 | 2022-02-28|\n|77World   |**2021-09-18** | **2021-09-19** | **2021-09-19** | 2022-02-28 | 2022-03-02|\n|86Hykdt   |2020-02-19 | 2021-10-21 | 2022-01-16 | 2022-04-03 | 2022-04-05|\n|  |  |  |  |  |  | \n",
    "1759238": "One of the many reasons why I believe the time of csv competitions should be finally over.",
    "1759154": "They seem to mostly hide their countries, which is unusual on kaggle.\n\nAlso, I found some paid tutoring services (experienced players with newbies who pay for medals), which may be the source of cheating: \nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156\n![1](https://i.ibb.co/wLC0kJj/043cb377ec04c713bea60f0c94084da.png)\n![2](https://i.ibb.co/tLrQFJW/ed328472e89ee56c87f68e4730b2db4.png)",
    "1759781": "Kaggle should organize a fraud detection competition for kaggle!\nThis competition will provide a lot of data :)",
    "1759843": "🙏🙏🙏🙏🙏🙏😭",
    "1759660": "Just add some \"Miracle\" here.  \nLook at **Lindapanpan** team and when they submitted.  Until April 9, all team members were inactive at least a month. Then, on April 9, they all decided to make sabmitions, April 10 - silence, April 11 again a lot of sabmitions from all team members. And then team up....  \n\nMay be it is \"Miracle of Coinscedence\"\nBut look at **Wesley** team - almost the same situation, even the same dates ...",
    "1761005": "> While I fail at deep learning, \n\nwell, if being 19th in a highly competitive challenge is failing then lots of people are failing with you!",
    "1759163": "Real shame, there's obviously some private sharing going on at the very least. Hopefully Kaggle can whittle it down, but there will always be some remaining who broke the rules undetected. Regardless, it's been a fun competition: learnt a lot of cool new stuff and can't wait to see the winning write ups.",
    "1760006": "I guess those students cannot get their money back if they were removed from the LB.\n\n`0 risk` for those institution 😂",
    "1760120": "I agree with other comments about making the analysis of cheating a part of the competition, with the end result being a more authentic competition. It's a shame what folks will do with/for money.",
    "1759146": "https://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635\nMelanoma competition V2.0?",
    "1759353": "I know for sure there is  paid tutoring business.  There are rich persons who like to buy medals for \"academic honor\".\n\n",
    "1759180": "You should write to compliance@kaggle.com if you have suspicion of cheating.",
    "1760359": "Strange situation with the competition. Each comes with different purposes. I'm here because of my interest in communicating with current professionals.",
    "1762063": "My huge congratulations to those who finally got their DESERVED medals after the LB was filtered!🙌",
    "1760571": "Good work @bakeryproducts\nGaining knowledge and experience matters the most at the end of the day.\n\nI request @tedcheese, @addisonhoward to kindly look into this and do the needful",
    "1760115": "Identify ~~whales and dolphins~~ *geese* by unique characteristics",
    "1759839": "Dream Team\n![](https://i.ibb.co/9yp2YXZ/19-04-2022-021116.jpg)",
    "1759439": "There is one more interesting thing. Many of these newbie accounts in top were online many days ago! That's really strange as well. Check it out.",
    "1762040": "Wow, Your data analysis played a role, and it seems like a lot of cheating teams were eliminated.  Now our team rank improve 3 place.  <25 -> 22> ",
    "1760510": "Seems You are initiating a new paradigm where we will use Miracle Learning ....🎉🎉 ",
    "1762173": "25 teams got removed. 2 gold teams among them!\ncongrats to all.",
    "1759898": "I believe Kaggle usually removes fraudulent teams when validating the final results; I'm sure they do analysis similar to this.\n\nMy position usually changes about 2-3 places from competition closes to announcing final results, which I assume means they cull that many teams.",
    "1759636": "So many geese together!",
    "1759233": "Great statistics, Gleb!\nWhen I went through some accounts of those strange teams I also found interesting data about their former activity.  These findings give evidence that similar practices didn't appear so suddenly and aren't limited only on current novices/contributors.\nOne team for all: \nLeeHann, a member of the **sarahclarks** team (joined 3 months ago) already has 2 medals with only 4 days of activity (counted before this competition)!\nCao,Xianzhi this year 2 medals with 7 days of activity!\n\nAnd it is similar for all the \"experts\" in those strange teams, e.g. team Jesus, team MilkMilk, team Gilber...",
    "1762048": "Gold zone also cancelled a team.",
    "1759760": "Reading all these comments and looking at the original description.... I think Kaggle will have a few busy days with cleaning out all the non-compliant users....",
    "1767219": "Now, this is a real data scientist at work. Neat piece of data investigation/EDA. Kaggle staff need to take cognizance of the observations",
    "1764471": "yes, great work",
    "1764088": "Wow, Interesting!",
    "1763529": "really good",
    "1763376": "Hello World\\\n",
    "1763183": "eeeeeeeeeee",
    "1762679": "Wow, that is very surprising to see. I wonder how often this strategy is used?",
    "1762424": "Need to replace these lines with a blank space?",
    "1761554": "great work!!",
    "1760281": "good finding",
    "1759700": "Seems Interesting. ",
    "1759358": "A lot of coincidence",
    "2934588": "",
    "1766196": "",
    "1765507": "",
    "1759262": "",
    "1759237": "",
    "1765163": "thanks for the kaggleteam",
    "1761472": "Thank You for your sharing"
  }
}