House of the Dragon, in the finale afterglow
The channel's two biggest Westeros lists of the summer ran at 5.8x and 7.9x. Season three ended on 9 August and the conversation is still arriving, so the post-finale reread window is open for a few weeks.
What was measured, and what it means here
Left, the four scores as computed for this subject. Right, a person's reading of each one for WatchMojo in particular.
- Strongest signal
- Momentum at 49
- Weakest signal
- Opening at 10
- Videos on the subject
- 22, 16 of them owned
- Comments naming it
- 1, 1 unprompted
Thin in comments, which is the caveat below. The demand is in the views: 124k and 168k against a 21k median.
Conversation up 1.6x; upload share down. The channel is under-supplying its own hit.
One competitor, six videos. Open.
Top 30 Game of Thrones Questions Answered in House of the Dragon at 5.8x. Top 20 Game of Thrones Revenge Kills at 7.9x. House of the Dragon Season 3: 10 Differences Between Book and Show at 2.0x.
Scores are 0 to 100 percentile-style ranks among the 262 subjects in this run, never absolute claims: a high demand score means most other subjects drew less conversation, not that most viewers want it. Fit is read from 16 owned uploads on the subject itself. Everything on the left is counted from the corpus; everything on the right was written by a person and cites its evidence further down the page.
When to publish
The call and the measurement side by side, so an override is visible rather than silent.
The season finished on 9 August (eight episodes from 21 June). Comment velocity on the subject is still up 1.6x, which is the finale afterglow; whole-season rereads and open-questions lists are the shape for it, and the window closes within weeks.
The measurement alone would say: Later in the cycle. The author overrode it, for the reason given beside this.
- Momentum
- 49
- of 100, rank in this run
- Recent mentions
- 1
- comments in the last 14 days, 0 in the 60 before
- Late share
- -
- no upload on the subject is old enough to tell
The rule is simple on purpose: hot conversation plus rising uploads reads as now; rising with less heat reads as within days; a subject still drawing comments a month after upload reads as evergreen; the rest waits for its moment in the cycle.
The shape the evidence points to
In the corpus's own vocabulary of ten format lanes, matched on titles.
Numbered 20 to 30, moments and questions. The book-versus-show comparison ran at 2.0x, the one comparison shape on the channel that does.
Index is views over each channel’s own median for the same video type, on uploads at least three days old. An index above one means the format beats the channel’s typical upload; the two bars share a scale.
Where the editorial team would start
These are title directions for the editorial team, not scripts. Each is written in the shape the channel's own record rewards; the writing stays with the writers.
- 01Top 30 House of the Dragon Season 3 Moments Ranked
- 0220 Questions the Season 3 Finale Left Open
- 03Top 20 Targaryen Deaths Ranked by How Deserved They Were
What could make this call wrong
The demand signal is low because the comment sample under these videos was thin, not because the audience is absent. Treat this as a views-led call.
Weakest measured signal: Opening at 10 of 100 among the 262 subjects in this run.
What the call was read from
First the comments and videos the author cites. Then the live record, pulled from the corpus at render time: what WatchMojo has already made on this subject, and what broke out for competitors.
Comments the call cites
0 citedThis call cites no comments; it rests on the video record.
Videos the call cites
3 citedWatchMojo’s own record on this subject
16 owned uploads old enough to index, top 8 by indexCompetitor breakouts on this subject
top 0 by indexNo competitor video on this subject counts as a breakout in this run.
Comments are the top-sorted sample YouTube returns for each video, capped at our own limit, never a video’s full comment count. The index on each video is its views over its own channel’s median for the same type, on uploads at least three days old, which is why a small channel’s breakout ranks above a large channel’s ordinary day. Competitor supply means the 15 competitor channels collected for this run, not YouTube as a whole, and everything here comes from a single collection on 2026-08-27.
4 of the 20 inputs in the brief
The brief lists twenty inputs and asks that they be combined, not presented. These are the ones this call rests on most, with how each was measured in this run.
Every video's views over its own channel's median for the same type, on videos at least three days old.
329 main-channel uploads from 25 May to 27 August, each indexed against the channel median, aggregated by subject, vertical and format.
Ten format lanes matched on titles; median index per format, WatchMojo against the field.
A suggested lane from momentum, velocity and late share, which the authored call may override with reasons.