The Swordfish And The Finned Swordfish: A Simple Explanation

Simon discusses the Swordfish and Finned Swordfish Sudoku pattern and explains how, with the help of Hodoku software, they could help solve last Friday's Diabolical Sudoku from The Daily Telegraph.

[@robertborcherding9085]

I understand the logic as you are explaining it; however, applying that logic in real situations, or even recognizing a situation is quite a different story.

[@eugenetswong]

I agree that it is easier to understand than find, but today I discovered what it takes. Every chain or structure can join to another chain or structure. In many cases, like the fin, it just adds to the complexity, but if we just focus on the cells that the structures see, then it makes sense.

[@user-zf9gg7do3c]

Wow... You are the first person that explained this so that I could fully grasp the concept. Thank you so much. I have been able to solve these tough puzzles consistently now. You have earned a new subscriber today!

[@GJvv2009]

Took me several times listening to understand. Thank you. Well done.

[@KoyasuNoBara]

Thank you, I found another video where you mentioned swordfish, and since I was unfamiliar with the technique, I was having trouble following the logic. Hopefully I can understand that other video now!

[@marykosanke5431]

Great explanation. I understand the logic, but I have never found one! I will have to see if my app allows me to hi-lite the numbers like that. I don't think I would ever notice a finned swordfish, although I understood you explanation.

[@n8likesmath]

Sweet explanation, I think the people bashing in the comments below don't understand swordfish of x wings in the first place. Thankfully I've spent enough time on your videos to understand normal swordfish, and this was an excellent explanation of the finned version. Thank you

[@darcash1738]

I mean sure it makes sense, but what’s the actual approach for these? I don’t want to have to try it for each number, so is there a way to get a feel for which number I should be trying to look for a swordfish for? If not, there’s got to be a quick way to consistently process through a swordfish I notice that 3 candidate rows must match each other perfectly in columns to work, and 3s (as I will now call them) must match to both candidate columns of 2s. From there if you do it systematically, you can either eliminate all possible swordfishes with 3s or come across a valid one. Then you are left with just the 2s, of which you must have at least one alignment for different 2s to be compatible-and of course if two alignments then that is simply and X wing. This systematic approach sounds like it would be efficient even though I haven’t put it into practice yet. But this wouldn’t work exactly with doing the fin. Then it seems like you need to make an allowance for 4 columns which would sort of mess up that system. How would you integrate the fin into this system, or do you have a more effective way of doing it? Because in this case he actually had two 2s, and a 3 that was singly matched with both of these 2s. Is this a common way that finned ones come up? Are there other common ways that fins show up?

[@n8likesmath]

@@darcash1738 look for numbers that are more restricted abd have columns of rows of 2-3

[@Proflaxis]

Wow, this looks like something I can use in my solves. However, I am generally solving NYT hard sudokus and the tools are a bit clumsy to use and there's no color that's available so kinda hard to detect. But I am gonna try. Thank you!!

[@LearnSciSnippets]

Best description and finally understand now. Thanks

[@PhilBoswell]

Throwback to when 17 minutes was a long video 🤯

[@thomaskivi]

I would rather use empty rectangle here in an actual solve, also on the same 7’s: if r4c4 is a 7, then r6c6 cannot be a 7. But if the 7 in that box is not in r4c4, it is either of the positions along c6 in that box. That starts a chain along the 7’s, which ends in box 4, causing r6c2 to be a 7, meaning that again r6c6 has to be a 1. Thanks for explaining the finned swordfish fundamentally! It’s not really much more complicated than a finned x-wing in the end. Just harder to spot.

[@MohamedMahmoud-ey9tj]

I have been using this idea without even knowing it has a name

[@ianwebster995]

Interesting stuff. Very clear and helpful. Hope there’s no weever fish.

[@monoelmono9476]

Nice explanation. My worst nightmare are the swordfishes. I am just hopeless at spotting them

[@monoelmono9476]

@Sam De Roeck I find XY chains pretty easy! I love chains. Swordfishes, Jellyfishes and finned varieties of them will stump me!!

[@alsecen5674]

It would be helpful if your examples were more realistic. The x wing example - Do they all have to be identical pairs? Can they have more than two numbers in each cell? Can I eliminate numbers in the row/column?

