潮回り · 물때 · ヒラマサ

Does the spring tide really catch fish — I checked the arithmetic against 378 days of records

2026-08-07 · 2026-08-08 updated · fishingmap.fun

Let me write the conclusion first. Whether or not fish are caught was hardly decided by the tide at all. If anything, the core of the spring tide is the lowest. The only thing that was decided was whether or not a big one came out.

This did not turn into a story that "the spring tide has no meaning", though. When I finished counting and showed it to the angler himself, a completely different reading came back. I write about that in the second half.

This is not statistics. It is only a recount of the records of one angler who has been fishing for 13 years. I lay out the sample size and all the biases. Even so, I think it is worth reading for anyone who builds a schedule believing that "if it is the spring tide, you will catch".

What kind of fishing these records are

Let me write down the scale and the character first, because the numbers alone do not carry the meaning.

There were 161 lure casting trips, 86% of them expeditions, and 378 actual fishing days in total.

A little over 7 out of 10 destinations were Nagasaki and Okinawa (Iki, Tsushima, Goto (Fukue Island) / Iriomote, the main island). Besides those, Hokkaido, Aomori, Korea, Malaysia and the Torres Strait.

It is fishing where I fly to a remote island and have three days on which I put a rod out. Travel days are not included in those three days. It is not fishing where I put a rod out every week nearby; one expedition carries a flight, a boat, lodging and three actual fishing days on it — so which tide the schedule is aimed at is directly whether the fishing succeeds or fails. What follows is the result of accumulating that.

Let me sort out the number of years first. I had been writing with this mixed up, so I am fixing it.

"13 years" is correct only for the records. Casting is 10 years, and the tide tables below are shorter still: the 9th year counted from Iki in April 2017.

Why I counted again

In the course of building This month's tides on this site, I ended up digging up all of my own records. The fishing notes that came from Facebook amounted to 168 trips over 13 years. Cross-checking against Instagram, trips that were not there came out as well. Another 40 from the calendar. A catch record is a thing that only survives for the days something was caught.

Now that I finally had a denominator that included the blanks, I checked the arithmetic on a long-standing question. Is "go around the spring tide and you will catch" true?

Check 1: whether a catch happens — it does not split by tide

I split the 116 sasoidashi expeditions into five bands by distance from Sari (사리), the spring-tide peak (new/full moon — the new moon or the full moon). For multi-day expeditions I judge by the day on which that window came closest to Sari.

From 58.5% to 69.2%. And the core (①) is at the bottom. Far from being an anticlimax, it was the opposite of intuition. An expedition has three actual fishing days, so one fish comes somewhere. "I did not catch because the tide was bad" does not hold, at least not in my records.

Instead, it splits clearly by species. Splitting the same 116 trips into hiramasa (yellowtail amberjack) and GT, hiramasa is 53 out of 75 (70.7%) and GT is 24 out of 41 (58.5%). Against a band-by-band spread of 58.5 to 69.2%, the difference between species is 12.2 points. More than the tide band, what I went out to target is what splits the catch rate.

I know the reason myself. Hiramasa I go after with some degree of aim, whereas GT is fishing where catching on 1 trip out of 2 is a good result. Even within the same "casting expedition", the design of the expected value is different.

I also looked at where the blank expeditions fell. There are 17 of them in the core ±1 day as well. The day at Iriomote when the leader was cut six times in a row, and the day at Tsushima when I dropped three fish in a row, both had the best possible tide. Even aimed at the core, when you cannot land them you cannot land them. What lost me those days was not the tide but knots and hooksets.

What the angler himself said on seeing the numbers — "the number of times" and katashio (one-way tide)

When I had written this far and showed it to him, he was surprised, but his view did not change. And the reason that came back is what I think is the most important thing in this article.

I believe in the spring tide. What I believe in is that the chances become more numerous. The number of times the tide turns over may be greater than in other tide phases.

On a spring tide where the rise and fall come out clearly, there are four chances in a day. That is the part I like.

That said, katashio itself is not bad. There are areas, like the western waters of Tsushima, where katashio is the normal state. There it is normal, and fish are caught that way.

This is not a claim that the probability of catching goes up. It is a claim that the number of at-bats goes up. The rise and the fall each come out clearly twice in a day. On three actual fishing days you can step through 12 tide changes — that is how he is counting.

Read that way, it does not contradict the flat table above. As long as you are looking at "does even one fish come in three days", whether there are 12 at-bats or 6, the answer is usually "yes". It is hitting the ceiling. What moves is not the average but what is in it when it does hit.

