2019年 · 壱岐 · 対馬

2019 — 71 days offshore, 448 hours. What was left in the records was 34 lines

2026-08-11 · fishingmap.fun

Days out at sea: 71. Hours spent at sea: 448. What was left in the records: 34 lines.

I rebuilt the year of 6 and a half years ago from three things: the photo EXIF, the posts of the time, and the angler's own memory. The dates and times come from the photos. The weights and the lures come from the records and from the angler's own account.

[Note first] The figures in this article are not yet all of them.

There are 66,579 actual files in the photo folder, but only 58,009 are in this index.

8,593 of them (13%) I have not been able to read yet. What is missing is concentrated in August 2017 to April 2019, and

of the photos from January to early April 2019 I have not seen a single one yet.

The day counts and hours below are all for mid-April onward. The real figures will be higher than these.

I wrote how this came about in the record of the AI development.

---

The skeleton of the year

Days out at sea by month

By area

Yaeyama and Okinawa are the most at 21 days. Next is Sagami Bay and Miura at 16 days. By day count, the far expeditions and the close-to-home trips are almost even.

But seen in hours, against Yaeyama's 156 hours, Tokyo Bay is 2 hours. Close to home I break off after a short time. Even for the same "one day", the contents are completely different.

And what was left in the records is 34 lines. In the photos it is 71 days. I have written down less than half. What I did not write down was the close-to-home side.

---

The full record of catches

The tides are worked out from the Japan Meteorological Agency's hourly estimated tide levels (nearest observation station). "Sari (사리)" is the number of days from the new/full moon, and "Band 1" is within ±1 day. "Multtae (물때)" is the Korean way of counting the tide cycle, the one the anglers of Tsushima and Iki use.

The record of zero catches

The days when nothing was caught are exactly the ones worth keeping.

---

Target fish

Besides these, the records also hold 18 GT at Christmas Island (9/13-19, γ250, 10 catches on the first and second days alone), a buri (Japanese amberjack) at Iki (4/24), a big haul of kuromutsu from Manazuru to Hatsushima (1/21), 37cm/33cm at Iioka (1/23), a four-fish hookup in mid-deep water (5/15), ×0 fish in the Tokara Islands (4/15) and Kakeroma (10/27).

This year I was not yet "the hiramasa and GT person". I was catching tuna, rockfish and mid-deep water fish as well.

---

Lures

The Damsel 75 and the BF100 are the mainstays. 9 of all 16 fish came on these two. The numbers are the weight in g, not the length. The BF100 is 100g / 200mm overall.

The KLL028A alone I still cannot identify. KLL is Carpenter's blanket name for the "special, small-batch slot"; it is not a model name. The Hayabusa also came out in this slot at first and became a regular product later. What the lure was that hooked the 8kg at Iki on the morning of April 8, the records leave only a number for.

---

Rods and reels

To be honest, almost none of this year's tackle is left. Of the 34 lines of records, the ones with a rod or a reel written down are only 5 lines.

Only for the heaviest fish of the year, the kihada 25kg of June 23, were both the rod and the reel left.

2019 was still the age of spinning

The shift to baitcasting is from 2022 onward. That is, the 16 fish of this year were all taken on spinning.

The tackle ledger has an entry for the DAIWA Saltiga 5000H saying "did good work at Iriomote and Genkai-nada" and "it suited me well". The areas match the main grounds of 2019, so it is likely that the fish of Iki and Iriomote were taken with it. But the year cannot be pinned down, so I do not state it as fact.

The Fishman BRIST 8.3XXXXH and BC4 8.3XXXH that are the mainstays now have their track record at 155cm at Jeju in 2023 and 34.80kg at Goto (Fukue Island) in 2024, so they have not appeared yet in 2019.

The records of the tackle disappear before the records of the fish.

The date the photos remember. The weight stays in the posts. The lure stays half the time.

The rod and the reel were left in only 5 lines out of 34.

---

Hit times

It is not skewed to the morning mazume. Just as many came in the 10 o'clock and the 11 o'clock hours.

And at Iki this year, the evening was strong. April 7 gave two 9kg fish in a row at 17:18 and 17:52, and March 28 as well gave 12kg at 16:37.

