It's my first time in a NCAA competition , and I've seen Ken Pom rankings being discussed in multiple discussions and notebooks, including ones in the previous years' competitions. Can someone explain what exactly these rankings are ?
", "messages": [ { "id": 753884, "postDate": "2020-02-22T19:19:43.407Z", "content": "Also please note that our contest data includes a file MMasseyOrdinals, which includes ordinal rankings (e.g. ranked #1, ranked #2, ..., ranked #351) for many seasons, many weeks, and many ranking systems, including the Pomeroy rankings. However, they are only ordinal rankings and do not try to distinguish e.g. how much better the #1 team is than the #2 team. That is why the underlying ratings may be more useful. Nevertheless, the ordinal rankings are already present in the contest data and easily accessed. You can read more about the Massey Ordinals on the contest Data page.
", "rawMarkdown": "Also please note that our contest data includes a file MMasseyOrdinals, which includes ordinal rankings (e.g. ranked #1, ranked #2, ..., ranked #351) for many seasons, many weeks, and many ranking systems, including the Pomeroy rankings. However, they are only ordinal rankings and do not try to distinguish e.g. how much better the #1 team is than the #2 team. That is why the underlying ratings may be more useful. Nevertheless, the ordinal rankings are already present in the contest data and easily accessed. You can read more about the Massey Ordinals on the contest Data page.", "votes": 1 }, { "id": 753760, "postDate": "2020-02-22T16:22:44.567Z", "content": "KenPom is a website with various NCAA hoops ratings and rankings. My opinion is that it isn't a good idea to use them because they are an aggregated number for the entire season including the tournament. This is the data leakage and risky. Apparently a lot of people people did well using them last year. If you buy a subscription, you could scrape some interesting game data however.
\n\nHere's a link to discussion about it in another thread. \nhttps://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/130586#749794
", "rawMarkdown": "KenPom is a website with various NCAA hoops ratings and rankings. My opinion is that it isn't a good idea to use them because they are an aggregated number for the entire season including the tournament. This is the data leakage and risky. Apparently a lot of people people did well using them last year. If you buy a subscription, you could scrape some interesting game data however. \n\nHere's a link to discussion about it in another thread. \nhttps://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/130586#749794", "votes": 1, "replies": [ { "id": 753795, "postDate": "2020-02-22T16:50:19.683Z", "content": "I see. Thanks :)
", "rawMarkdown": "I see. Thanks :)" }, { "id": 754748, "postDate": "2020-02-24T02:15:36.507Z", "rawMarkdown": "", "votes": 1, "isDeleted": true }, { "id": 754757, "postDate": "2020-02-24T02:47:00.433Z", "content": "You might be right. It's a good reminder to double check the rules. I haven't read through them but I'm not using any KenPom myself. The only paid data I think might be useful is the team experience or continuity potion but one could extract it from the play by play files with a little guile. The only reason that data has any value is there is no leakage.
", "rawMarkdown": "You might be right. It's a good reminder to double check the rules. I haven't read through them but I'm not using any KenPom myself. The only paid data I think might be useful is the team experience or continuity potion but one could extract it from the play by play files with a little guile. The only reason that data has any value is there is no leakage. " }, { "id": 754802, "postDate": "2020-02-24T04:31:23.710Z", "content": "There's a public notebook called ken pom scraper or something, does that work?
", "rawMarkdown": "There's a public notebook called ken pom scraper or something, does that work? " }, { "id": 757661, "postDate": "2020-02-27T02:11:17.587Z", "content": "one way to find out...
", "rawMarkdown": "one way to find out..." } ] }, { "id": 753738, "postDate": "2020-02-22T15:58:27.077Z", "content": "It's my first time in a NCAA competition , and I've seen Ken Pom rankings being discussed in multiple discussions and notebooks, including ones in the previous years' competitions. Can someone explain what exactly these rankings are ?
", "rawMarkdown": "It's my first time in a NCAA competition , and I've seen Ken Pom rankings being discussed in multiple discussions and notebooks, including ones in the previous years' competitions. Can someone explain what exactly these rankings are ?", "votes": 2 } ], "comments": [ { "id": 753884, "author_name": "Jeff Sonas", "author_url": "", "post_date": "2020-02-22T19:19:43.407000", "content": "Also please note that our contest data includes a file MMasseyOrdinals, which includes ordinal rankings (e.g. ranked #1, ranked #2, ..., ranked #351) for many seasons, many weeks, and many ranking systems, including the Pomeroy rankings. However, they are only ordinal rankings and do not try to distinguish e.g. how much better the #1 team is than the #2 team. That is why the underlying ratings may be more useful. Nevertheless, the ordinal rankings are already present in the contest data and easily accessed. You can read more about the Massey Ordinals on the contest Data page.
", "votes": 1, "replies": [] }, { "id": 753760, "author_name": "Ben Tesch", "author_url": "", "post_date": "2020-02-22T16:22:44.567000", "content": "KenPom is a website with various NCAA hoops ratings and rankings. My opinion is that it isn't a good idea to use them because they are an aggregated number for the entire season including the tournament. This is the data leakage and risky. Apparently a lot of people people did well using them last year. If you buy a subscription, you could scrape some interesting game data however.
\n\nHere's a link to discussion about it in another thread. \nhttps://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/130586#749794
", "votes": 1, "replies": [ { "id": 753795, "author_name": "Satwik", "author_url": "", "post_date": "2020-02-22T16:50:19.683000", "content": "I see. Thanks :)
", "votes": 0, "replies": [] }, { "id": 754748, "author_name": "", "author_url": "", "post_date": "2020-02-24T02:15:36.507000", "content": "", "votes": 1, "replies": [] }, { "id": 754757, "author_name": "Ben Tesch", "author_url": "", "post_date": "2020-02-24T02:47:00.433000", "content": "You might be right. It's a good reminder to double check the rules. I haven't read through them but I'm not using any KenPom myself. The only paid data I think might be useful is the team experience or continuity potion but one could extract it from the play by play files with a little guile. The only reason that data has any value is there is no leakage.
", "votes": 0, "replies": [] }, { "id": 754802, "author_name": "Satwik", "author_url": "", "post_date": "2020-02-24T04:31:23.710000", "content": "There's a public notebook called ken pom scraper or something, does that work?
", "votes": 0, "replies": [] }, { "id": 757661, "author_name": "S Charlesworth", "author_url": "", "post_date": "2020-02-27T02:11:17.587000", "content": "one way to find out...
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "753884": "Also please note that our contest data includes a file MMasseyOrdinals, which includes ordinal rankings (e.g. ranked #1, ranked #2, ..., ranked #351) for many seasons, many weeks, and many ranking systems, including the Pomeroy rankings. However, they are only ordinal rankings and do not try to distinguish e.g. how much better the #1 team is than the #2 team. That is why the underlying ratings may be more useful. Nevertheless, the ordinal rankings are already present in the contest data and easily accessed. You can read more about the Massey Ordinals on the contest Data page.", "753760": "KenPom is a website with various NCAA hoops ratings and rankings. My opinion is that it isn't a good idea to use them because they are an aggregated number for the entire season including the tournament. This is the data leakage and risky. Apparently a lot of people people did well using them last year. If you buy a subscription, you could scrape some interesting game data however. \n\nHere's a link to discussion about it in another thread. \nhttps://www.kaggle.com/c/google-cloud-ncaa-march-madness-2020-division-1-mens-tournament/discussion/130586#749794", "753738": "It's my first time in a NCAA competition , and I've seen Ken Pom rankings being discussed in multiple discussions and notebooks, including ones in the previous years' competitions. Can someone explain what exactly these rankings are ?" } }