{
  "id": 135057,
  "title": "Know Everything About The Competition [Summary]",
  "url": "/competitions/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/135057",
  "author_name": "",
  "post_date": "2020-03-11T20:15:38.240432400Z",
  "votes": 4,
  "comment_count": 4,
  "views": 0,
  "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3045065%2Faf8a3a8c50627441dc51e9ca3cee78f0%2Flockup_cloud.png?generation=1583955428827644&amp;alt=media\" alt=\"\"></p>\n\n<p>I almost read all discussions here , and i see that many competitors confused about the rules , the timeline , stage 1 and stage 2 etc...\nso i try to write this guide to help avoiding misunderstood the problem and  solve correctly the task : </p>\n\n<h3>First of all you should know that This competition will be conducted in 2 phases :</h3>\n\n<h3>Stage 1 :</h3>\n\n<ul>\n<li>Deadline : 14th March</li>\n<li>you need to prediction the 2015-2019 seasons</li>\n<li>you should  build a model that best predicts these seasons, according data from 1985-2014 seasons</li>\n<li>if you find any external stats/resources useful, Post it publicly as a data-set on this thread</li>\n<li>organizers of the competition give us the opportunity to test the accuracy of our model, since they give us access to all the true answers to the Stage 1 , so you can reach a high score but it's useless !</li>\n<li>Finally the trained model will be used for stage 2 , so stage 2 worth the most !</li>\n</ul>\n\n<h3>Stage 2 :</h3>\n\n<ul>\n<li>Deadline : 16th March : You will only have a few days during Stage 2 in which to submit prediction files. Pairings are announced on Sunday afternoon, and we will release the Stage 2 data (including a Stage 2 sample submission file) either Sunday night or Monday .</li>\n<li>You need to forecast outcomes of all possible match-ups in 2020 NCAA games .</li>\n<li>This is the final test in this competition to predict 2020 match results .</li>\n</ul>\n\n<h3>Useful Tips :</h3>\n\n<ul>\n<li>the average Stage 2 score last five years : \n2015: 0.475455\n2016: 0.546592\n2017: 0.477630\n2018: 0.587862\n2019: 0.463723</li>\n</ul>\n\n<p>Average across five years: 0.510252</p>\n\n<ul>\n<li><p>So 0.46 is a nice benchmark to have !</p></li>\n<li><p>NCAA rule changes for 2019-2020 season : </p></li>\n</ul>\n\n<p>1 - the 3-point line was moved to the international basketball distance of 22 feet, 1¾ inches\n2 -the shot clock resets at 20 seconds after a field goal attempt hits the rim and the offensive team rebounds the ball in the front court </p>\n\n<ul>\n<li>a team that has rested longer since the last game has an advantage (set difference by DayNum);\nhow often has the team won by way of overtime (see NumOT)? Total? How many (how share from total) games in your history or in the previous season?\nmentioned by <a href=\"/vbmokin\">@vbmokin</a></li>\n</ul>\n\n<p>Thanks Everyone and best of luck ! </p>\n\n<h2>Upvote if you liked it and found useful.</h2>",
  "messages": [
    {
      "id": "769351",
      "postDate": "03/11/2020 20:15:38",
      "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3045065%2Faf8a3a8c50627441dc51e9ca3cee78f0%2Flockup_cloud.png?generation=1583955428827644&amp;alt=media\" alt=\"\"></p>\n\n<p>I almost read all discussions here , and i see that many competitors confused about the rules , the timeline , stage 1 and stage 2 etc...\nso i try to write this guide to help avoiding misunderstood the problem and  solve correctly the task : </p>\n\n<h3>First of all you should know that This competition will be conducted in 2 phases :</h3>\n\n<h3>Stage 1 :</h3>\n\n<ul>\n<li>Deadline : 14th March</li>\n<li>you need to prediction the 2015-2019 seasons</li>\n<li>you should  build a model that best predicts these seasons, according data from 1985-2014 seasons</li>\n<li>if you find any external stats/resources useful, Post it publicly as a data-set on this thread</li>\n<li>organizers of the competition give us the opportunity to test the accuracy of our model, since they give us access to all the true answers to the Stage 1 , so you can reach a high score but it's useless !</li>\n<li>Finally the trained model will be used for stage 2 , so stage 2 worth the most !</li>\n</ul>\n\n<h3>Stage 2 :</h3>\n\n<ul>\n<li>Deadline : 16th March : You will only have a few days during Stage 2 in which to submit prediction files. Pairings are announced on Sunday afternoon, and we will release the Stage 2 data (including a Stage 2 sample submission file) either Sunday night or Monday .</li>\n<li>You need to forecast outcomes of all possible match-ups in 2020 NCAA games .