---

Tides

The tide data could be obtained only for the 6 fish from June onward, where the offshore coordinates are known from the photos.

The best fish of the year, the GT of 20kg, came on 조금 (chogeum, the neap tide). It was the exact opposite of the most widely believed thing in the fishing world, "the spring tide catches".

The 10kg hiramasa of March 6 and March 8 were at Sari 0.5 and 1.0, both of them the core of the spring tide. The 12kg of March 28, on the other hand, is at Sari 6.5, 13물, far from the core. On the same boat at the same Iki, double figures came out on the spring tide and on the neap tide alike.

In does the spring tide really catch fish I checked 500 days' worth, and

to give only the conclusion, "the spring tide catches" could not be confirmed even with 1,274 hours as the denominator.

This table of 2019 is the first fragment of that.

---

What was not left in the records

The contents of "×1 fish", understood after 6 and a half years

February 24, Iriomote. The records had only "×1 fish, popper". No species, no weight. The answer is 10:08, GT 10kg, Damsel 75. The Damsel is a popper, so the record was not wrong. It was only not enough.

The GT of 20kg that went missing with its catch column blank

June 10, Iriomote, Nakanokamishima. The records had only "Iriomote, Nakanokamishima, Bluefish, Shimadoraku", and the catch column was empty. In the photos, 12 burst shots over 12 minutes from 11:00 were left. Once the "BF" of BF160 turned out to be the Bluefish, it connected to the record.

The largest GT of this year did not exist in the records for 6 and a half years.

Three days of blanks, and the hole in the photos

November 30 to December 2, Iriomote. Three days, only small fish.

But at 16:00 on December 1, within 1 km south of Kamishima, 3 bites in a row, cut off 3 times in a row. A brand-new PE was cut. The records say "the 2 best bites", but in fact there were 3.

And in that time band, not a single photo is left. I was shooting at 09:50, 10:08 and 11:53, and the next is 17:29. The 16 o'clock hour is a gaping hole.

While losing fish, a person does not hold a camera.

This was a finding for the analysis as well. "Time with no photos = time not fishing" is not true. If anything, the hotter the time, the fewer the photos.

---

The figures of this year

---

The dates and times are from the photo EXIF; the weights, lures and tackle rest on the records of the time and the angler's own account. The tides are calculated from the Japan Meteorological Agency's hourly estimated tide levels (16 points, 2019-2026). For before mid-April 2019 the index is unfinished, so it rests on the records and the memory alone (the photos themselves exist). Fish landed by companions are not included.

This article, the text and the tallies alike, is written by an AI. How it was made, and where it went wrong and was corrected, is in the record of the AI development.

Correction, 2026-08-15 (made before the English translation, by recounting the summary tables from the article's own catch table): (1) In the lure table, the 25 kg kihada of June 23 was missing from the Damsel 75 row (4 times to 5). Following from that, "8 of all 16 fish" becomes 9. (2) In the hit-time table, 10-12 and 16-18 were each one fish short (3 to 4: 10:08 / 10:23 / 11:00 / 11:19, and 16:37 / 17:06 / 17:18 / 17:52). The total goes from 13 to 15 (March 6 has no time left in the record, so 15 of the 16 fish in the catch table). The basis is the catch table in this article itself; the counting was done mechanically, in the same way as src/audit_blog_facts.py.

Two figures are still unsettled. The "hours offshore" column of the area table adds up to 450 hours, 2 hours off the 448 hours in the headline. The one akahata in the target-species table does not appear in the catch table (presumably because no weight or time survives, but this has not been confirmed). Both need the primary record to be checked, so neither has been touched in this correction.

Added 2026-08-15 (confirmed with yusu): the single akahata has been removed from the target-species table. It is not in the catch table, and on checking with the angler it turned out to have been put in without any record behind it. This piece is about how only 34 lines survived in the log, so something with no record cannot stay in the count. The target-species total is now 16, matching the catch table. / 448 hours is the correct figure (confirmed with yusu). The area breakdown adds up to 450 hours, and the source of the 2-hour gap has not been identified. It is written in the table as it stands, unresolved.

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