{
  "id": 38818,
  "title": "Leaderboard is fixed",
  "url": "/competitions/carvana-image-masking-challenge/discussion/38818",
  "author_name": "",
  "post_date": "2017-08-31T18:25:27.248222800Z",
  "votes": 26,
  "comment_count": 18,
  "views": 0,
  "content": "<p>The (limited precision) leaderboard should now be working correctly. \"Correctly\" means that scores are being truncated, not rounded, and that ties are broken out to full precision. Thanks for your patience while we patched the bug and let us know if you see any strange behavior.</p>",
  "messages": [
    {
      "id": "217712",
      "postDate": "08/31/2017 18:25:27",
      "content": "<p>The (limited precision) leaderboard should now be working correctly. \"Correctly\" means that scores are being truncated, not rounded, and that ties are broken out to full precision. Thanks for your patience while we patched the bug and let us know if you see any strange behavior.</p>",
      "rawMarkdown": "The (limited precision) leaderboard should now be working correctly. \"Correctly\" means that scores are being truncated, not rounded, and that ties are broken out to full precision. Thanks for your patience while we patched the bug and let us know if you see any strange behavior.",
      "votes": null
    },
    {
      "id": "217717",
      "postDate": "08/31/2017 18:31:46",
      "content": "<p>Wow, what a shakeup..  Hoping full precision is coming soon.</p>",
      "rawMarkdown": "Wow, what a shakeup..  Hoping full precision is coming soon.",
      "votes": null
    },
    {
      "id": "217720",
      "postDate": "08/31/2017 18:37:19",
      "content": "<p>Wow, yes! Great news - both in terms of having the new LB, and our position on it. :D</p>",
      "rawMarkdown": "Wow, yes! Great news - both in terms of having the new LB, and our position on it. :D",
      "votes": null
    },
    {
      "id": "217723",
      "postDate": "08/31/2017 18:48:00",
      "content": "<p>So people that previously had 0,997 in reality has more than 0,9965, but less than 0,997 <br>\nI think another precision digit is needed...</p>",
      "rawMarkdown": "So people that previously had 0,997 in reality has more than 0,9965, but less than 0,997   \nI think another precision digit is needed...",
      "votes": null
    },
    {
      "id": "217727",
      "postDate": "08/31/2017 18:57:44",
      "content": "<p>@William Cukierski, Could you tell us any good reason to keep insisting on not adding additional digit?</p>",
      "rawMarkdown": "William Cukierski, Could you tell us any good reason to keep insisting on not adding additional digit?",
      "votes": null
    },
    {
      "id": "217729",
      "postDate": "08/31/2017 19:00:14",
      "content": "<p>Thanks for the great efforts. However, my previous submissions are still in old scores. They are not reflected this change yet. (Saying my best score in LB is corrected to 0.996 from 0.997 but that one displayed in \"my submunitions\" is still 0.997).</p>",
      "rawMarkdown": "Thanks for the great efforts. However, my previous submissions are still in old scores. They are not reflected this change yet. (Saying my best score in LB is corrected to 0.996 from 0.997 but that one displayed in \"my submunitions\" is still 0.997).",
      "votes": null
    },
    {
      "id": "217734",
      "postDate": "08/31/2017 19:18:21",
      "content": "<p>Same Here !!</p>",
      "rawMarkdown": "Same Here !!",
      "votes": null
    },
    {
      "id": "217735",
      "postDate": "08/31/2017 19:18:59",
      "content": "<p>yes, and it would be good to know if the \"my submissions\" rank order are now based on full precision or only truncated.</p>",
      "rawMarkdown": "yes, and it would be good to know if the \"my submissions\" rank order are now based on full precision or only truncated.",
      "votes": null
    },
    {
      "id": "217736",
      "postDate": "08/31/2017 19:19:22",
      "content": "<p>We agree and have increased to 4.</p>",
      "rawMarkdown": "We agree and have increased to 4.",
      "votes": null
    },
    {
      "id": "217737",
      "postDate": "08/31/2017 19:19:54",
      "content": "<p>Thanks, will investigate. Hang tight.</p>",
      "rawMarkdown": "Thanks, will investigate. Hang tight.",
      "votes": null
    },
    {
      "id": "217755",
      "postDate": "08/31/2017 20:09:58",
      "content": "<p>Haha, and now Kyle and bestfitting are first (again like in the Planet Challenge). Hopefully I can beat you this time! :D</p>",
      "rawMarkdown": "Haha, and now Kyle and bestfitting are first (again like in the Planet Challenge). Hopefully I can beat you this time! :D",
      "votes": null
    },
    {
      "id": "217766",
      "postDate": "08/31/2017 20:59:36",
      "content": "<p>Another piece of great news. It's been a great day in Carvanaland. :D </p>",
      "rawMarkdown": "Another piece of great news. It's been a great day in Carvanaland. :D",
      "votes": null
    },
    {
      "id": "217771",
      "postDate": "08/31/2017 21:14:07",
      "content": "<p>It seems to be fixed and updated to 4 digits right now. Thanks!</p>",
