This is a breakdown of the proposed rule changes by popularity.
\n\n11+ proposals\n* Add new blocking/tackling penalties\n* Enforce changes to punt formations (e.g. number of jammers)\n* Incentivize fair catches
\n\n2-10\n* Change how the game is refereed (e.g. add another ref)\n* Eliminate punt returns\n* Incentivize fourth-down \"go for it\" attempts\n* Incentivize out-of-bounds punts\n* Incentivize touchbacks\n* Install a mandatory concussions training program\n* Restrict the actions of punt receivers
\n\nUnique\n* Enforce a specific turf type \n* Incentivize field goals from 60+ yards\n* Limit number of pre-season games\n* Limit number of punts per game\n* Move sidelines during punts\n* Protect punt receivers during catch
\n\nAlso, at least one kernel tries to list all possible rule changes:\nhttps://www.kaggle.com/ericfreeman/evaluation-of-proposed-punt-rules
\n\n=====
\n\nOriginal Post:\nDoes someone already have the stats? If not, I would like to collect them over the weekend.
", "messages": [ { "id": "454472", "postDate": "01/11/2019 17:24:57", "content": "This is a breakdown of the proposed rule changes by popularity.
\n\n11+ proposals\n* Add new blocking/tackling penalties\n* Enforce changes to punt formations (e.g. number of jammers)\n* Incentivize fair catches
\n\n2-10\n* Change how the game is refereed (e.g. add another ref)\n* Eliminate punt returns\n* Incentivize fourth-down \"go for it\" attempts\n* Incentivize out-of-bounds punts\n* Incentivize touchbacks\n* Install a mandatory concussions training program\n* Restrict the actions of punt receivers
\n\nUnique\n* Enforce a specific turf type \n* Incentivize field goals from 60+ yards\n* Limit number of pre-season games\n* Limit number of punts per game\n* Move sidelines during punts\n* Protect punt receivers during catch
\n\nAlso, at least one kernel tries to list all possible rule changes:\nhttps://www.kaggle.com/ericfreeman/evaluation-of-proposed-punt-rules
\n\n=====
\n\nOriginal Post:\nDoes someone already have the stats? If not, I would like to collect them over the weekend.
", "rawMarkdown": "This is a breakdown of the proposed rule changes by popularity.\n\n11+ proposals\n* Add new blocking/tackling penalties\n* Enforce changes to punt formations (e.g. number of jammers)\n* Incentivize fair catches\n\n2-10\n* Change how the game is refereed (e.g. add another ref)\n* Eliminate punt returns\n* Incentivize fourth-down \"go for it\" attempts\n* Incentivize out-of-bounds punts\n* Incentivize touchbacks\n* Install a mandatory concussions training program\n* Restrict the actions of punt receivers\n\nUnique\n* Enforce a specific turf type \n* Incentivize field goals from 60+ yards\n* Limit number of pre-season games\n* Limit number of punts per game\n* Move sidelines during punts\n* Protect punt receivers during catch\n\nAlso, at least one kernel tries to list all possible rule changes:\nhttps://www.kaggle.com/ericfreeman/evaluation-of-proposed-punt-rules\n\n=====\n\nOriginal Post:\nDoes someone already have the stats? If not, I would like to collect them over the weekend.", "votes": null }, { "id": "454689", "postDate": "01/12/2019 02:34:52", "content": "That'd be awesome!
\n\nI've read through 15 or so, and it looks like there were a lot of really awesome analyses and beautiful visualizations. However, most led back to fairly consistent and simple suggestions.
\n\nSeems like a good proportion of proposed changes included some kind of fair catch incentive, most proposing a 5-10 yard bonus. Mine included some qualification criteria regarding punt LOS having to be < kicking teams 45 yard line to keep the balance of the game/ability to pin a team deep in tact. Looks like there are also a few gunner/coverage/formation rules which I need to re-read to better understand how they address correlation/causation regarding those formations and how they relate to concussion rates. There are also some interesting proposals suggesting harsher penalties for blindside blocks, etc. I'm pretty sure a few people suggested banning punts.
\n\nIt'll be interesting to see how this pans out. There is one analysis I've seen that will almost surely make the top 4, the rest seem up for grabs. This competition is made more interesting by this article about the possibility of replacing the kickoff with a 4th and long scenario: https://slate.com/culture/2018/12/nfl-kickoff-rule-change-4th-and-15-play.html\nReally cool idea.
