Why a clip can top the day and miss the week
The daily board and the weekly board are not the same measurement at two sizes. They ask different questions of different fields, so a clip can honestly win one and never appear on the other.
Three boards, three separate contests
Community-ranked clips look like one ranking viewed at three zoom levels. They are not. The daily board asks what people watched on one day, judged only against the other clips from that day. The weekly board asks what held attention across a field roughly seven times larger, one that already contains every daily winner. The Hall of Fame asks what people still come back to long after the day it was posted. Winning the first tells you very little about the second, and almost nothing about the third.
The daily field is small, and that is deliberate
A clip on the daily board only had to beat one day. That is a feature: it means a good play from a quiet Tuesday gets a place instead of being buried under a loud weekend. It also means the daily board is the easiest of the three to reach, and being the top clip on a Tuesday makes you one of seven candidates for the week, not a lock for it. Most daily winners do not make the weekly board, and nothing has gone wrong when that happens.
Attention decays, and not at the same rate for every clip
Two clips can finish a day with almost identical engagement and then separate completely. One was a moment: watched, enjoyed, finished. The other kept being shared, re-posted and found by people who missed it, so it was still collecting attention on day three. The daily board reads both at the same age, twenty four hours old, where they look the same. The weekly board reads them at different points on their curve, where they do not. That is the thing the daily board structurally cannot see, because it stops looking at midnight. Durability is a separate property from impact, and only the longer windows measure it.
Each window corrects one distortion and adds another
The daily board is unfair to clips posted at quiet hours, which is covered in detail here. The weekly board partly fixes that by giving a clip more time to be found. In exchange it introduces the opposite unfairness: a clip posted on a Monday has most of the window ahead of it, and one posted late Saturday is judged on a fraction of its life against clips that have been accumulating for six days. The daily board treats those two fairly, because each is scored against its own day. So neither window is the honest one. They are honest about different things, and reading them together is what gets you close to the truth.
Why a clip sometimes surfaces later than the day it was posted
Nothing leaves. The archive stays browsable by date, the categories keep collecting plays by what happens in them rather than by when they landed, and every featured creator has a page holding the clips of theirs that ranked. A clip that finished third on its day and then kept being watched can end up on the weekly board without ever having led a daily one. That is not a correction applied by hand, and no clip is promoted because somebody liked it. It is the same engagement signal read over a longer window, where slow growth is finally visible.
What this changes if you make clips
Read the daily board as a snapshot, not a verdict. If a clip placed and then disappeared, the useful question is not why it was demoted, because it was not: it was scored again against a different field. Check the archive for the date it landed and the play is still there, still credited, exactly where it finished.
The practical difference between a clip that wins a day and one that wins a week is usually rewatchability. A fail that lands hard is funny once. A clutch that reveals something the second time through, a build sequence with a decision in it, a moment where the interesting part is what nearly happened: those keep earning attention after the first watch, and the longer windows are the only place that shows up as a number.
How to read all three at once
Take the daily list as what a lot of people found worth watching today. Take the week as what survived being watched more than once. Take the Hall of Fame as survivorship, with the bias that implies: it favours plays that were legible to a large audience at the time, and it is the slowest of the three boards to admit anything new. Three windows on the same signal is not the same as three opinions, and none of the three is a skill leaderboard.