Welcome to the competition. Feel free to ask any questions. If you do not know how to start here are some useful links:
\n\nKaggle is also full of resources. Here is a sample.
\nGood luck!
", "messages": [ { "id": 1671307, "postDate": "2022-02-01T11:01:08.390Z", "content": "Welcome to the competition. Feel free to ask any questions. If you do not know how to start here are some useful links:
\n\nKaggle is also full of resources. Here is a sample.
\nGood luck!
", "rawMarkdown": "Welcome to the competition. Feel free to ask any questions. If you do not know how to start here are some useful links:\n- [Competition submission example](https://www.kaggle.com/octosportio/logistic-regression-submission-example)\n- [Octosport medium blog](https://octosport.medium.com/)\n- [Octosport github](https://github.com/octosport) \n\nKaggle is also full of resources. Here is a sample.\n- [European Soccer Database](https://www.kaggle.com/hugomathien/soccer)\n- [Match Outcome Prediction in Football](https://www.kaggle.com/saife245/football-match-prediction)\n- [Deep Learning Starter : Simple LSTM](https://www.kaggle.com/theoviel/deep-learning-starter-simple-lstm)\n- [Everything you can do with a time series](https://www.kaggle.com/thebrownviking20/everything-you-can-do-with-a-time-series)\n\nGood luck!", "votes": 6 }, { "id": 1684762, "postDate": "2022-02-10T17:38:47.970Z", "content": "Hi there! This competition looks really interesting. I'd love to know if you would want Kaggle team to share your competition on our social media? Also, if you had any issues or other feedback to share with us, I'd love to hear from you -- my email is meg@kaggle.com
", "rawMarkdown": "Hi there! This competition looks really interesting. I'd love to know if you would want Kaggle team to share your competition on our social media? Also, if you had any issues or other feedback to share with us, I'd love to hear from you -- my email is meg@kaggle.com", "votes": 3, "replies": [ { "id": 1685419, "postDate": "2022-02-11T08:57:59.003Z", "content": "Thanks, we have sent you an email.
", "rawMarkdown": "Thanks, we have sent you an email." } ] }, { "id": 1709928, "postDate": "2022-03-02T15:52:23.320Z", "content": "Quick question regarding the ground truth. Is the ground truth a binary value for win/draw/loss based on the real outcome of the game or is there a probability associated with each that serves as the ground truth?
\nAs an example, for a game in the test set where the outcome was a home win, would a prediction of 1,0,0 be the ground truth or is the goal to predict some triple of non-binary probabilities?
\nIf the latter, where do these probabilities come from?
", "rawMarkdown": "Quick question regarding the ground truth. Is the ground truth a binary value for win/draw/loss based on the real outcome of the game or is there a probability associated with each that serves as the ground truth?\n\nAs an example, for a game in the test set where the outcome was a home win, would a prediction of 1,0,0 be the ground truth or is the goal to predict some triple of non-binary probabilities?\n\nIf the latter, where do these probabilities come from?", "replies": [ { "id": 1710072, "postDate": "2022-03-02T18:11:58.283Z", "content": "It is the prediction of 1,0,0 indeed.
", "rawMarkdown": "It is the prediction of 1,0,0 indeed." } ] }, { "id": 1701727, "postDate": "2022-02-23T02:52:04.680Z", "content": "Sounds fun! As a football fan I'll defo give this competition a try. I'm getting started on my EDA right away ;)
", "rawMarkdown": "Sounds fun! As a football fan I'll defo give this competition a try. I'm getting started on my EDA right away ;)" }, { "id": 1697172, "postDate": "2022-02-19T12:26:34.670Z", "content": "hi man this is insanely good it is really interesting to work with football data when you a huge fan of football this make me working harder to exploit my football watching knowledge to stand out
", "rawMarkdown": "hi man this is insanely good it is really interesting to work with football data when you a huge fan of football this make me working harder to exploit my football watching knowledge to stand out ", "replies": [ { "id": 1697343, "postDate": "2022-02-19T14:58:15.210Z", "content": "Thanks, good luck to you, do not hesitate to share 🙏.
", "rawMarkdown": "Thanks, good luck to you, do not hesitate to share 🙏." } ] }, { "id": 1683474, "postDate": "2022-02-09T20:21:41.700Z", "content": "Hi. This competition seems very fun.
\nWould you know a value of log loss that would be good in practical?
\nLet's say that we build a model that can get a log loss of 0.8.
\nThis would be good in practical (e. g. make bets, for instance)?
We put the bookmaker's log-loss as a benchmark in the leaderboard. It is about 0.97 so anything below that would be good for betting.
