https://www.ft.com/content/bfb7e6b8-d57b-11e1-af40-00144feabdc0\n > \nIn 1906, the great statistician Francis Galton observed a competition to guess the weight of an ox at a country fair. Eight hundred people entered. Galton, being the kind of man he was, ran statistical tests on the numbers. He discovered that the average guess (1,197lb) was extremely close to the actual weight (1,198lb) of the ox. This story was told by James Surowiecki, in his entertaining book The Wisdom of Crowds.
\n\nWe all are believing there is going to be shake-up. lets say it follows some distribution. what do you think will be
\n\n> Mean shake up ( absolute number of places )
\n\nif you care to guess and answer i will compile the answer and compare after the competition. let's see how it goes.
\n\nPS: please don't influence your guess by someone else's guess. (ie. don't read the comments before commenting)
\n\nUPDATE:
\n\nhere is the analysis\nhttps://www.kaggle.com/mks2192/shake-up
", "messages": [ { "id": 540307, "postDate": "2019-05-31T10:44:10.053Z", "content": "https://www.ft.com/content/bfb7e6b8-d57b-11e1-af40-00144feabdc0\n > \nIn 1906, the great statistician Francis Galton observed a competition to guess the weight of an ox at a country fair. Eight hundred people entered. Galton, being the kind of man he was, ran statistical tests on the numbers. He discovered that the average guess (1,197lb) was extremely close to the actual weight (1,198lb) of the ox. This story was told by James Surowiecki, in his entertaining book The Wisdom of Crowds.
\n\nWe all are believing there is going to be shake-up. lets say it follows some distribution. what do you think will be
\n\n> Mean shake up ( absolute number of places )
\n\nif you care to guess and answer i will compile the answer and compare after the competition. let's see how it goes.
\n\nPS: please don't influence your guess by someone else's guess. (ie. don't read the comments before commenting)
\n\nUPDATE:
\n\nhere is the analysis\nhttps://www.kaggle.com/mks2192/shake-up
", "rawMarkdown": "https://www.ft.com/content/bfb7e6b8-d57b-11e1-af40-00144feabdc0\n > \nIn 1906, the great statistician Francis Galton observed a competition to guess the weight of an ox at a country fair. Eight hundred people entered. Galton, being the kind of man he was, ran statistical tests on the numbers. He discovered that the average guess (1,197lb) was extremely close to the actual weight (1,198lb) of the ox. This story was told by James Surowiecki, in his entertaining book The Wisdom of Crowds.\n\n\n\nWe all are believing there is going to be shake-up. lets say it follows some distribution. what do you think will be\n\n > Mean shake up ( absolute number of places )\n\nif you care to guess and answer i will compile the answer and compare after the competition. let's see how it goes.\n\nPS: please don't influence your guess by someone else's guess. (ie. don't read the comments before commenting)\n\n\n\n\n**UPDATE:**\n\nhere is the analysis\nhttps://www.kaggle.com/mks2192/shake-up", "votes": 6 }, { "id": 540390, "postDate": "2019-05-31T12:52:01.210Z", "content": "For those who do not know about Scott E. Page's Diversity Prediction Theorem
\n\nCrowd's square error = the mean square error - the diversity of the crowd
210
", "rawMarkdown": "210", "votes": 1 }, { "id": 541104, "postDate": "2019-06-01T18:16:23.873Z", "content": "280
", "rawMarkdown": "280", "votes": 1 }, { "id": 540946, "postDate": "2019-06-01T12:15:42.027Z", "content": "230
", "rawMarkdown": "230", "votes": 1 }, { "id": 540921, "postDate": "2019-06-01T11:22:56.753Z", "content": "246
", "rawMarkdown": "246", "votes": 1 }, { "id": 540914, "postDate": "2019-06-01T11:15:31.523Z", "content": "350 (with gaussian distribution)
", "rawMarkdown": "350 (with gaussian distribution)", "votes": 1 }, { "id": 540680, "postDate": "2019-05-31T20:48:40.457Z", "content": "In my case? Approximately 4112.
\n\nIn general? I think around 500.
