{
  "id": 371948,
  "title": "Should the model predict player Ids?",
  "url": "/competitions/nfl-player-contact-detection/discussion/371948",
  "author_name": "",
  "post_date": "2022-12-13T11:30:09.354708600Z",
  "votes": 6,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi, from what I understand reading the competition Data, the inputs will be just the contact_ids, which include the 2 player ids, game, play and step.<br>\nSo we don't input the video? <br>\nIs the idea just to check the given data csvs for the appropiate video frame and position from the players using the tracking data? Or are we suposed to be able to detect and track each player as well as predict their player id in a new video file and not check the csvs at all?</p>",
  "messages": [
    {
      "id": "2063869",
      "postDate": "12/13/2022 11:30:09",
      "content": "<p>Hi, from what I understand reading the competition Data, the inputs will be just the contact_ids, which include the 2 player ids, game, play and step.<br>\nSo we don't input the video? <br>\nIs the idea just to check the given data csvs for the appropiate video frame and position from the players using the tracking data? Or are we suposed to be able to detect and track each player as well as predict their player id in a new video file and not check the csvs at all?</p>",
      "rawMarkdown": "Hi, from what I understand reading the competition Data, the inputs will be just the contact_ids, which include the 2 player ids, game, play and step.\nSo we don't input the video? \nIs the idea just to check the given data csvs for the appropiate video frame and position from the players using the tracking data? Or are we suposed to be able to detect and track each player as well as predict their player id in a new video file and not check the csvs at all?",
      "votes": null
    },
    {
      "id": "2064137",
      "postDate": "12/13/2022 14:46:04",
      "content": "<p>Great question Richard. The goal of this competition is to predict when contact occurs between:<br>\n1) Player pairs (any moment they are in physical contact)<br>\n2) Player and the ground (When any part of the player's body except their feet or hands are touching the ground)</p>\n<p>You are asked to predict this for steps of 0.1 seconds within the play.</p>\n<p>The <code>contact_id</code> is simply a unique identifier for each moment you are expected to predict.</p>\n<p>It consists of either : </p>\n<ul>\n<li><code>{game_play}_{step}_{nfl_player_id_1}_{nfl_player_id_2}</code> (where nfl_player_id_1 &lt; nfl_player_id_2) for player pair contact <strong>or</strong></li>\n<li><code>{game_play}_{step}_{nfl_player_id_1}_G</code> for player to ground contact. <br>\nYou will need to predict for every player pair and player/ground for each time step in the play (from step 0 to the last step in tracking or video data). Once submitted your notebook will have access to a <code>sample_submission.csv</code> which contains all of the <code>contact_ids</code> you are expected to predict for. </li>\n</ul>\n<p>As for the data provided, you can use some or all of them to help you make predictions. They include:</p>\n<ul>\n<li>Tracking data which provides information about each player (position, speed, etc) at each step in the play.</li>\n<li>Video of the play from three views (Sideline, Endzone and All29). Sideline and Endzone can be synced with the <code>step</code>. All29 is not guaranteed to be time synced.</li>\n<li>Imperfect helmet boxes detected in the Sideline/Endzone videos which can be used or combined with the video to help your predictions.</li>\n</ul>\n<p>How you choose to use the provided data is up to you. One of the things we are hoping to learn from this competition are the novel approaches kagglers take in using this data.</p>\n<p>Hope that helps!</p>",
      "rawMarkdown": "Great question Richard. The goal of this competition is to predict when contact occurs between:\n1) Player pairs (any moment they are in physical contact)\n2) Player and the ground (When any part of the player's body except their feet or hands are touching the ground)\n\nYou are asked to predict this for steps of 0.1 seconds within the play.\n\nThe `contact_id` is simply a unique identifier for each moment you are expected to predict.\n\nIt consists of either : \n- `{game_play}_{step}_{nfl_player_id_1}_{nfl_player_id_2}` (where nfl_player_id_1 < nfl_player_id_2) for player pair contact **or**\n- `{game_play}_{step}_{nfl_player_id_1}_G` for player to ground contact. \nYou will need to predict for every player pair and player/ground for each time step in the play (from step 0 to the last step in tracking or video data). Once submitted your notebook will have access to a `sample_submission.csv` which contains all of the `contact_ids` you are expected to predict for. \n\nAs for the data provided, you can use some or all of them to help you make predictions. They include:\n- Tracking data which provides information about each player (position, speed, etc) at each step in the play.\n- Video of the play from three views (Sideline, Endzone and All29). Sideline and Endzone can be synced with the `step`. All29 is not guaranteed to be time synced.\n- Imperfect helmet boxes detected in the Sideline/Endzone videos which can be used or combined with the video to help your predictions.\n\nHow you choose to use the provided data is up to you. One of the things we are hoping to learn from this competition are the novel approaches kagglers take in using this data.\n\nHope that helps!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2064137,