      "rawMarkdown": "It seems to be fixed and updated to 4 digits right now. Thanks!",
      "votes": null
    },
    {
      "id": "217781",
      "postDate": "08/31/2017 22:35:21",
      "content": "<p>thanks!!</p>",
      "rawMarkdown": "thanks!!",
      "votes": null
    },
    {
      "id": "217851",
      "postDate": "09/01/2017 08:08:22",
      "content": "<p>thanks</p>",
      "rawMarkdown": "thanks",
      "votes": null
    },
    {
      "id": "217863",
      "postDate": "09/01/2017 08:38:25",
      "content": "<p>thanks, but why not display 5 significant figures? The top LB teams still have the same score.</p>",
      "rawMarkdown": "thanks, but why not display 5 significant figures? The top LB teams still have the same score.",
      "votes": null
    },
    {
      "id": "219311",
      "postDate": "09/07/2017 17:47:49",
      "content": "<p>@William Cukierski.</p>\n\n<p>I just made a submission and the my LB score shows 0.9957 even though your best entry statement is showing 0.9958 i.e.\n\"Your submission scored 0.9958, which is an improvement of your previous score of 0.8220. Great job!\"</p>\n\n<p>Is there still a problem with the LB?</p>",
      "rawMarkdown": "William Cukierski.\n\nI just made a submission and the my LB score shows 0.9957 even though your best entry statement is showing 0.9958 i.e.\n\"Your submission scored 0.9958, which is an improvement of your previous score of 0.8220. Great job!\"\n\nIs there still a problem with the LB?",
      "votes": null
    },
    {
      "id": "219313",
      "postDate": "09/07/2017 17:56:34",
      "content": "<p>The best entry message was still using the outdated code path with rounding. It should be fixed now.</p>",
      "rawMarkdown": "The best entry message was still using the outdated code path with rounding. It should be fixed now.",
      "votes": null
    },
    {
      "id": "219568",
      "postDate": "09/08/2017 16:21:07",
      "content": "<p>@Williams Thanks.</p>\n\n<p>By the way I also noticed likes of comments in the discussions section does not show under one's profiles for days now. Was the issue you fixed affecting this competition or the whole of Kaggle i.e all running competitions?</p>",
      "rawMarkdown": "Williams Thanks.\n\nBy the way I also noticed likes of comments in the discussions section does not show under one's profiles for days now. Was the issue you fixed affecting this competition or the whole of Kaggle i.e all running competitions?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 217717,
      "author_name": "heyt0ny",
      "author_url": "",
      "post_date": "08/31/2017 18:31:46",
      "content": "<p>Wow, what a shakeup..  Hoping full precision is coming soon.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 217720,
      "author_name": "tunguz",
      "author_url": "",
      "post_date": "08/31/2017 18:37:19",
      "content": "<p>Wow, yes! Great news - both in terms of having the new LB, and our position on it. :D</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 217723,
      "author_name": "ironbar",
      "author_url": "",
      "post_date": "08/31/2017 18:48:00",
      "content": "<p>So people that previously had 0,997 in reality has more than 0,9965, but less than 0,997 <br>\nI think another precision digit is needed...</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 217727,
      "author_name": "pavelgonchar",
      "author_url": "",
      "post_date": "08/31/2017 18:57:44",
      "content": "<p>@William Cukierski, Could you tell us any good reason to keep insisting on not adding additional digit?</p>",
      "votes": null,
      "replies": [
        {
          "id": 217736,
          "author_name": "wcukierski",
          "author_url": "",
          "post_date": "08/31/2017 19:19:22",
          "content": "<p>We agree and have increased to 4.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217755,
          "author_name": "timjoseph",
          "author_url": "",
          "post_date": "08/31/2017 20:09:58",
          "content": "<p>Haha, and now Kyle and bestfitting are first (again like in the Planet Challenge). Hopefully I can beat you this time! :D</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217766,
          "author_name": "tunguz",
          "author_url": "",
          "post_date": "08/31/2017 20:59:36",
          "content": "<p>Another piece of great news. It's been a great day in Carvanaland. :D </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217781,
          "author_name": "cjansen",
          "author_url": "",
          "post_date": "08/31/2017 22:35:21",
          "content": "<p>thanks!!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 217729,
      "author_name": "cttsai",
      "author_url": "",
      "post_date": "08/31/2017 19:00:14",
      "content": "<p>Thanks for the great efforts. However, my previous submissions are still in old scores. They are not reflected this change yet. (Saying my best score in LB is corrected to 0.996 from 0.997 but that one displayed in \"my submunitions\" is still 0.997).</p>",
      "votes": null,
      "replies": [
        {
          "id": 217734,
          "author_name": "rrqqmm",