", "rawMarkdown": "In my case? Approximately 4112.\n\nIn general? I think around 500.", "votes": 1 }, { "id": 540643, "postDate": "2019-05-31T19:36:18.650Z", "content": "100
", "rawMarkdown": "100", "votes": 1 }, { "id": 540589, "postDate": "2019-05-31T17:40:21.963Z", "content": "I think major quake will be closer to the top only - about 800-1200 positions, where scores are tight. Probably quite a few blended overfitted models will go down severely. Looking down the leaderboard I guess mainly it will not change much.\nMy estimate is around 100 :)
", "rawMarkdown": "I think major quake will be closer to the top only - about 800-1200 positions, where scores are tight. Probably quite a few blended overfitted models will go down severely. Looking down the leaderboard I guess mainly it will not change much.\nMy estimate is around 100 :)", "votes": 1 }, { "id": 540567, "postDate": "2019-05-31T17:01:05.993Z", "content": "666 ;)
", "rawMarkdown": "666 ;)", "votes": 1 }, { "id": 540544, "postDate": "2019-05-31T16:06:57.817Z", "content": "400
", "rawMarkdown": "400\n", "votes": 1 }, { "id": 540382, "postDate": "2019-05-31T12:39:08.900Z", "content": "Wisdom of crowds is an excellent book. The craziest thing in there is the submarine story in the beginning.
\n\nI think mean shake will be 10, with a few very extreme (+/- 1000) cases.
", "rawMarkdown": "Wisdom of crowds is an excellent book. The craziest thing in there is the submarine story in the beginning.\n\nI think mean shake will be 10, with a few very extreme (+/- 1000) cases.", "votes": 1, "replies": [ { "id": 540383, "postDate": "2019-05-31T12:41:14.497Z", "content": "Thanks for guessing. in that case median can be compared.
", "rawMarkdown": "Thanks for guessing. in that case median can be compared." } ] }, { "id": 540319, "postDate": "2019-05-31T11:01:23.843Z", "content": "Isn't there a database/compilation of past shakeups? And wouldn't the statistics from other competitions be affected by the number of participants N? Based on this competition I think It will be closer to 0, but I'll say 150. I like around numbers.
", "rawMarkdown": "Isn't there a database/compilation of past shakeups? And wouldn't the statistics from other competitions be affected by the number of participants N? Based on [this competition](https://www.kaggle.com/c/sberbank-russian-housing-market/discussion/35522) I think It will be closer to 0, but I'll say 150. I like around numbers.", "votes": 1, "replies": [ { "id": 540320, "postDate": "2019-05-31T11:02:30.477Z", "content": "Not really, thats where exactly wisdom of crowd will come, from your past experience.
", "rawMarkdown": "Not really, thats where exactly wisdom of crowd will come, from your past experience." } ] }, { "id": 540920, "postDate": "2019-06-01T11:22:53.243Z", "content": "400
", "rawMarkdown": "400", "votes": 2 }, { "id": 543755, "postDate": "2019-06-04T19:12:58.833Z", "content": "pretty far off guess\nhttps://www.kaggle.com/mks2192/shake-up
", "rawMarkdown": "pretty far off guess\nhttps://www.kaggle.com/mks2192/shake-up" }, { "id": 542466, "postDate": "2019-06-04T00:14:53.530Z", "content": "hahaha, i was way off
", "rawMarkdown": "hahaha, i was way off" }, { "id": 542463, "postDate": "2019-06-04T00:13:20.953Z", "content": "I was overly optimistic/pessimistic (depends on how you look at it ;) - I couldn't even imagine so many people jumping thousands of spots, not mentioning hundreds...
", "rawMarkdown": "I was overly optimistic/pessimistic (depends on how you look at it ;) - I couldn't even imagine so many people jumping thousands of spots, not mentioning hundreds...\n", "replies": [ { "id": 543667, "postDate": "2019-06-04T17:20:40.447Z", "content": "That because a lot of other people insanely blended public kernels which was obviously overfited.
", "rawMarkdown": "That because a lot of other people insanely blended public kernels which was obviously overfited." } ] }, { "id": 540311, "postDate": "2019-05-31T10:51:17.883Z", "content": "My guess : 45
", "rawMarkdown": "My guess : 45", "replies": [ { "id": 540543, "postDate": "2019-05-31T16:05:06.230Z", "content": "\n\n\nMy guess : 45
\n
no typo here?