", "rawMarkdown": "That'd be awesome!\n\nI've read through 15 or so, and it looks like there were a lot of really awesome analyses and beautiful visualizations. However, most led back to fairly consistent and simple suggestions.\n\nSeems like a good proportion of proposed changes included some kind of fair catch incentive, most proposing a 5-10 yard bonus. Mine included some qualification criteria regarding punt LOS having to be < kicking teams 45 yard line to keep the balance of the game/ability to pin a team deep in tact. Looks like there are also a few gunner/coverage/formation rules which I need to re-read to better understand how they address correlation/causation regarding those formations and how they relate to concussion rates. There are also some interesting proposals suggesting harsher penalties for blindside blocks, etc. I'm pretty sure a few people suggested banning punts. \n\nIt'll be interesting to see how this pans out. There is one analysis I've seen that will almost surely make the top 4, the rest seem up for grabs. This competition is made more interesting by this article about the possibility of replacing the kickoff with a 4th and long scenario: https://slate.com/culture/2018/12/nfl-kickoff-rule-change-4th-and-15-play.html\nReally cool idea.", "votes": null }, { "id": "454794", "postDate": "01/12/2019 08:22:07", "content": "About correlation-causation of formations, causation is almost always elusive... And here even correlation is tricky, given the small number of events!
\n\nI screened exhaustively for Information Value and looks like punt return formations do matter a lot. That said, the question \"why does it matter?\" proved harder to answer. Of lots of possible descriptors of punt return formations (Symmetry, density of different sections) only one showed predictive power, weak but worth considering in my opinion.
", "rawMarkdown": "About correlation-causation of formations, causation is almost always elusive... And here even correlation is tricky, given the small number of events!\n\nI screened exhaustively for Information Value and looks like punt return formations do matter a lot. That said, the question \"why does it matter?\" proved harder to answer. Of lots of possible descriptors of punt return formations (Symmetry, density of different sections) only one showed predictive power, weak but worth considering in my opinion.", "votes": null }, { "id": "454960", "postDate": "01/12/2019 16:30:25", "content": "OK, I've done the work. If someone's rule proposal is missing, please, don't hesitate to inform us.
", "rawMarkdown": "OK, I've done the work. If someone's rule proposal is missing, please, don't hesitate to inform us.", "votes": null }, { "id": "455000", "postDate": "01/12/2019 18:11:15", "content": "Thanks for pulling the stats. I remember one proposal for implementing the college rule for punts (i.e, all players, not just gunners can release immediately). Here's an article about the concept. \nhttps://www.usatoday.com/story/sports/ncaaf/2014/11/20/rugby-punts-bounding-through-college-football/19311797/
", "rawMarkdown": "Thanks for pulling the stats. I remember one proposal for implementing the college rule for punts (i.e, all players, not just gunners can release immediately). Here's an article about the concept. \nhttps://www.usatoday.com/story/sports/ncaaf/2014/11/20/rugby-punts-bounding-through-college-football/19311797/", "votes": null }, { "id": "455001", "postDate": "01/12/2019 18:12:36", "content": "@miguel perez Was that the number of jammers? I struggled with correlation and on the last day changed my presentation to go with a fair catch incentive instead of jammer restrictions.
", "rawMarkdown": "miguel perez Was that the number of jammers? I struggled with correlation and on the last day changed my presentation to go with a fair catch incentive instead of jammer restrictions.", "votes": null }, { "id": "455027", "postDate": "01/12/2019 19:25:09", "content": "To me, that sounds like a formation change.
", "rawMarkdown": "To me, that sounds like a formation change.", "votes": null }, { "id": "455038", "postDate": "01/12/2019 20:02:33", "content": "@EricFreeman, for -penalized- Information Value strongest feature was not number jammers, but asymmetry of central \"box\" considering the two lines of defense, first 3 x2 players first line and 2 players second line.\n(This asymmetry is sure correlated with number of jammers though, as \"missing\" central players have to go somewhere, but strongest predictor was central box asymmetry.)
