Today I tried to get the amount of papers (roughly) since 2004. But while playing around in Google Scholar I noticed that I get more results when I filter with the "since 2016" button than with the "since 2015" button.
In my understanding, the "since 2015" filter should include everything that was published after 2015, so 2016 as well.
And Google didn't mess up, because the "since 2015" results contain elements that were published in 2016.
Another fun fact: When you enter a custom timeframe and search for articles from 2015-2016, you get 100 results more than searching from 2015-blank.