</li>\n<li>This is the final test in this competition to predict 2020 match results .</li>\n</ul>\n\n<h3>Useful Tips :</h3>\n\n<ul>\n<li>the average Stage 2 score last five years : \n2015: 0.475455\n2016: 0.546592\n2017: 0.477630\n2018: 0.587862\n2019: 0.463723</li>\n</ul>\n\n<p>Average across five years: 0.510252</p>\n\n<ul>\n<li><p>So 0.46 is a nice benchmark to have !</p></li>\n<li><p>NCAA rule changes for 2019-2020 season : </p></li>\n</ul>\n\n<p>1 - the 3-point line was moved to the international basketball distance of 22 feet, 1¾ inches\n2 -the shot clock resets at 20 seconds after a field goal attempt hits the rim and the offensive team rebounds the ball in the front court </p>\n\n<ul>\n<li>a team that has rested longer since the last game has an advantage (set difference by DayNum);\nhow often has the team won by way of overtime (see NumOT)? Total? How many (how share from total) games in your history or in the previous season?\nmentioned by <a href=\"/vbmokin\">@vbmokin</a></li>\n</ul>\n\n<p>Thanks Everyone and best of luck ! </p>\n\n<h2>Upvote if you liked it and found useful.</h2>",
      "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3045065%2Faf8a3a8c50627441dc51e9ca3cee78f0%2Flockup_cloud.png?generation=1583955428827644&amp;alt=media)\n\nI almost read all discussions here , and i see that many competitors confused about the rules , the timeline , stage 1 and stage 2 etc...\nso i try to write this guide to help avoiding misunderstood the problem and  solve correctly the task : \n\n### First of all you should know that This competition will be conducted in 2 phases : \n\n### Stage 1 : \n\n- Deadline : 14th March\n- you need to prediction the 2015-2019 seasons\n-  you should  build a model that best predicts these seasons, according data from 1985-2014 seasons\n- if you find any external stats/resources useful, Post it publicly as a data-set on this thread\n- organizers of the competition give us the opportunity to test the accuracy of our model, since they give us access to all the true answers to the Stage 1 , so you can reach a high score but it's useless !\n- Finally the trained model will be used for stage 2 , so stage 2 worth the most !\n\n### Stage 2 : \n\n- Deadline : 16th March : You will only have a few days during Stage 2 in which to submit prediction files. Pairings are announced on Sunday afternoon, and we will release the Stage 2 data (including a Stage 2 sample submission file) either Sunday night or Monday .\n- You need to forecast outcomes of all possible match-ups in 2020 NCAA games .\n- This is the final test in this competition to predict 2020 match results .\n\n### Useful Tips :\n - the average Stage 2 score last five years : \n  2015: 0.475455\n  2016: 0.546592\n  2017: 0.477630\n  2018: 0.587862\n  2019: 0.463723\n\nAverage across five years: 0.510252\n\n- So 0.46 is a nice benchmark to have !\n \n- NCAA rule changes for 2019-2020 season : \n\n1 - the 3-point line was moved to the international basketball distance of 22 feet, 1¾ inches\n2 -the shot clock resets at 20 seconds after a field goal attempt hits the rim and the offensive team rebounds the ball in the front court \n\n\n- a team that has rested longer since the last game has an advantage (set difference by DayNum);\nhow often has the team won by way of overtime (see NumOT)? Total? How many (how share from total) games in your history or in the previous season?\nmentioned by @vbmokin\n\nThanks Everyone and best of luck ! \n\n## Upvote if you liked it and found useful.",
      "votes": null
    },
    {
      "id": "769362",
      "postDate": "03/11/2020 20:38:57",
      "content": "<blockquote>\n  <p>DayNum &amp; NumOT can be the most useful features on stage 2 ! </p>\n</blockquote>\n\n<p>🤔 </p>",
      "rawMarkdown": "&gt; DayNum &amp; NumOT can be the most useful features on stage 2 ! \n\n🤔",
      "votes": null
    },
    {
      "id": "769372",
      "postDate": "03/11/2020 21:04:42",
      "content": "<p>thanks <a href=\"/btesch\">@btesch</a> for mention ! \ni edited the topic !</p>",
      "rawMarkdown": "thanks @btesch for mention ! \ni edited the topic !",
      "votes": null
    },
    {
      "id": "769423",
      "postDate": "03/11/2020 22:19:02",
      "content": "<p>Good that you liked my comment about DayNum &amp; NumOT: <a href=\"https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248\">https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248</a></p>",