\n\nBiggest insight though was that initial formations matter. I didn't recommend enforcing limitations, though, not before more testing. But I did suggest to share with coaches that defense formations possibly are more than tactics, also related to risk. And, of course monitor the pattern real test time... only then applying a rule change that favours symmetry/compactness, to some extent.
", "rawMarkdown": "EricFreeman, for -penalized- Information Value strongest feature was not number jammers, but asymmetry of central \"box\" considering the two lines of defense, first 3 x2 players first line and 2 players second line.\n(This asymmetry is sure correlated with number of jammers though, as \"missing\" central players have to go somewhere, but strongest predictor was central box asymmetry.)\n\nBiggest insight though was that initial formations matter. I didn't recommend enforcing limitations, though, not before more testing. But I did suggest to share with coaches that defense formations possibly are more than tactics, also related to risk. And, of course monitor the pattern real test time... only then applying a rule change that favours symmetry/compactness, to some extent.", "votes": null }, { "id": "455042", "postDate": "01/12/2019 20:08:51", "content": "That's fair.
", "rawMarkdown": "That's fair.", "votes": null }, { "id": "455312", "postDate": "01/13/2019 15:15:46", "content": "My two suggestions were to:\n1) Wait Rule: Not allow gunners to leave the line of scrimmage until the punt was in the air and kicked\n2) Barrier Rule: Give the punt returner a 3-yard circular barrier on the punt catch, if a player from the Punting Team enters that barrier before the punt is caught, call the player dead where the punt is caught or where the ball hits the ground.
", "rawMarkdown": "My two suggestions were to:\n1) Wait Rule: Not allow gunners to leave the line of scrimmage until the punt was in the air and kicked\n2) Barrier Rule: Give the punt returner a 3-yard circular barrier on the punt catch, if a player from the Punting Team enters that barrier before the punt is caught, call the player dead where the punt is caught or where the ball hits the ground.", "votes": null }, { "id": "455332", "postDate": "01/13/2019 16:04:06", "content": "I want to make sure I understand your barrier rule. I'm reading it as a reward for the punting team and as a way to get awarded a tackle without contact. It's almost an extension of the in the grasp rule for QB's except \"in the grasp\" is defined as within 3 yards. Is that right?
", "rawMarkdown": "I want to make sure I understand your barrier rule. I'm reading it as a reward for the punting team and as a way to get awarded a tackle without contact. It's almost an extension of the in the grasp rule for QB's except \"in the grasp\" is defined as within 3 yards. Is that right?", "votes": null }, { "id": "455342", "postDate": "01/13/2019 16:23:02", "content": "Yes exactly. Trying to prevent those concussions at the punt catch area and after the catch situations during a run, which we know many of the concussions occur.
", "rawMarkdown": "Yes exactly. Trying to prevent those concussions at the punt catch area and after the catch situations during a run, which we know many of the concussions occur.", "votes": null }, { "id": "455463", "postDate": "01/14/2019 00:46:43", "content": "Very interesting @miguelpm , thanks for sharing. Initial formations are so heavily entangled with extraneous variables like field position, proficiency of the return team, score, etc. - it doesn't surprise me that establishing causation between formation and concussions would be a slippery proposition. I didn't dig into it heavily, but I'm glad I wasn't missing something obvious.
", "rawMarkdown": "Very interesting @miguelpm , thanks for sharing. Initial formations are so heavily entangled with extraneous variables like field position, proficiency of the return team, score, etc. - it doesn't surprise me that establishing causation between formation and concussions would be a slippery proposition. I didn't dig into it heavily, but I'm glad I wasn't missing something obvious.", "votes": null }, { "id": "455589", "postDate": "01/14/2019 07:58:55", "content": "\"Entangled\" is a perfect word for describing it @ColinMc , your intuitions are good, average riskier formations plots showed that field position mattered. But also \"normalized\" plots, disconnected from field position were different, so structure of formation matters too...! About score, I found no relationship (though it also was a good candidate for me).