And one more thing: that he did not write katashio down as a "bad tide" is what I think matters. There are waters, like the western side of Tsushima, where it is the normal state. Good and bad in tides is decided only together with the body of water — put that way, my own method of lining every area up in a single table starts to look the cruder of the two.

Where the upside came out. That is the next check.

Check 2: only fish over 20kg split by tide

From the same 116 trips I pull out only the expeditions on which a fish over 20kg came. Here the view changed completely.

⚠ 2026-08-09, I rewrote this section. Until then I decided the band by "expanding the expedition

across its days and taking the day with the smallest Sari distance among them".

That means I was deciding the band not by the day the fish came, but by whichever of the three days was closest to the spring tide.

Structurally that pulls things toward the core.

The photo EXIF told me the day of the catch itself, so I replaced it. The conclusion changed.

For hiramasa (Iki, Tsushima, Goto, Genkai-nada, Jeju, Geomundo), fish over 20kg came on 3 expeditions, 6 fish. The core (Sari ±1 day) accounts for 2 of the 3 expeditions.

The 2 fish at Jeju came from entering the same shoal on two consecutive days, and are counted as one expedition (in the records they appeared as two entries, October 12 and October 16, but those were posting dates).

Exposure to the core is 35.3% overall. The probability that 2 of 3 expeditions land there by chance is 28.6%. That is within the margin of error. I can no longer say that "the upside clusters at the core".

The shape does remain, though. Both Jeju and Goto came within the few days building up toward Sari (Jeju was the first 2 days of a 5-day run going 2.5 → 1.5 → 0.5 → 0.0 → 1.0). The way of building a schedule aimed at the spring tide is unchanged; what changed is the conclusion that "it only comes on the day of the core itself".

For GT (Iriomote, Okinawa), fish over 20kg came on 5 expeditions, 5 fish. The day of the catch is confirmed for all 5 expeditions, and the core (Sari ±1 day) is 0 times.

All 5 expeditions miss the core (0/5). The probability of missing by that much or more by chance is 11%. It is within the margin of error, but it is certainly lined up in the opposite direction from the hiramasa side.

2026-03-01 was "March 2" in the records, but yusu's check established March 1, 15:26 (out of Hateruma).

It is the third date correction, after Jeju and Goto.

I cannot go as far as "GT misses the core". All I can say is that no evidence has come out that the core of the tide is working for GT.

On the GT side, the day of the catch is confirmed for all 5 expeditions, through photos and yusu's check.

"That will be because for hiramasa he picks the spring tide to go", you would think. I thought so too. The share of expeditions falling within Sari ±2 days is 45.3% for hiramasa and 41.5% for GT. Almost the same, and it is not that hiramasa alone is pushed toward the core. The same person goes with the same way of building a schedule, and only the results split.

Why the difference appears (my own reading)

Hiramasa is fishing where you put the main current onto an offshore shoal. The tide hits the shoal, a boil forms, baitfish gather, and big fish hold there. Whether that structure stands up is decided by the volume of the tide itself. The 20kg class only opens its mouth on a day when that structure has stood up completely. The feeling that "the more times it turns over, the more chances there are" is, I think, a statement about how that structure stands up.

On the GT side there is less that can be said clearly. Here, please do not decide that GT is "a reef fish". It is not that GT can only be caught on the reef. It is only that the places I go to happen to be skewed toward waters with many reefs and channels. Because Yaeyama and Okinawa have that kind of topography, GT fishing there becomes "fishing by topography and time of day" — that is all it is, and it is not an explanation of the nature of the fish.

On that premise, all I can say is that in the places I go it does not ride on the day-scale tide phase. In fact, in my records the big GT came from a channel 2m deep and from dead shallow water in the neap period. In Yaeyama on a spring tide there are days when the current is too fast to hold the boat. If the place changes, the story should change.

What ought to be written honestly — confounding with the tackle

This conclusion has a weak point. All the hiramasa over 20kg are from 2023 onward. They exist only after the year I changed reels. Through the old-tackle era I kept aiming at the core and had zero fish over 20kg.

So whether they came "because it was the core of the tide" or "because the tackle began to reach", these records cannot separate the two. The big fish are concentrated only on trips where both were in place — that is the accurate way to put it. I wrote about the tackle in the article on line retrieve per crank.

There is another hole that no way of counting will fill. This sample size is built from what remains in social media and the fishing notes. But there is fishing that cannot be left in the records. Tuna is the main example: there are a number of expeditions I did not post about for reasons of circumstance. There are also places like Kyotango where I went but there is not a single entry in the archive.