      "rawMarkdown": "Good that you liked my comment about DayNum &amp; NumOT: https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248",
      "votes": null
    },
    {
      "id": "769427",
      "postDate": "03/11/2020 22:28:24",
      "content": "<p><a href=\"/vbmokin\">@vbmokin</a> \nOf course it is one of the most informative topics</p>",
      "rawMarkdown": "vbmokin \nOf course it is one of the most informative topics",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 769362,
      "author_name": "btesch",
      "author_url": "",
      "post_date": "03/11/2020 20:38:57",
      "content": "<blockquote>\n  <p>DayNum &amp; NumOT can be the most useful features on stage 2 ! </p>\n</blockquote>\n\n<p>🤔 </p>",
      "votes": null,
      "replies": [
        {
          "id": 769372,
          "author_name": "yassinehamdaoui1",
          "author_url": "",
          "post_date": "03/11/2020 21:04:42",
          "content": "<p>thanks <a href=\"/btesch\">@btesch</a> for mention ! \ni edited the topic !</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 769423,
          "author_name": "vbmokin",
          "author_url": "",
          "post_date": "03/11/2020 22:19:02",
          "content": "<p>Good that you liked my comment about DayNum &amp; NumOT: <a href=\"https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248\">https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248</a></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 769427,
          "author_name": "yassinehamdaoui1",
          "author_url": "",
          "post_date": "03/11/2020 22:28:24",
          "content": "<p><a href=\"/vbmokin\">@vbmokin</a> \nOf course it is one of the most informative topics</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "769351": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3045065%2Faf8a3a8c50627441dc51e9ca3cee78f0%2Flockup_cloud.png?generation=1583955428827644&amp;alt=media)\n\nI almost read all discussions here , and i see that many competitors confused about the rules , the timeline , stage 1 and stage 2 etc...\nso i try to write this guide to help avoiding misunderstood the problem and  solve correctly the task : \n\n### First of all you should know that This competition will be conducted in 2 phases : \n\n### Stage 1 : \n\n- Deadline : 14th March\n- you need to prediction the 2015-2019 seasons\n-  you should  build a model that best predicts these seasons, according data from 1985-2014 seasons\n- if you find any external stats/resources useful, Post it publicly as a data-set on this thread\n- organizers of the competition give us the opportunity to test the accuracy of our model, since they give us access to all the true answers to the Stage 1 , so you can reach a high score but it's useless !\n- Finally the trained model will be used for stage 2 , so stage 2 worth the most !\n\n### Stage 2 : \n\n- Deadline : 16th March : You will only have a few days during Stage 2 in which to submit prediction files. Pairings are announced on Sunday afternoon, and we will release the Stage 2 data (including a Stage 2 sample submission file) either Sunday night or Monday .\n- You need to forecast outcomes of all possible match-ups in 2020 NCAA games .\n- This is the final test in this competition to predict 2020 match results .\n\n### Useful Tips :\n - the average Stage 2 score last five years : \n  2015: 0.475455\n  2016: 0.546592\n  2017: 0.477630\n  2018: 0.587862\n  2019: 0.463723\n\nAverage across five years: 0.510252\n\n- So 0.46 is a nice benchmark to have !\n \n- NCAA rule changes for 2019-2020 season : \n\n1 - the 3-point line was moved to the international basketball distance of 22 feet, 1¾ inches\n2 -the shot clock resets at 20 seconds after a field goal attempt hits the rim and the offensive team rebounds the ball in the front court \n\n\n- a team that has rested longer since the last game has an advantage (set difference by DayNum);\nhow often has the team won by way of overtime (see NumOT)? Total? How many (how share from total) games in your history or in the previous season?\nmentioned by @vbmokin\n\nThanks Everyone and best of luck ! \n\n## Upvote if you liked it and found useful.",
    "769362": "&gt; DayNum &amp; NumOT can be the most useful features on stage 2 ! \n\n🤔",
    "769372": "thanks @btesch for mention ! \ni edited the topic !",
    "769423": "Good that you liked my comment about DayNum &amp; NumOT: https://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/134248",
    "769427": "vbmokin \nOf course it is one of the most informative topics"
  },
  "source": "meta"
}