We parsed every episode-by-episode air-date list in the catalogue. The weekly assumption fails on 51 percent of them.
Forty-six of the 90 do not follow a clean weekly schedule. That is 51 percent. Not a handful of oddities. More than half.
Forty-six of the 90 BL and GL titles in this catalogue with a full episode date list do not follow a clean weekly schedule. That is 51 percent. Everyone knows how a BL drama comes out: one episode a week, same night, until it ends. That is the shape the fandom plans around, the shape every episode guide assumes, and the shape a countdown widget is built on. We went and checked, and it holds for slightly under half the catalogue.
A note on method first, because the number is only worth something if the definition behind it is stated.
Every title record on this site has an episode dates field: one line per episode, each with the date that episode was released. Of 121 title records, 30 carry no list at all and one carries a single date, which leaves no gap to measure. That leaves 90. We computed the number of days between each episode and the next, then sorted every title into exactly one bucket, testing in the order below so that no title is counted twice.
| Bucket | Definition | Titles |
|---|---|---|
| Full dump | Every episode shares a single date. Tested first, because a full dump also satisfies the next definition. | 4 |
| Clean weekly | Every gap is exactly seven days. No exceptions, no skipped weeks, no doubles. | 44 |
| Same-day batch | At least one pair of episodes shares a date, and every gap that is not zero is exactly seven days. | 17 |
| Irregular gap | At least one gap that is neither seven days nor zero. Skipped weeks, split seasons, long hiatuses. | 25 |
| Other | Fits none of the above. Nothing landed here. | 0 |
| Total | The complete analysed set. | 90 |
Those five bucket counts sum to 90, which is the whole analysed set. If you think a full dump should not count as a schedule at all, drop those four and the figure is 42 of 90. It is still not close to the weekly assumption.
This is the most common deviation and the one that silently corrupts every third-party episode list.
Top Form ran eleven episodes across only nine Thursdays. Episodes 1 and 2 both went up on 20 March 2025, and episodes 7 and 8 both went up on 24 April. No week was skipped and the series ran straight through Songkran. But anyone filling in dates weekly from the premiere will have every episode from the second onward on the wrong day. That failure mode bites hardest on a series still in its run, which is what our running list of ongoing BL series keeps track of.
Connecting to You put thirteen episodes on six Fridays: a four-episode premiere on 14 August 2026, then two per week, then a single finale. Double Helix opened with two on the same Friday and then ran a Friday and Saturday pair every week after, so its gaps alternate one day, six days, one day, six days, and never once hit seven.
Korean cable has a shape of its own. Four Hands, Two Sonatas aired Saturday and Sunday every week for six weeks. Love Class Season 3 did the same thing on Thursday and Friday for eight weeks. Both are perfectly regular. Neither ever has a seven-day gap.
This matters because a naive “is it weekly?” check marks these as broken when they are the most disciplined schedules in the set.
Two of the most-watched titles in our catalogue skipped exactly one week, and for the same category of reason.
The Heart Killers ran twelve Wednesdays with a fourteen-day gap between episode 6 on 25 December 2024 and episode 7 on 8 January 2025. That is the Christmas and New Year break. KinnPorsche has a single fourteen-day gap after episode 2, between 9 and 23 April 2022. That is Songkran. The Heart Killers episode dates either side of its Christmas break are listed in full in our cast and episode guide for The Heart Killers.
Worth noting alongside them: Pit Babe did not take that break. Episodes 6 and 7 aired on 22 and 29 December 2023, straight through both holidays. The break is a choice, not a rule.
Some records look catastrophically irregular until you see what the gap is.
Fragrance of the First Flower has a 1,166-day gap in the middle of its episode list: six episodes across five Fridays in November 2021, then nothing until February 2025, when twelve more arrive two per Tuesday. That is two seasons more than three years apart, stored as one run.
GELBOYS has a 504-day gap between episode 7 and episode 8. Either side of it, the schedule is perfectly weekly Saturdays.
The Untamed released fifty episodes across only twenty distinct dates, inside six weeks. A six-episode premiere on Thursday 27 June 2019, then two episodes a day on Monday, Tuesday and Wednesday each week, then an eight-episode finale dump on 7 August.
