{"cells":[{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"My proposed change is for NFL punt plays to continue to be played under existing \nrules, except to eliminate the option for the receiving team to return punts.\n\nMy analysis was conducted using Microsoft Excel. I do not know R or Python.\nShould that invalidate my entry if I have a plausible proposal?\n\nI have never used Kaggle or kernels before. What appears below is a written summary of \nthe analysis supporting my proposed change to reduce the frequency of concussions\nin NFL games.\n\n\n\nData analyzed were:\n    \nNFL Punt Analytics Competition data provided on Kaggle site\n2016 and 2017 pre-season, regular season, and post-season games\nGame, play, player, and concussion data\nVideo footage of punt plays when a concussion occurred\n\nFrom NFL website\nFor 2016 and 2017, punting statistics\n\nFrom Play Smart, Play Safe website\nIncidence of concussion statistics, 2012-2017\n\nFrom Pro Football Reference website\nData for all plays using Team Game Finder database\n\n\n\nIn 2016 and 2017, there were 6680 punts.\n\n2977 punts were returned; 203 were muffed.\n\n3703 punts were fair caught, went into end zone or out-of-bounds, downed by \nkicking team, called dead, blocked, or faked.\n\nPenalties were called on 626 plays where punt was returned (21.0%).\n\nAnecdotal observation suggests that most penalties during punt returns \nare\u000bcalled on receiving team, negating the return, but further analysis is \nrequired.\n\nOnly 2211 non-muffed punts were returned with no penalties called\n(33.1% of 6680 punts) [includes returns with fumbles]\n\nAverage yards gained on those returns:  8.1 yards\n\n\n \n\nIn 2016 and 2017, there were 37 concussions on punt plays.\n\nBased on interpretation of punt plays in videos:\n    \nA  Concussion occurred during punt return (22)\nB  Concussion occurred but punt was not returned (5)\nC  Concussion did occur on punt play – but video inconclusive about when it occurred (8)\nD  Punt was returned on play but concussion was not related to return (2)\n\nAt least 22/37 (59.5%) of concussions on punt plays occurred during returns –\ncould be higher if contexts of type C concussions were determined.\n\n\n\n\nThere were 452 in-game concussions in all games 2016 and 2017.\n\nAt least 22 concussions occurred during punt returns, representing 4.9% \nof all in-game concussions.\n\n\nThere were 83194 plays in 2016 and 2017 (regular/post-season only).\n\n4968 of those plays were punt plays; assuming same 33.1% ratio \nfrom above = 1644 punt returns, so all other plays = 81550.\n\nOf 361 concussions in these games, at least 13 occurred during punt \nreturns in those games; remaining ones (348) occurred on all other plays.\n\nFrequency of concussion on punt plays with returns = 13/1644 = 0.79%\n\nFrequency of concussion on all other plays = 348/81550 = 0.43%\n\n\n\n\nThe analysis showed that most concussions on punt plays occurred during \npunt returns (at least 59.5% in 2016 and 2017),\n\nand that the frequency of concussion was at least 1.8 times higher \nduring punt returns (0.79%) than on all other plays (0.44%).\n\nTherefore, the proposed change is:\nTo eliminate the option for the receiving team to return punts.\n\nOther existing rules and strategies regarding punt plays would remain.\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"ff298b1df12cd73f981151cf783ca025b121e0ca"},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}