The bias of "only the days something was caught survive" can be filled in by digging up the blanks, but this one does not exist as a whole expedition, so there is nothing to dig up. There is fishing that was never in this article's sample size to begin with — please read it with that in mind.

And I have not counted "how many tide changes there were on the day". The records do not distinguish a day where the rise and fall came out clearly from a day that ended as katashio. It can be judged after the fact from a tide table, but that is a different thing from "how it actually came out in that body of water". In waters where katashio is the normal state, like the western side of Tsushima, the meaning of that judgment changes in the first place. From next season I will put it into the items I record.

I should add that I am talking about n=4 and n=5. I never intended to call this statistics.

So how to use this

• If you are after a record-class hiramasa, hang the window on the core, Sari ±1 day. The catch rate does not go up. What goes up is the expected value of the size
• What you are aiming at is "a day when the rise and fall come out clearly". Rather than the spring tide itself, the meaning is in being able to step through four tide changes in a day. Taking three days is also to earn that number of times
• Do not dislike katashio. There are waters, like the western side of Tsushima, where it is the normal state and it works. Judge the tide together with the body of water
• For GT, do not pick the day from a tide table. At least not in the waters I go to. Pick by tide level, time of day and topography
• First, stop saying "I did not catch because of the tide". Most of what lost me my days was knots, hooksets and lure rotation

Which three days of this month the Sari window falls on is put out automatically every month by This month's tides. This article is also an explanation of where that table came from.

---

The figures in this article come from the site owner's personal fishing records (2013–2026, 280 consolidated records), except that the sasoidashi population of 116 expeditions used for the tide tables runs from April 2017 to March 2026 (casting itself begins with the Torres Strait in December 2015). The period is output by 【期間】 in verify_stats.py. There is no statistical significance. Every figure given can be recalculated with data/author/fishing-profile/verify_stats.py, and the same script lists the fish over 20kg from the data. A multi-day expedition is counted as one case for the whole expedition and judged by the maximum weight on that expedition. The day counts are actual fishing days and do not include travel days. Fish landed by companions are excluded from the tallies (the (他者) marker in the catch field). The tides are counted in the Korean south-coast way (the 8물때 system), which shifts by ±1 day depending on the school. The framework itself does not apply to shallow ground, to waters where the tide runs too hard, or to the diurnal tides of Southeast Asia. Expeditions to Malaysia, Thailand, Taiwan, Vietnam and India (the Andamans), and tuna casting in places such as the Tsugaru Strait, are excluded from the sample size because they are not sasoidashi over the shoals (they are included in the 161 casting trips).

The two entries 2020-05-26 and 2022-07-01 come only from Instagram records, and there is no corresponding record on the fishing database side.

In August 2026 I corrected dates on the basis of the original posts (2 for Goto, 1 for Nakanokamishima). This includes places where a record previously treated as "one day's catch" was in fact the total over three days. In the same month I removed 7 entries where a companion's catch had been mixed into my own records and 52 duplicates against Instagram, and recalculated every figure. The tallies are generated automatically from the vault with vault_sync.py. Also in the same month I excluded from the sample size a single Langkawi entry that had slipped in through a difference in area naming, and 4 tuna-targeting expeditions (122→117).

2026-08-09, I corrected the date of Goto, March 2024, using the photo EXIF (capture time and GPS). The original post said "March 22-24", but in the photos 3/22 is on the land side of Fukue, and the actual fishing was 3/23, 3/24 and 3/25. On 3/25 there are burst shots offshore at 07:05 and 07:26–07:30, matching the HIT of 07:16 in the records. On the basis of yusu's point that posts are written afterwards and so the dates drift, from here on the photos are taken as correct. With this the sample size became 117→116 expeditions and the related figures were recalculated (the conclusion that Band 1 is at the bottom, and the conclusion that all 4 expeditions with fish over 20kg were Band 1, have not changed).

2026-08-08 (2), I corrected the wording "accumulated over 13 years". What is 13 years is the records as a whole; casting is 10 years from December 2015, and the population for the tide tables is from April 2017.

2026-08-08, on the site owner's own point I corrected three things. ① that the day count is 3 actual fishing days and does not include travel days ② the place where I had written GT as "a reef fish" (the places I go are simply skewed toward reefs; it is not an explanation of the nature of the fish) ③ that the owner's own view on the spring tide was added to the body text.

2026-08-08, added. I made the owner's view concrete down to "four tide changes in a day" and "katashio is the normal state in some waters". The previous version wrote katashio as uniformly bad, and that was wrong.

TideThis month's tide — read in Korean 물때 and a 3-day window→