At the other end of the scale, You Are My Present released all twenty of its episodes on a single day, 8 August 2025. They run about a minute each, made vertically for a phone screen. The whole series is roughly the length of one ordinary drama episode.
And Love Class put its six episodes out on six consecutive days in May 2022. Daily, not weekly.
Our working assumption going in was that streaming platforms schedule irregularly and broadcasters do not. The raw numbers appear to confirm it. Sorting the 89 titles whose network list is filled in, those that list only streaming services deviate 33 times out of 42, or 79 percent, while those carrying at least one linear broadcast channel deviate 13 times out of 47, or 28 percent. A two-sided Fisher exact test on that table returns p = 0.000002.
It is also, we think, not true as stated.
The two groups are not comparable populations. All sixteen of the Chinese, Taiwanese and Hong Kong titles in our set are streaming-only, and all sixteen deviate, because those industries do not use a one-episode-per-week broadcast slot at all. Sort the same 90 titles by country instead.
| Country | Titles | Deviating | Share |
|---|---|---|---|
| Thailand | 51 | 16 | 31% |
| South Korea | 11 | 10 | 91% |
| Japan | 10 | 3 | 30% |
| China | 9 | 9 | 100% |
| Taiwan | 5 | 5 | 100% |
| Hong Kong | 2 | 2 | 100% |
| Vietnam | 1 | 1 | 100% |
| Philippines | 1 | 0 | 0% |
Now control for country. Within Thailand alone, where broadcast and streaming-only delivery genuinely coexist, broadcast titles deviate 9 times out of 37, or 24 percent, and streaming-only titles deviate 7 times out of 13, or 54 percent. The same Fisher test on that table returns p = 0.08. The distance between 0.000002 and 0.08 is the whole point: strip out the country composition and the association stops clearing the usual bar.
We want to be exact about what that does and does not show. Thirteen streaming-only Thai titles is a small group, and p = 0.08 is not evidence of no effect. It is a result that will not carry a claim. There may be a real platform effect inside Thailand that a catalogue this size cannot resolve. What we can say is that the pooled figure is not measuring it.
So the pooled predictor is not the platform. It is the production industry and the format it works in. Thai and Japanese broadcast-slot dramas hold a weekly rhythm because a slot is a slot. Chinese, Taiwanese and Korean web series are built around multi-episode drops from the start, and were never weekly to begin with.
We are stating the thing we did not find as clearly as the thing we did, because the misleading version of this claim is easy to write and we nearly wrote it.
If you want to see the range, these are the records where the schedule is doing something genuinely unusual. Every line below was read off the stored dates, not summarised from a synopsis.
Three practical consequences.
There is a reason this article was possible to write, and it is not that we are especially clever. It is that this site records every episode date individually rather than storing a premiere date and a weekly interval. The same field-by-field record keeping is what lets the drama-by-drama adaptations index name the novel, manga or manhwa each series came from.
The calculated approach is much less work and it is right about 49 percent of the time. It produces a page that looks authoritative and is quietly wrong, and the error compounds: a show with a double drop in episode two has every subsequent date off by a week, and a show with a season gap has them off by years.
Storing dates individually costs more to maintain and occasionally exposes our own gaps. Thirty of our 121 titles do not yet have a date list at all, and one more has a list with a single entry, and we would rather show those as absences than fill them with arithmetic.
Two of our own records store their dates as bare dates without episode numbers rather than in the numbered form the other 89 use, and eight records use a different line ending from the rest of the set. We found all of them while doing this count. None is visible to a reader and all of them parse correctly, but they are the kind of thing that breaks a script six months later, so they are now noted.
Fifty-seven of the 91 dated records contain at least one 2026 date. The catalogue is weighted toward the present, which means the country breakdown above reflects what we have covered as much as what those industries produce. A wider sample might move the Thai and Japanese figures. It is unlikely to move the Chinese, Taiwanese and Hong Kong ones, because those are already at 100 percent and the reason is structural.
The 51 percent headline is a measurement of our own catalogue, stated as such. We would rather publish the number with its limits attached than round it into a claim about the whole genre.