\n\nSo even if there was something there, big difficulty was to find a quantifiable parameter, of many many candidates... with the time we had I only found the mentioned symmetry one but there could be more in that \"entanglement\" of features. Glad you found it interesting :-)
", "rawMarkdown": "\"Entangled\" is a perfect word for describing it @ColinMc , your intuitions are good, average riskier formations plots showed that field position mattered. But also \"normalized\" plots, disconnected from field position were different, so structure of formation matters too...! About score, I found no relationship (though it also was a good candidate for me).\n\nSo even if there was something there, big difficulty was to find a quantifiable parameter, of many many candidates... with the time we had I only found the mentioned symmetry one but there could be more in that \"entanglement\" of features. Glad you found it interesting :-)", "votes": null }, { "id": "455814", "postDate": "01/14/2019 16:36:02", "content": "To me, 1) sounds like a change to the punt formation.\nAs for the 2), how does a more generic \"protect punt receivers during catch\" sound?
", "rawMarkdown": "To me, 1) sounds like a change to the punt formation.\nAs for the 2), how does a more generic \"protect punt receivers during catch\" sound?", "votes": null } ], "comments": [ { "id": 454689, "author_name": "colin33", "author_url": "", "post_date": "01/12/2019 02:34:52", "content": "That'd be awesome!
\n\nI've read through 15 or so, and it looks like there were a lot of really awesome analyses and beautiful visualizations. However, most led back to fairly consistent and simple suggestions.
\n\nSeems like a good proportion of proposed changes included some kind of fair catch incentive, most proposing a 5-10 yard bonus. Mine included some qualification criteria regarding punt LOS having to be < kicking teams 45 yard line to keep the balance of the game/ability to pin a team deep in tact. Looks like there are also a few gunner/coverage/formation rules which I need to re-read to better understand how they address correlation/causation regarding those formations and how they relate to concussion rates. There are also some interesting proposals suggesting harsher penalties for blindside blocks, etc. I'm pretty sure a few people suggested banning punts.
\n\nIt'll be interesting to see how this pans out. There is one analysis I've seen that will almost surely make the top 4, the rest seem up for grabs. This competition is made more interesting by this article about the possibility of replacing the kickoff with a 4th and long scenario: https://slate.com/culture/2018/12/nfl-kickoff-rule-change-4th-and-15-play.html\nReally cool idea.
", "votes": null, "replies": [ { "id": 454794, "author_name": "miguelpm", "author_url": "", "post_date": "01/12/2019 08:22:07", "content": "About correlation-causation of formations, causation is almost always elusive... And here even correlation is tricky, given the small number of events!
\n\nI screened exhaustively for Information Value and looks like punt return formations do matter a lot. That said, the question \"why does it matter?\" proved harder to answer. Of lots of possible descriptors of punt return formations (Symmetry, density of different sections) only one showed predictive power, weak but worth considering in my opinion.
", "votes": null, "replies": [] }, { "id": 454960, "author_name": "sp007c", "author_url": "", "post_date": "01/12/2019 16:30:25", "content": "OK, I've done the work. If someone's rule proposal is missing, please, don't hesitate to inform us.
", "votes": null, "replies": [] }, { "id": 455001, "author_name": "ericfreeman", "author_url": "", "post_date": "01/12/2019 18:12:36", "content": "@miguel perez Was that the number of jammers? I struggled with correlation and on the last day changed my presentation to go with a fair catch incentive instead of jammer restrictions.
", "votes": null, "replies": [] }, { "id": 455038, "author_name": "miguelpm", "author_url": "", "post_date": "01/12/2019 20:02:33", "content": "@EricFreeman, for -penalized- Information Value strongest feature was not number jammers, but asymmetry of central \"box\" considering the two lines of defense, first 3 x2 players first line and 2 players second line.\n(This asymmetry is sure correlated with number of jammers though, as \"missing\" central players have to go somewhere, but strongest predictor was central box asymmetry.)
\n\nBiggest insight though was that initial formations matter. I didn't recommend enforcing limitations, though, not before more testing. But I did suggest to share with coaches that defense formations possibly are more than tactics, also related to risk. And, of course monitor the pattern real test time... only then applying a rule change that favours symmetry/compactness, to some extent.
", "votes": null, "replies": [] }, { "id": 455463, "author_name": "colin33", "author_url": "", "post_date": "01/14/2019 00:46:43", "content": "Very interesting @miguelpm , thanks for sharing. Initial formations are so heavily entangled with extraneous variables like field position, proficiency of the return team, score, etc. - it doesn't surprise me that establishing causation between formation and concussions would be a slippery proposition. I didn't dig into it heavily, but I'm glad I wasn't missing something obvious.