      "author_name": "robikscube",
      "author_url": "",
      "post_date": "12/13/2022 14:46:04",
      "content": "<p>Great question Richard. The goal of this competition is to predict when contact occurs between:<br>\n1) Player pairs (any moment they are in physical contact)<br>\n2) Player and the ground (When any part of the player's body except their feet or hands are touching the ground)</p>\n<p>You are asked to predict this for steps of 0.1 seconds within the play.</p>\n<p>The <code>contact_id</code> is simply a unique identifier for each moment you are expected to predict.</p>\n<p>It consists of either : </p>\n<ul>\n<li><code>{game_play}_{step}_{nfl_player_id_1}_{nfl_player_id_2}</code> (where nfl_player_id_1 &lt; nfl_player_id_2) for player pair contact <strong>or</strong></li>\n<li><code>{game_play}_{step}_{nfl_player_id_1}_G</code> for player to ground contact. <br>\nYou will need to predict for every player pair and player/ground for each time step in the play (from step 0 to the last step in tracking or video data). Once submitted your notebook will have access to a <code>sample_submission.csv</code> which contains all of the <code>contact_ids</code> you are expected to predict for. </li>\n</ul>\n<p>As for the data provided, you can use some or all of them to help you make predictions. They include:</p>\n<ul>\n<li>Tracking data which provides information about each player (position, speed, etc) at each step in the play.</li>\n<li>Video of the play from three views (Sideline, Endzone and All29). Sideline and Endzone can be synced with the <code>step</code>. All29 is not guaranteed to be time synced.</li>\n<li>Imperfect helmet boxes detected in the Sideline/Endzone videos which can be used or combined with the video to help your predictions.</li>\n</ul>\n<p>How you choose to use the provided data is up to you. One of the things we are hoping to learn from this competition are the novel approaches kagglers take in using this data.</p>\n<p>Hope that helps!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2063869": "Hi, from what I understand reading the competition Data, the inputs will be just the contact_ids, which include the 2 player ids, game, play and step.\nSo we don't input the video? \nIs the idea just to check the given data csvs for the appropiate video frame and position from the players using the tracking data? Or are we suposed to be able to detect and track each player as well as predict their player id in a new video file and not check the csvs at all?",
    "2064137": "Great question Richard. The goal of this competition is to predict when contact occurs between:\n1) Player pairs (any moment they are in physical contact)\n2) Player and the ground (When any part of the player's body except their feet or hands are touching the ground)\n\nYou are asked to predict this for steps of 0.1 seconds within the play.\n\nThe `contact_id` is simply a unique identifier for each moment you are expected to predict.\n\nIt consists of either : \n- `{game_play}_{step}_{nfl_player_id_1}_{nfl_player_id_2}` (where nfl_player_id_1 < nfl_player_id_2) for player pair contact **or**\n- `{game_play}_{step}_{nfl_player_id_1}_G` for player to ground contact. \nYou will need to predict for every player pair and player/ground for each time step in the play (from step 0 to the last step in tracking or video data). Once submitted your notebook will have access to a `sample_submission.csv` which contains all of the `contact_ids` you are expected to predict for. \n\nAs for the data provided, you can use some or all of them to help you make predictions. They include:\n- Tracking data which provides information about each player (position, speed, etc) at each step in the play.\n- Video of the play from three views (Sideline, Endzone and All29). Sideline and Endzone can be synced with the `step`. All29 is not guaranteed to be time synced.\n- Imperfect helmet boxes detected in the Sideline/Endzone videos which can be used or combined with the video to help your predictions.\n\nHow you choose to use the provided data is up to you. One of the things we are hoping to learn from this competition are the novel approaches kagglers take in using this data.\n\nHope that helps!"
  },
  "source": "meta"
}