", "rawMarkdown": "We put the bookmaker's log-loss as a benchmark in the leaderboard. It is about 0.97 so anything below that would be good for betting.", "votes": 1 }, { "id": 1684775, "postDate": "2022-02-10T17:54:31.650Z", "content": "Ok. Thank you
", "rawMarkdown": "Ok. Thank you" } ] }, { "id": 1683575, "postDate": "2022-02-09T22:02:04.967Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 1684762, "author_name": "Meg Risdal", "author_url": "", "post_date": "2022-02-10T17:38:47.970000", "content": "Hi there! This competition looks really interesting. I'd love to know if you would want Kaggle team to share your competition on our social media? Also, if you had any issues or other feedback to share with us, I'd love to hear from you -- my email is meg@kaggle.com
", "votes": 3, "replies": [ { "id": 1685419, "author_name": "octosport.io", "author_url": "", "post_date": "2022-02-11T08:57:59.003000", "content": "Thanks, we have sent you an email.
", "votes": 0, "replies": [] } ] }, { "id": 1709928, "author_name": "rdolphin", "author_url": "", "post_date": "2022-03-02T15:52:23.320000", "content": "Quick question regarding the ground truth. Is the ground truth a binary value for win/draw/loss based on the real outcome of the game or is there a probability associated with each that serves as the ground truth?
\nAs an example, for a game in the test set where the outcome was a home win, would a prediction of 1,0,0 be the ground truth or is the goal to predict some triple of non-binary probabilities?
\nIf the latter, where do these probabilities come from?
", "votes": 0, "replies": [ { "id": 1710072, "author_name": "octosport.io", "author_url": "", "post_date": "2022-03-02T18:11:58.283000", "content": "It is the prediction of 1,0,0 indeed.
", "votes": 0, "replies": [] } ] }, { "id": 1701727, "author_name": "Youssef Taoudi", "author_url": "", "post_date": "2022-02-23T02:52:04.680000", "content": "Sounds fun! As a football fan I'll defo give this competition a try. I'm getting started on my EDA right away ;)
", "votes": 0, "replies": [] }, { "id": 1697172, "author_name": "Ahmed Mohamed11", "author_url": "", "post_date": "2022-02-19T12:26:34.670000", "content": "hi man this is insanely good it is really interesting to work with football data when you a huge fan of football this make me working harder to exploit my football watching knowledge to stand out
", "votes": 0, "replies": [ { "id": 1697343, "author_name": "octosport.io", "author_url": "", "post_date": "2022-02-19T14:58:15.210000", "content": "Thanks, good luck to you, do not hesitate to share 🙏.
", "votes": 0, "replies": [] } ] }, { "id": 1683474, "author_name": "Igor Kuivjogi Fernandes", "author_url": "", "post_date": "2022-02-09T20:21:41.700000", "content": "Hi. This competition seems very fun.
\nWould you know a value of log loss that would be good in practical?
\nLet's say that we build a model that can get a log loss of 0.8.
\nThis would be good in practical (e. g. make bets, for instance)?
We put the bookmaker's log-loss as a benchmark in the leaderboard. It is about 0.97 so anything below that would be good for betting.
", "votes": 1, "replies": [] }, { "id": 1684775, "author_name": "Igor Kuivjogi Fernandes", "author_url": "", "post_date": "2022-02-10T17:54:31.650000", "content": "Ok. Thank you
", "votes": 0, "replies": [] } ] }, { "id": 1683575, "author_name": "", "author_url": "", "post_date": "2022-02-09T22:02:04.967000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1671307": "Welcome to the competition. Feel free to ask any questions. If you do not know how to start here are some useful links:\n- [Competition submission example](https://www.kaggle.com/octosportio/logistic-regression-submission-example)\n- [Octosport medium blog](https://octosport.medium.com/)\n- [Octosport github](https://github.com/octosport) \n\nKaggle is also full of resources. Here is a sample.\n- [European Soccer Database](https://www.kaggle.com/hugomathien/soccer)\n- [Match Outcome Prediction in Football](https://www.kaggle.com/saife245/football-match-prediction)\n- [Deep Learning Starter : Simple LSTM](https://www.kaggle.com/theoviel/deep-learning-starter-simple-lstm)\n- [Everything you can do with a time series](https://www.kaggle.com/thebrownviking20/everything-you-can-do-with-a-time-series)\n\nGood luck!", "1684762": "Hi there! This competition looks really interesting. I'd love to know if you would want Kaggle team to share your competition on our social media? Also, if you had any issues or other feedback to share with us, I'd love to hear from you -- my email is meg@kaggle.com", "1709928": "Quick question regarding the ground truth. Is the ground truth a binary value for win/draw/loss based on the real outcome of the game or is there a probability associated with each that serves as the ground truth?\n\nAs an example, for a game in the test set where the outcome was a home win, would a prediction of 1,0,0 be the ground truth or is the goal to predict some triple of non-binary probabilities?\n\nIf the latter, where do these probabilities come from?", "1701727": "Sounds fun! As a football fan I'll defo give this competition a try. I'm getting started on my EDA right away ;)", "1697172": "hi man this is insanely good it is really interesting to work with football data when you a huge fan of football this make me working harder to exploit my football watching knowledge to stand out ", "1683474": "Hi. This competition seems very fun. \n\nWould you know a value of log loss that would be good in practical?\n\nLet's say that we build a model that can get a log loss of 0.8. \nThis would be good in practical (e. g. make bets, for instance)? ", "1683575": "" } }