", "votes": null, "replies": [] }, { "id": 455589, "author_name": "miguelpm", "author_url": "", "post_date": "01/14/2019 07:58:55", "content": "\"Entangled\" is a perfect word for describing it @ColinMc , your intuitions are good, average riskier formations plots showed that field position mattered. But also \"normalized\" plots, disconnected from field position were different, so structure of formation matters too...! About score, I found no relationship (though it also was a good candidate for me).
\n\nSo even if there was something there, big difficulty was to find a quantifiable parameter, of many many candidates... with the time we had I only found the mentioned symmetry one but there could be more in that \"entanglement\" of features. Glad you found it interesting :-)
", "votes": null, "replies": [] } ] }, { "id": 455000, "author_name": "ericfreeman", "author_url": "", "post_date": "01/12/2019 18:11:15", "content": "Thanks for pulling the stats. I remember one proposal for implementing the college rule for punts (i.e, all players, not just gunners can release immediately). Here's an article about the concept. \nhttps://www.usatoday.com/story/sports/ncaaf/2014/11/20/rugby-punts-bounding-through-college-football/19311797/
", "votes": null, "replies": [ { "id": 455027, "author_name": "sp007c", "author_url": "", "post_date": "01/12/2019 19:25:09", "content": "To me, that sounds like a formation change.
", "votes": null, "replies": [] }, { "id": 455042, "author_name": "ericfreeman", "author_url": "", "post_date": "01/12/2019 20:08:51", "content": "That's fair.
", "votes": null, "replies": [] } ] }, { "id": 455312, "author_name": "predact", "author_url": "", "post_date": "01/13/2019 15:15:46", "content": "My two suggestions were to:\n1) Wait Rule: Not allow gunners to leave the line of scrimmage until the punt was in the air and kicked\n2) Barrier Rule: Give the punt returner a 3-yard circular barrier on the punt catch, if a player from the Punting Team enters that barrier before the punt is caught, call the player dead where the punt is caught or where the ball hits the ground.
", "votes": null, "replies": [ { "id": 455332, "author_name": "ericfreeman", "author_url": "", "post_date": "01/13/2019 16:04:06", "content": "I want to make sure I understand your barrier rule. I'm reading it as a reward for the punting team and as a way to get awarded a tackle without contact. It's almost an extension of the in the grasp rule for QB's except \"in the grasp\" is defined as within 3 yards. Is that right?
", "votes": null, "replies": [] }, { "id": 455342, "author_name": "predact", "author_url": "", "post_date": "01/13/2019 16:23:02", "content": "Yes exactly. Trying to prevent those concussions at the punt catch area and after the catch situations during a run, which we know many of the concussions occur.
", "votes": null, "replies": [] }, { "id": 455814, "author_name": "sp007c", "author_url": "", "post_date": "01/14/2019 16:36:02", "content": "To me, 1) sounds like a change to the punt formation.\nAs for the 2), how does a more generic \"protect punt receivers during catch\" sound?