          "author_url": "",
          "post_date": "08/31/2017 19:18:21",
          "content": "<p>Same Here !!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217735,
          "author_name": "tivfrvqhs5",
          "author_url": "",
          "post_date": "08/31/2017 19:18:59",
          "content": "<p>yes, and it would be good to know if the \"my submissions\" rank order are now based on full precision or only truncated.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217737,
          "author_name": "wcukierski",
          "author_url": "",
          "post_date": "08/31/2017 19:19:54",
          "content": "<p>Thanks, will investigate. Hang tight.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 217771,
          "author_name": "cttsai",
          "author_url": "",
          "post_date": "08/31/2017 21:14:07",
          "content": "<p>It seems to be fixed and updated to 4 digits right now. Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 217851,
      "author_name": "nsoojin",
      "author_url": "",
      "post_date": "09/01/2017 08:08:22",
      "content": "<p>thanks</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 217863,
      "author_name": "gw00207",
      "author_url": "",
      "post_date": "09/01/2017 08:38:25",
      "content": "<p>thanks, but why not display 5 significant figures? The top LB teams still have the same score.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 219311,
      "author_name": "sheriytm",
      "author_url": "",
      "post_date": "09/07/2017 17:47:49",
      "content": "<p>@William Cukierski.</p>\n\n<p>I just made a submission and the my LB score shows 0.9957 even though your best entry statement is showing 0.9958 i.e.\n\"Your submission scored 0.9958, which is an improvement of your previous score of 0.8220. Great job!\"</p>\n\n<p>Is there still a problem with the LB?</p>",
      "votes": null,
      "replies": [
        {
          "id": 219313,
          "author_name": "wcukierski",
          "author_url": "",
          "post_date": "09/07/2017 17:56:34",
          "content": "<p>The best entry message was still using the outdated code path with rounding. It should be fixed now.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 219568,
          "author_name": "sheriytm",
          "author_url": "",
          "post_date": "09/08/2017 16:21:07",
          "content": "<p>@Williams Thanks.</p>\n\n<p>By the way I also noticed likes of comments in the discussions section does not show under one's profiles for days now. Was the issue you fixed affecting this competition or the whole of Kaggle i.e all running competitions?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "217712": "The (limited precision) leaderboard should now be working correctly. \"Correctly\" means that scores are being truncated, not rounded, and that ties are broken out to full precision. Thanks for your patience while we patched the bug and let us know if you see any strange behavior.",
    "217717": "Wow, what a shakeup..  Hoping full precision is coming soon.",
    "217720": "Wow, yes! Great news - both in terms of having the new LB, and our position on it. :D",
    "217723": "So people that previously had 0,997 in reality has more than 0,9965, but less than 0,997   \nI think another precision digit is needed...",
    "217727": "William Cukierski, Could you tell us any good reason to keep insisting on not adding additional digit?",
    "217729": "Thanks for the great efforts. However, my previous submissions are still in old scores. They are not reflected this change yet. (Saying my best score in LB is corrected to 0.996 from 0.997 but that one displayed in \"my submunitions\" is still 0.997).",
    "217734": "Same Here !!",
    "217735": "yes, and it would be good to know if the \"my submissions\" rank order are now based on full precision or only truncated.",
    "217736": "We agree and have increased to 4.",
    "217737": "Thanks, will investigate. Hang tight.",
    "217755": "Haha, and now Kyle and bestfitting are first (again like in the Planet Challenge). Hopefully I can beat you this time! :D",
    "217766": "Another piece of great news. It's been a great day in Carvanaland. :D",
    "217771": "It seems to be fixed and updated to 4 digits right now. Thanks!",
    "217781": "thanks!!",
    "217851": "thanks",
    "217863": "thanks, but why not display 5 significant figures? The top LB teams still have the same score.",
    "219311": "William Cukierski.\n\nI just made a submission and the my LB score shows 0.9957 even though your best entry statement is showing 0.9958 i.e.\n\"Your submission scored 0.9958, which is an improvement of your previous score of 0.8220. Great job!\"\n\nIs there still a problem with the LB?",
    "219313": "The best entry message was still using the outdated code path with rounding. It should be fixed now.",
    "219568": "Williams Thanks.\n\nBy the way I also noticed likes of comments in the discussions section does not show under one's profiles for days now. Was the issue you fixed affecting this competition or the whole of Kaggle i.e all running competitions?"
  },
  "source": "meta"
}