Methodology
We get data about the videos we track using the YouTube API.
We check for new videos every 10 minutes.
We check like counts, dislike counts, and view counts on a schedule that varies by video age, as described by the table below.
To calculate the real dislikes on a video, we tabulate only increases to the dislikes, and ignore decreases. On normal YouTube videos (e.g., on PewDiePie's videos), the official YouTube stats and our real stats completely agree. On many White House videos, there is a huge discrepancy between official dislikes and our calculated real dislikes.
Open-source data: Our data is freely available for download at the bottom of every page in three convenient formats. You can download it and use it in your own reports, charts, and graphics. If you want to validate our data, you can simply monitor the likes and dislikes we report, live, to the video's actual YouTube page. You could also validate our data by using the YouTube API. If there are any data scientists out there, please contact us if you would like us to link to your project that evaluates the accuracy of our data.
Video age Check frequency
<1 day every 10 minutes
<3 days every 20 minutes
<7 days every 30 minutes
<14 days every 6 hours
<28 days every 12 hours
How does the site obtain the actual upvote/downvote info?
Methodology We get data about the videos we track using the YouTube API. We check for new videos every 10 minutes. We check like counts, dislike counts, and view counts on a schedule that varies by video age, as described by the table below. To calculate the real dislikes on a video, we tabulate only increases to the dislikes, and ignore decreases. On normal YouTube videos (e.g., on PewDiePie's videos), the official YouTube stats and our real stats completely agree. On many White House videos, there is a huge discrepancy between official dislikes and our calculated real dislikes. Open-source data: Our data is freely available for download at the bottom of every page in three convenient formats. You can download it and use it in your own reports, charts, and graphics. If you want to validate our data, you can simply monitor the likes and dislikes we report, live, to the video's actual YouTube page. You could also validate our data by using the YouTube API. If there are any data scientists out there, please contact us if you would like us to link to your project that evaluates the accuracy of our data. Video age Check frequency <1 day every 10 minutes <3 days every 20 minutes <7 days every 30 minutes <14 days every 6 hours <28 days every 12 hours