Published 10 September 2026, corrected the same day. The first version of this article reported 47 deviating titles out of 91, with a bucket table of 42, 17, 4, 25 and 1. Those bucket counts summed to 89, not 91, which is how the error was caught. Re-running the count from the stored dates gives 90 analysable titles, because one record holds a single date and therefore no gap to measure, and a table of 44, 17, 4, 25 and 0 that sums correctly. The headline figure is 46 of 90, or 51 percent, not 47 of 91.
The same pass corrected four further figures: the platform split, now 79 percent against 28 percent at p = 0.000002 where the first version said 70 against 32; the within-Thailand control, now p = 0.08 where the first version described the effect as well inside noise; the Thai country row, now 51 titles and 16 deviating rather than 52 and 17; and a claim that one record used an unusual line ending when in fact eight do. Every per-title example in the article was re-checked against the stored dates and all of them stand unchanged.
The direction of the finding did not move. Its precision did, and a number that does not add up is not a number anyone should have been asked to trust.
Bucket counts, gap measurements and the country breakdown were computed from this site’s own stored episode date fields on 10 September 2026, and every figure was independently re-derived before publishing. Episode dates are compiled from official broadcaster and platform sources and cross-referenced against TMDB. Last updated: September 10, 2026.
Of the 90 titles in the BLGLFLIX catalogue that carry two or more stored episode dates, 44 follow a clean weekly schedule where every gap between consecutive episodes is exactly seven days. That is 49 percent. The remaining 46 titles, 51 percent, deviate in at least one measurable way.
We sorted each title into exactly one bucket, testing in a fixed order so nothing is counted twice. Full dump means every episode shares a single date; 4 titles. Same-day batch means at least one pair of episodes shares a date and every non-zero gap is exactly seven days; 17 titles. Irregular gap means at least one gap that is neither seven days nor zero, covering skipped weeks, split seasons and hiatuses; 25 titles. Nothing fell outside those definitions. The five bucket counts sum to 90.
The most common deviation is the silent double drop, where two episodes go up on the same date without the schedule being announced as changed. Top Form ran eleven episodes across only nine Thursdays because episodes 1 and 2 shared 20 March 2025 and episodes 7 and 8 shared 24 April. Any guide that assumes one episode per week from the premiere date drifts out of alignment from that point onward and never recovers.
The pooled association is strong — streaming-only titles deviate 33 times out of 42 against 13 of 47 for titles carrying a linear broadcast channel, Fisher exact p = 0.000002 — but it does not survive a control for country. Within Thai titles alone the same test returns p = 0.08, which will not carry the claim. All sixteen Chinese, Taiwanese and Hong Kong titles in the set are streaming-only and all sixteen deviate, so platform here is largely a proxy for production industry. We publish this as a non-finding because the misleading version is easy to write.
Every title record on BLGLFLIX has an episode dates field holding one line per episode with that episode’s release date. Of 121 title records, 30 have no list, one has a single date, and 90 carry two or more. Dates are compiled from TMDB and official broadcaster and platform sources. The measurement was run over the catalogue as of September 2026 and the bucket definitions are stated in full in the article.
That is a definitional choice, so we give both figures. Four titles release every episode on a single date. If you exclude them as having no schedule to deviate from, the deviation count is 42 of 90 rather than 46 of 90. Either way the weekly assumption fails on close to half the catalogue.
Yes, on the day of publication. The first version reported 47 of 91 with a bucket table that summed to 89 rather than 91. Re-running the count from the stored dates gives 90 analysable titles and 46 deviating. The platform percentages, the within-Thailand control and the Thai country row were corrected in the same pass. A full correction notice sits at the foot of the article.
Metadata and images in this article come from TMDB and official sources, and remain the property of their copyright holders. This product uses the TMDB API but is not endorsed or certified by TMDB. BLGLFLIX is an independent site with no affiliation to the productions covered. Full copyright, correction and takedown policy: Disclaimer.
Discussion 0 comments
Join other fans talking about Half of BL Doesn’t Air Weekly: What 90 Release Schedules Actually Show. Be kind, no spoilers without tags, no spam.
No comments yet.
Share the first reaction below.