{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"markdown","source":"# **NFL Punt Analytics Competition**\n## *Analyze NFL game data and suggest rules to improve player safety during punt plays*\n### Team: Mark and Merrick\n### Deliverables: 1) Kaggle kernel of data analysis, 2) Deck summarizing analysis and rule change\n### Submission Deadline: January 9th, 2019\n\n\n"},{"metadata":{"_uuid":"fafcaa35caaaa59fe0678943053faac32ac007da"},"cell_type":"markdown","source":"# Data Analysis Plan:\n* Step 1: Clean datasets\n* Step 2: Specify research questions\n* Step 3: Establish inclusion criteria for study population\n* Step 4: Code qualitative data for concussion population and come up with categories\n* Step 5: Determine best statistical methods for dataset/research questions\n* Step 6: Execute statistical methods\n* Step 7: Combine quant and qual findings to deepen understanding of data\n* Step 8: Based on findings, determine rule change to reduce concussions\n* Step 9: Create deck summarizing findings and rule changes\n* Step 10: Submit Kernels and Deck to Kaggle"},{"metadata":{"_uuid":"30dbfdce1ca114082a42dc6262e24f0784ce17a3"},"cell_type":"markdown","source":"### Step 1: Clean datasets\nThe following datasets were cleaned and merged:\n* video_footage_control\n* video_footage_injury\n    * This is the dataset that will be used for qual coding\n* play_information\n* play_player_role_data\n* game_data\n    * Required the most cleaning; has very useful covariates\n* player_punt_data\n* video_review\n\n**Unique rows will be  “game ID” + “player ID” + \"GSISID\" where cleaned datasets will be combined utilizing this unique ID.**"},{"metadata":{"_uuid":"f9e3a2a39a3de650b66c5962846302901d33f234"},"cell_type":"markdown","source":"### Step 2: Specify research questions\n\n**Null Hypothesis 1:** \n* Players with ball possession and players with no ball position have the same association for increased concussions during punt plays.\n\n**Null Hypothesis 2:** \n* NFL games on turf fields and grass fields have the same association for increased concussions.\n\n**Null Hypothesis 3:** \n* 'Gunners' have the same association for increased concussion as other punt posiitons.\n\n**Null Hypothesis 4:** \n* Players on the receiving side and kicking side of the play have the same association for increased concussions."},{"metadata":{"_uuid":"7fd118142f96fc5aff4f35ed364d80de281743be"},"cell_type":"markdown","source":"### Step 3: Establish inclusion criteria for study population\n* NFL players\n* Players during the 2016 and 2017 season (includes pre, regular, post)\n* Players involved in punt plays\n"},{"metadata":{"_uuid":"004f7e4b8923568979a16a440442405a89e0ae15"},"cell_type":"markdown","source":"### Step 4: Code qual data for concussion population and come up with categories\n* After reviewing the concussion videos a second time, I came up with a few key observations:\n    * Some of the concussion occurred when the blocker engaged the gunner behind the ball carrier after the ball carrier fielded the punt\n    * Two gunners collided on a tackle, even when coming from the same direction\n    * Punting team linebackers were the most likely to be concussed.\n\nUltimately decided to not include qualitative data due to challenge of attributing video footage content to player that received the concussion,\n"},{"metadata":{"_uuid":"1ad332387cebbabcd0eaf407cf22281a15198c75"},"cell_type":"markdown","source":"### Step 5: Determine best statistical methods for dataset\n\nDue to most of the variables being binary or categorical, we were limited to what statistical tests we could use. We debated between using a chi-square test and logistic regression. Though chi-square is more \"straight-forward\", we ultimately decided on using logistic regression as we wanted to pursue the opportunity for accounting for observed confounders."},{"metadata":{"_uuid":"8a3fc5a657b807313a222e037542b00d71cad4b7"},"cell_type":"markdown","source":"### Step 6: Execute statistical methods\n[Please refer to our data analysis script kernal](https://www.kaggle.com/mellamomark/nfl-punt-analytics-analysis-code-r?scriptVersionId=9369213)"},{"metadata":{"_uuid":"e784822d2351c45f02cadbcb9fc421194103adff"},"cell_type":"markdown","source":"### Step 7: Combine quant and qual findings to deepen understanding of data\nUltimately decided to not include qualitative data due to challenge of attributing video footage content to player that received the concussion,"},{"metadata":{"_uuid":"286a921e76a3b1e04a44b9665df9e75586be5fc4"},"cell_type":"markdown","source":"### Steps 8 - 10:\n[Please refer to our data analysis notebook kernal](https://kaggle.com/mellamomark/nfl-punt-analytics-fair-catch-or-it-s-a-wrap?scriptVersionId=9374741)"}],"metadata":{"kernelspec":{"display_name":"R","language":"R","name":"ir"},"language_info":{"mimetype":"text/x-r-source","name":"R","pygments_lexer":"r","version":"3.4.2","file_extension":".r","codemirror_mode":"r"}},"nbformat":4,"nbformat_minor":1}