", "rawMarkdown": "> My guess : 45\n\nno typo here?", "votes": 2 }, { "id": 540628, "postDate": "2019-05-31T19:06:52.940Z", "content": "hahaha , zero is missing maybe ;).\nI still have a gut feeling your team will be in top 0-10
", "rawMarkdown": "hahaha , zero is missing maybe ;).\nI still have a gut feeling your team will be in top 0-10", "votes": 1 } ] }, { "id": 540310, "postDate": "2019-05-31T10:50:46.907Z", "content": "Subscription to FT is needed to read this. Your credentials might help if you could share ๐
", "rawMarkdown": "Subscription to FT is needed to read this. Your credentials might help if you could share ๐ ", "replies": [ { "id": 540312, "postDate": "2019-05-31T10:51:47.033Z", "content": "i wrote the content. look again
", "rawMarkdown": "i wrote the content. look again" } ] }, { "id": 540917, "postDate": "2019-06-01T11:20:19.960Z", "rawMarkdown": "", "votes": 2, "isDeleted": true } ], "comments": [ { "id": 540390, "author_name": "steubk", "author_url": "", "post_date": "2019-05-31T12:52:01.210000", "content": "For those who do not know about Scott E. Page's Diversity Prediction Theorem
\n\nCrowd's square error = the mean square error - the diversity of the crowd
210
", "votes": 1, "replies": [] }, { "id": 541104, "author_name": "Nickolay Safronov", "author_url": "", "post_date": "2019-06-01T18:16:23.873000", "content": "280
", "votes": 1, "replies": [] }, { "id": 540946, "author_name": "Ilham Firdausi Putra", "author_url": "", "post_date": "2019-06-01T12:15:42.027000", "content": "230
", "votes": 1, "replies": [] }, { "id": 540921, "author_name": "Kha Vo", "author_url": "", "post_date": "2019-06-01T11:22:56.753000", "content": "246
", "votes": 1, "replies": [] }, { "id": 540914, "author_name": "Bellรฒ Filippo", "author_url": "", "post_date": "2019-06-01T11:15:31.523000", "content": "350 (with gaussian distribution)
", "votes": 1, "replies": [] }, { "id": 540680, "author_name": "RNA", "author_url": "", "post_date": "2019-05-31T20:48:40.457000", "content": "In my case? Approximately 4112.
\n\nIn general? I think around 500.
", "votes": 1, "replies": [] }, { "id": 540643, "author_name": "Sergey Zlobin", "author_url": "", "post_date": "2019-05-31T19:36:18.650000", "content": "100
", "votes": 1, "replies": [] }, { "id": 540589, "author_name": "Alexey Kotlik", "author_url": "", "post_date": "2019-05-31T17:40:21.963000", "content": "I think major quake will be closer to the top only - about 800-1200 positions, where scores are tight. Probably quite a few blended overfitted models will go down severely. Looking down the leaderboard I guess mainly it will not change much.\nMy estimate is around 100 :)
", "votes": 1, "replies": [] }, { "id": 540567, "author_name": "Kain", "author_url": "", "post_date": "2019-05-31T17:01:05.993000", "content": "666 ;)
", "votes": 1, "replies": [] }, { "id": 540544, "author_name": "CPMP", "author_url": "", "post_date": "2019-05-31T16:06:57.817000", "content": "400
", "votes": 1, "replies": [] }, { "id": 540382, "author_name": "interneuron", "author_url": "", "post_date": "2019-05-31T12:39:08.900000", "content": "Wisdom of crowds is an excellent book. The craziest thing in there is the submarine story in the beginning.
\n\nI think mean shake will be 10, with a few very extreme (+/- 1000) cases.
", "votes": 1, "replies": [ { "id": 540383, "author_name": "Manoj", "author_url": "", "post_date": "2019-05-31T12:41:14.497000", "content": "Thanks for guessing. in that case median can be compared.
", "votes": 0, "replies": [] } ] }, { "id": 540319, "author_name": "Tim Yee", "author_url": "", "post_date": "2019-05-31T11:01:23.843000", "content": "Isn't there a database/compilation of past shakeups? And wouldn't the statistics from other competitions be affected by the number of participants N? Based on this competition I think It will be closer to 0, but I'll say 150. I like around numbers.
", "votes": 1, "replies": [ { "id": 540320, "author_name": "Manoj", "author_url": "", "post_date": "2019-05-31T11:02:30.477000", "content": "Not really, thats where exactly wisdom of crowd will come, from your past experience.
", "votes": 0, "replies": [] } ] }, { "id": 540920, "author_name": "Felipe Loque", "author_url": "", "post_date": "2019-06-01T11:22:53.243000", "content": "400
", "votes": 2, "replies": [] }, { "id": 543755, "author_name": "Manoj", "author_url": "", "post_date": "2019-06-04T19:12:58.833000", "content": "pretty far off guess\nhttps://www.kaggle.com/mks2192/shake-up
", "votes": 0, "replies": [] }, { "id": 542466, "author_name": "interneuron", "author_url": "", "post_date": "2019-06-04T00:14:53.530000", "content": "hahaha, i was way off
", "votes": 0, "replies": [] }, { "id": 542463, "author_name": "Alexey Kotlik", "author_url": "", "post_date": "2019-06-04T00:13:20.953000", "content": "I was overly optimistic/pessimistic (depends on how you look at it ;) - I couldn't even imagine so many people jumping thousands of spots, not mentioning hundreds...