[@user-gn5fq1zu9f]

Deep deep deep! Thanks for great video.

[@johndoe-ow2ns]

This was excellent. Thanks 👍

[@TrinityEcho]

KZbin algorithm is fascinating. It showed me this video today. Was able to use what I learned in a NYT puzzle several hours later...

[@The_Cali_Dude_88]

I learned a lot of 'techniques' with some other 'soduko' learning guides which are and still is helpful... however I like your approach as pro-solvers as a bases of time and discovery of these 'methodologies' that can be utilized on the hardest puzzles at the higjest degree so I watch and attempt your styles to get better.. and faster... I make mistakes but learn each time... the NY hard is one you steered me towards... like that with one exception... don't close out the tab... 😁

[@ericwallhagen3146]

Don't you also have a finned swordfish in Rows 2, 6, and 7, with the "fin" being r6c2? Following this would also force R6C6 into a 1. This was the one I spotted, to the same conclusion.

[@wossaaaat]

I suck with this kind of logic, but I thiiiiink that arrangement would eliminate the 7 from R4C3 (which in turn would eliminate it from R6C6, but through pointing rather than the wing). Not sure if that's what you meant anyway, but just in case, I don't think the fin on R6C2 would eliminate the 7 from R6C6 as part of the actual swordfish logic in itself, because either R6C2 is the 7 and R6C6 is eliminated, or it's not the 7 and you have the swordfish, but R6C6 is now a part of that swordfish and _can't_ be eliminated.

[@andres3725]

Great stuff 🤙💯

[@nihatbekiroglu8004]

I don't understand how the pattern of 4s at 7:36 is a valid swordfish ; if the 4 in R2C4 turns out to be the real one, it eliminates the 4s in R2C7 and R8C4, but then what if 4 goes into R5C6 ? It eliminates the 4s in R8C6 and R5C7, and so at the end the two 4s in C7 are gone...Where do you place 4 in C7 ? Can you please clarify this ?

[@Gsudi]

The 4 in R8C4 would be still available. You must not put both a 4 in R2C4 and R5C6. With this pattern you can eliminate all the other 4s in C4, C6 and C7, which aren't in the example. So if you can put a 4 in only those 6 positions you will have a useless swordfish, like the useless X-Wings later in the video. That's what I understood.

[@vernonharmon4007]

A swordfish does two things: first it allows you to eliminate possibilities in the opposite direction it was created (if it's a swordfish in columns, you eliminate in the rows, or if it's in rows, you eliminate in columns), but because of sudoku it also means that if you solve any of the positions for the number in the swordfish, you can immediately eliminate at least 2 other candidates -- along the row and column -- which will result in either an x-wing or a solvable angle configuration like the r8c6/r5c6/r5c7 arrangement you're talking about. In that configuration, r5c6 CANNOT contain the 4, because as you noted it eliminates all valid possibilities. This is why sparser swordfish are more powerful.

[@TheDmntdmnky]

I'm still hella confused 😅

[@jimbak478]

What is the app that you use for this puzzle? I can’t find one that shows candidates

[@fernandomaano2703]

beyond the league of my understanding

[@chloemaxwell2628]

There is a swordfish with 3's as well

[@djgulston]

10:55 Didn't you miss a swordfish for the 3's? Look at rows 1, 4 and 6. Or does the extra 3 in row 4 on the far right disqualify it from being a swordfish?

[@slashclaw14]

That’s not a swordfish because yep the 3 candidate in r4c8 is not confined within c2, c3, and c4 like the other cells. If 3 wasn’t a candidate in that cell, then there would be one

[@djgulston]

@@slashclaw14 Ah, alright! Perfect! Thank you! I've been trying to wrap my head around this. I just discovered swordfish today from his recent video.

[@kevinruggles206]

I am unclear as to how you use the knowledge that the cell you identified (6,6) is affected with either 7 or no 7 in the other cell(4,4) because it would be different numbers (in 6,6) so different logic would follow. I very much enjoy these videos where you explain techniques without solving an entire puzzle. Could you put together a play list of such videos? Previously I have seen a list of all the logic types around 8 techniques I think. It would be fantastic if you could do a video on them. It did not include swordfish or variants or jellyfish or worms. The 2 latter are other techniques I would like to see demonstrated.

