Fishing with lines?

[Zabby]

No matter how much time i tribute for trying to understand this technique, i never get.

If you are able to draw a line for each missing digit either horizontally or vertically in such a way that no two lines cross any of the remaining candidates, you are allowed to remove all digits not covered by a line.

Draw a line for each missing digit? Does it mean i have to draw a line for each raw and column that doesn't have a specific digit yet? If so, then lines need to cross at every cell with a candidate for the digit, so this technique is just drawing lines without any reason Anyways, i can see in the example that Ruud (i think it was him who wrote the guide) DIDN'T draw a line for EVERY raw and column that has cells with candidate 9. So i think that number of lines to be equal to the number of missing 9s, am I right? If so, there's only one 9 place on the grid and 8 remaining, and there are exactly 9 lines drawn. So i think I know the secret. Now, if I'm right, (but i'm not sure) then how do i know where to draw them? Or does is have to be at least 1 way to place the lines that they don't cross each other at the cells with candidates? If so, won't it require a lot of trying to place them and isn't it then even easier to scan the grid and look for X-Wings, Swordfishes and other raw/column subset techniques? Or maybe it's not too hard to find a way to draw the lines?

What i really would need is practice, yet, I don't really have puzzles which i could use to practice, so I'm soooo confused that I don't wanna even read the solving guide further until i master fishing.

So, any information, tips, explanation?

[Sudtyro]

I have similar difficulties with most methods used to find fish patterns. However, not long ago I ran across Arcilla’s article in the other forum entitled, "a new (?) view of fish (naked or hidden)" http://www.sudoku.com/boards/viewtopic.php?t=5017. His technique is easy to apply and seems to find both regular and certain other finned and “almost” fish patterns by using simple row/column listings.

Ruud’s Jellyfish example in the Sudoku Solving Guide has the following unresolved 9’s grid:

Code: Select all

9 . . | . . 9 | . . .
. . 9 | . . 9 | . . .
. . . | . . . | . . .
------+-------+------
. . . | 9 . . | 9 . .
9 9 9 | . 9 . | 9 . .
. . 9 | . 9 . | . . .
------+-------+------
. . . | 9 . . | . 9 .
9 . 9 | . 9 . | . . .
. 9 9 | . . . | . 9 .

Arcilla’s two lists for the above grid would appear as:

Col numbers, per row: (16)(36)(-)(47)(12357)(35)(48)(135)(238)
Row numbers, per col: (158)(59)(25689)(47)(568)(12)(45)(79)(-)

As pointed out in this forum by Myth Jellies (7/4/07), Arcilla’s technique converts fish into either hidden or naked sets, which are then relatively easy to spot. For example, the above list of column numbers (per row) reveals the naked “quad”, (16)(36)(35)(135), which is simply the (row) Jellyfish in rows 1,2,6 and 8, respectively.

Similarly, the above list of row numbers (per column) reveals the naked “quad”, (59)(47)(45)(79), which is the complementary (column) Jellyfish in columns 2,4,7 and 8, respectively.

It appears that each of the above lists contains both a naked Jellyfish and its complementary hidden Jellyfish. One can reveal the hidden complement by eliminating the naked “candidates” from the other “cells” in the list, as is routinely done with naked/hidden subsets in a regular Sudoku house. The eliminated “candidates” are the same as those that would normally be eliminated by the Jellyfish.