", "votes": 0, "replies": [ { "id": 543667, "author_name": "Nickolay Safronov", "author_url": "", "post_date": "2019-06-04T17:20:40.447000", "content": "That because a lot of other people insanely blended public kernels which was obviously overfited.
", "votes": 0, "replies": [] } ] }, { "id": 540311, "author_name": "Manoj", "author_url": "", "post_date": "2019-05-31T10:51:17.883000", "content": "My guess : 45
", "votes": 0, "replies": [ { "id": 540543, "author_name": "CPMP", "author_url": "", "post_date": "2019-05-31T16:05:06.230000", "content": "\n\n\nMy guess : 45
\n
no typo here?
", "votes": 2, "replies": [] }, { "id": 540628, "author_name": "Manoj", "author_url": "", "post_date": "2019-05-31T19:06:52.940000", "content": "hahaha , zero is missing maybe ;).\nI still have a gut feeling your team will be in top 0-10
", "votes": 1, "replies": [] } ] }, { "id": 540310, "author_name": "cyberia", "author_url": "", "post_date": "2019-05-31T10:50:46.907000", "content": "Subscription to FT is needed to read this. Your credentials might help if you could share ๐
", "votes": 0, "replies": [ { "id": 540312, "author_name": "Manoj", "author_url": "", "post_date": "2019-05-31T10:51:47.033000", "content": "i wrote the content. look again
", "votes": 0, "replies": [] } ] }, { "id": 540917, "author_name": "", "author_url": "", "post_date": "2019-06-01T11:20:19.960000", "content": "", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "540307": "https://www.ft.com/content/bfb7e6b8-d57b-11e1-af40-00144feabdc0\n > \nIn 1906, the great statistician Francis Galton observed a competition to guess the weight of an ox at a country fair. Eight hundred people entered. Galton, being the kind of man he was, ran statistical tests on the numbers. He discovered that the average guess (1,197lb) was extremely close to the actual weight (1,198lb) of the ox. This story was told by James Surowiecki, in his entertaining book The Wisdom of Crowds.\n\n\n\nWe all are believing there is going to be shake-up. lets say it follows some distribution. what do you think will be\n\n > Mean shake up ( absolute number of places )\n\nif you care to guess and answer i will compile the answer and compare after the competition. let's see how it goes.\n\nPS: please don't influence your guess by someone else's guess. (ie. don't read the comments before commenting)\n\n\n\n\n**UPDATE:**\n\nhere is the analysis\nhttps://www.kaggle.com/mks2192/shake-up", "540390": "For those who do not know about Scott E. Page's Diversity Prediction Theorem\n\n`Crowd's square error = the mean square error - the diversity of the crowd`\n\n[Diversity Prediction Theorem](https://www.coursera.org/lecture/model-thinking/diversity-prediction-theorem-rVbCg)\n", "541654": "210", "541104": "280", "540946": "230", "540921": "246", "540914": "350 (with gaussian distribution)", "540680": "In my case? Approximately 4112.\n\nIn general? I think around 500.", "540643": "100", "540589": "I think major quake will be closer to the top only - about 800-1200 positions, where scores are tight. Probably quite a few blended overfitted models will go down severely. Looking down the leaderboard I guess mainly it will not change much.\nMy estimate is around 100 :)", "540567": "666 ;)", "540544": "400\n", "540382": "Wisdom of crowds is an excellent book. The craziest thing in there is the submarine story in the beginning.\n\nI think mean shake will be 10, with a few very extreme (+/- 1000) cases.", "540319": "Isn't there a database/compilation of past shakeups? And wouldn't the statistics from other competitions be affected by the number of participants N? Based on [this competition](https://www.kaggle.com/c/sberbank-russian-housing-market/discussion/35522) I think It will be closer to 0, but I'll say 150. I like around numbers.", "540920": "400", "543755": "pretty far off guess\nhttps://www.kaggle.com/mks2192/shake-up", "542466": "hahaha, i was way off", "542463": "I was overly optimistic/pessimistic (depends on how you look at it ;) - I couldn't even imagine so many people jumping thousands of spots, not mentioning hundreds...\n", "540311": "My guess : 45", "540310": "Subscription to FT is needed to read this. Your credentials might help if you could share ๐ ", "540917": "" } }