[@wokkawicca]

You're unclear on why 6,6 cannot contain a 7? Obviously if 4,4 *is* 7, 6,6 is ruled out by sharing the same box. If 4,4 is *not* 7, then there is a valid swordfish in rows 2/4/7. It's an incomplete 222 fish instead of the full 333 type which would also include 2,6; 4,8; and 7,3, but the logic still holds: no other 7 candidates allowed in columns 3, 6, or 8 apart from those rows. This applies to 6,6 once again, so either way, 6,6 cannot be 7, and you can eliminate it as a candidate in that cell. (apologies if you were asking a different question and I'm simply restating the points from the video) Beyond that, you simply have to move on with whatever advantage that gains you. I am realizing there are many 'trivial' examples of these patterns where it's pointless to find them because they don't eliminate anything. For instance, there is a second finned swordfish on 7s in rows 2,6, and 7, using the same columns, with 6,2 as the 'fin'. It isn't at all useful, because there are no other 7 candidates to eliminate in box 4 where the fin is found. (The finned fish only operates on cells both the fin and the possible fish can 'see'--ie, in the same box as the fin. It doesn't affect, say, 1,2.)

[@billdwyer4129]

Great. Thanks.

[@h0ctApmoingay]

Can someone pls explain how when there's no 7 in R4C4, then the 7 in R6C6 is eliminated please. Thank you.

[@Has39.bfpo43]

I think he is saying that if there was no 7 in R4C4 then the it would be a ‘normal’ swordfish and all the green 7’s in C3, C6 & C8 would be eliminated.

[@h0ctApmoingay]

thank you for your reply@@Has39.bfpo43 , now I'm gonna try and make sense of it and watch this video again, haha, I'm just so slow at this

[@djgulston]

11:44 Could rows 1, 4 and 6 also equate to a finned swordfish?

[@mrsulaman9901]

Ding, ding, ding, you are absolutely correct sir. Well done.

[@groslou9997]

Much easier to see

[@imransaifi2399]

If you eliminate 7 from R4C4 somehow, there will be naked single 7 left in R8C4 which dismantle the swordfish pattern. Please clarify

[@CrackingTheCryptic]

The key is that eliminating the 7 from r4c4 is actually not straightforward at all. The beauty of the Finned Swordfish technique is that we're able to say "whether or not there is a 7 is r4c4, I don't care - because I know that, either way, I can logically eliminate the possibility of a 7 in r6c6".

[@imransaifi2399]

Thanks for reply

[@roccoxxxx1]

I think the unfinned fish are more rare, but at the same time recognizing that there is actually finned fish is really hard

[@eliseocid7457]

👏👏👏👏

[@altcommand]

I am no better at spotting these.

[@The_Cali_Dude_88]

where can you get the same software?

[@CrackingTheCryptic]

Just look up 'Hodoku'

[@The_Cali_Dude_88]

thx!

[@The_Cali_Dude_88]

well next to be safe I see and realize your using but when the site says 'not secure' I get iffy... are you comfortable with the suggestion with a cautionary yes?

[@CrackingTheCryptic]

This is where I downloaded Hodoku from... I'm not IT-competent enough to guarantee the software but it hasn't broken my computer yet :) hodoku.sourceforge.net/en/index.php

[@richardcranium0]

@@The_Cali_Dude_88 Here's the link to the SSL-encrypted site: sourceforge.net/projects/hodoku/

[@groslou9997]

We can see in the Middle square we have the 368 triplet

[@aok76_]

But but.. I found another swordfish with the eights. Rows 5, 8 and 9. I was waiting so eagerly to see if I was right. :(((

[@OzymandiasSaysHi]

@Amr Okasha Unfortunately that is not a swordfish. All the possible entries in the 3 rows have to lineup in EXACTLY 3 columns. Your example has 4 columns occupied.

[@alanclarke4646]

Nope, I didn't see the swordfish, even after staring at for several minutes.

[@kevinruggles206]

worms == a squirm bag

[@hansleber6090]

Häääääää?

[@Morphineck]

Worst swordfish explanation ever.