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "454472": "This is a breakdown of the proposed rule changes by popularity.\n\n11+ proposals\n* Add new blocking/tackling penalties\n* Enforce changes to punt formations (e.g. number of jammers)\n* Incentivize fair catches\n\n2-10\n* Change how the game is refereed (e.g. add another ref)\n* Eliminate punt returns\n* Incentivize fourth-down \"go for it\" attempts\n* Incentivize out-of-bounds punts\n* Incentivize touchbacks\n* Install a mandatory concussions training program\n* Restrict the actions of punt receivers\n\nUnique\n* Enforce a specific turf type \n* Incentivize field goals from 60+ yards\n* Limit number of pre-season games\n* Limit number of punts per game\n* Move sidelines during punts\n* Protect punt receivers during catch\n\nAlso, at least one kernel tries to list all possible rule changes:\nhttps://www.kaggle.com/ericfreeman/evaluation-of-proposed-punt-rules\n\n=====\n\nOriginal Post:\nDoes someone already have the stats? If not, I would like to collect them over the weekend.", "454689": "That'd be awesome!\n\nI've read through 15 or so, and it looks like there were a lot of really awesome analyses and beautiful visualizations. However, most led back to fairly consistent and simple suggestions.\n\nSeems like a good proportion of proposed changes included some kind of fair catch incentive, most proposing a 5-10 yard bonus. Mine included some qualification criteria regarding punt LOS having to be < kicking teams 45 yard line to keep the balance of the game/ability to pin a team deep in tact. Looks like there are also a few gunner/coverage/formation rules which I need to re-read to better understand how they address correlation/causation regarding those formations and how they relate to concussion rates. There are also some interesting proposals suggesting harsher penalties for blindside blocks, etc. I'm pretty sure a few people suggested banning punts. \n\nIt'll be interesting to see how this pans out. There is one analysis I've seen that will almost surely make the top 4, the rest seem up for grabs. This competition is made more interesting by this article about the possibility of replacing the kickoff with a 4th and long scenario: https://slate.com/culture/2018/12/nfl-kickoff-rule-change-4th-and-15-play.html\nReally cool idea.", "454794": "About correlation-causation of formations, causation is almost always elusive... And here even correlation is tricky, given the small number of events!\n\nI screened exhaustively for Information Value and looks like punt return formations do matter a lot. That said, the question \"why does it matter?\" proved harder to answer. Of lots of possible descriptors of punt return formations (Symmetry, density of different sections) only one showed predictive power, weak but worth considering in my opinion.", "454960": "OK, I've done the work. If someone's rule proposal is missing, please, don't hesitate to inform us.", "455000": "Thanks for pulling the stats. I remember one proposal for implementing the college rule for punts (i.e, all players, not just gunners can release immediately). Here's an article about the concept. \nhttps://www.usatoday.com/story/sports/ncaaf/2014/11/20/rugby-punts-bounding-through-college-football/19311797/", "455001": "miguel perez Was that the number of jammers? I struggled with correlation and on the last day changed my presentation to go with a fair catch incentive instead of jammer restrictions.", "455027": "To me, that sounds like a formation change.", "455038": "EricFreeman, for -penalized- Information Value strongest feature was not number jammers, but asymmetry of central \"box\" considering the two lines of defense, first 3 x2 players first line and 2 players second line.\n(This asymmetry is sure correlated with number of jammers though, as \"missing\" central players have to go somewhere, but strongest predictor was central box asymmetry.)\n\nBiggest insight though was that initial formations matter. I didn't recommend enforcing limitations, though, not before more testing. But I did suggest to share with coaches that defense formations possibly are more than tactics, also related to risk. And, of course monitor the pattern real test time... only then applying a rule change that favours symmetry/compactness, to some extent.", "455042": "That's fair.", "455312": "My two suggestions were to:\n1) Wait Rule: Not allow gunners to leave the line of scrimmage until the punt was in the air and kicked\n2) Barrier Rule: Give the punt returner a 3-yard circular barrier on the punt catch, if a player from the Punting Team enters that barrier before the punt is caught, call the player dead where the punt is caught or where the ball hits the ground.", "455332": "I want to make sure I understand your barrier rule. I'm reading it as a reward for the punting team and as a way to get awarded a tackle without contact. It's almost an extension of the in the grasp rule for QB's except \"in the grasp\" is defined as within 3 yards. Is that right?", "455342": "Yes exactly. Trying to prevent those concussions at the punt catch area and after the catch situations during a run, which we know many of the concussions occur.", "455463": "Very interesting @miguelpm , thanks for sharing. Initial formations are so heavily entangled with extraneous variables like field position, proficiency of the return team, score, etc. - it doesn't surprise me that establishing causation between formation and concussions would be a slippery proposition. I didn't dig into it heavily, but I'm glad I wasn't missing something obvious.", "455589": "\"Entangled\" is a perfect word for describing it @ColinMc , your intuitions are good, average riskier formations plots showed that field position mattered. But also \"normalized\" plots, disconnected from field position were different, so structure of formation matters too...! About score, I found no relationship (though it also was a good candidate for me).\n\nSo even if there was something there, big difficulty was to find a quantifiable parameter, of many many candidates... with the time we had I only found the mentioned symmetry one but there could be more in that \"entanglement\" of features. Glad you found it interesting :-)", "455814": "To me, 1) sounds like a change to the punt formation.\nAs for the 2), how does a more generic \"protect punt receivers during catch\" sound?" }, "source": "meta" }