Hodoku’s Multiple Fin Krakens

This post finds multiple fin kraken swordfish, with additional removals, where Hodoku overlooks kraken analysis. This prompts an examination of the Hodoku kraken fish examples in its Last Resorts section, which offer no guidance for human solving.

Hodoku Finned 6The homework was to discover the two double fin swordfish defined by suset 146/1289 in the second example of Hodoku’s Finned/Sashimi Fish page. This one is a double fin kraken swordfish making additional eliminations. The c8 potentials see both fins. The SW potentials all turn on 8r9c9, turning off both fins. Only r3c2 escapes, and in an interesting way. It turns off the 8 group in c9 leaving only the two fins. One of them must be true. We don’t care which. The fish does not bite, regardless.

Hodoku Finned 5In the second one , we have the fish 146/1289 with two fins in c2. Two krakens are saved by confirming one of the fins. To be removed, a potential must see all of the fin cells. This is what makes multiple finned fish removals so rare.

While potential r9c9 could not be accused of seeing both fins, it did see enough to get into trouble, by stripping the North box of 8-candidates. It’s one of Andrew Stuart’s Unit Forcing Chains (Dec 2012).

The first swordfish was completely resolved, but this one is not. The available finned fish have left r8c9 unresolved. Has the 8-panel raised a question that it cannot answer? If so, we can look on the grid for forcing chains by which this candidate can see any of the toxic fin sets.

I did that, in effect,Hodoku finned jelly tr simply by doing a trial trace on a true 8r8c9. It ends with

8r8c9 => fin 8r6c2

and my simplest inference path looking like this:

Hodoku finned jelly grid

Looking at that, I realized that I could have stopped short by recognizing that 8r8c9 sees 8r3c2. But that fact should have shown up on the 8-panel!

Well, guess what. It does.  Hodoku finned jelly panelAfter the other removals, the 8-panel shows that, if 8r8c9 is true, then one of the two fins must be true also, and that 8r8c9 escapes the school of finned fish.

I didn’t have to show you how I stumbled into that, but it does make my point that the grid just may back you up when the X-panel fails to resolve a finned fish potential victim.

It also reminds us how removals accumulate. If a potential is removed by any fish it stays removed, even though its presence may enable a removal in another fish. However, potentials are subject to double jeopardy. If they escape one fish, they can still be eliminated on the next one.

Hodoku krakwing 1

Hodoku’s Last Resort page gives us a glimpse of the kraken analysis that a solver can do on today’s powerful computers. It demonstrates the use of multivalue forcing chains on the entire grid.

His first kraken fish example, a Franken sashimi 2-wing with two wing cells, is definitely computer solving technology parading as humanly practical technique. Would anyone dream up this Franken finned fish with the idea that a potential victim could see both fins?

Hodoku krakwing 2So here is a sysudokie rendition of the computer printout. You can trace the grouped forcing chains through the ALS nodes from potential victim 2r8c7 to each fin.

Remember that this can be done without elaborate chain building by a trial trace toward the fin targets, as we saw above.

But the most telling fact about this case is revealed on the 2-panel.

Hodoku krakwing 1 safeThe removal claimed by the above result is obvious on the panel.  But not in the way that Hobiger intends. Each potential, if true, removes 2r6c7, and therefore forces one of the fins to be true. The fish fails to remove any for that reason.

So how does 2r6c7 remove itself by seeing both fins via forcing chains? Simply by not being true, and starting no chain. Any potential that sees both fins also removes all 2’s from r6. It becomes a victim of an Andrew Stuart unit forcing chain.

What an elaborately staged and terribly flawed example of kraken fish.

The Hodoku comment on this example also misleading. Hobiger says:

  • “If both fins are false, the fish is true and r1c7, r5c7, and r8c7 can be eliminated
  • If fin r6c2 is true, 2 can be eliminated from r8c7 as proven by the first chain
  • If fin r6c9 is true, 2 can be eliminated from r8c7 as proven by the second chain”

Just remain aware that successful removals by a finned fish do not mean that the fins are true. The first statement means nothing more than if all the fins are discovered later to be false, the fish becomes regular, and all previous free passes to potentials are revoked. The last two statements simply state that if a fin is later found to be true, then candidates that saw it are false. There is no justification in that for Hodoku’s example.

Hodoku Kraken 3Hodoku labels his second kraken fish a Type 2. It puts a two-finned sashimi 4-wing to a new purpose. It’s not about eliminating 4’s!

Both fins see a 7-candidate. If it is true, both fins are false, and the sashimi 4-wing removes 4r3c2. But that makes 4r5c2 true and contradicts the premise. The conclusion is that 7r5c8 is false!

Finally, a case where Hodoku has a reason for a premise. Note the Medusa ANL created by this argument. It’s a new type of toxic set.

Hodoku Kraken 4Type 2 may explain why Hobiger starts chains with the fins, rather than from potentials needing to see them.

But the idea of spending this much human solving effort on such a unlikely prospect is absurd.





In summary, Hodoku totally misses the most productive techniques for finned fishing. Single fin kraken analysis riddles his prime Franken and finned fish examples. X-panel and suset based kraken analysis reveals multiple fin examples that his solver missed. His finned fish techniques section should be revised, and the kraken analysis of the Hodoku Last Resort section, simply taken down.

Next time, we take a revealing second swipe at Hodoku’s suset-cooked Franken fish.


About Sudent

My real name is John Welch. I'm a happily married, retired professor (computer engineering), timeshare traveling, marathon running father of 3 wonderful daughters and granddad to 7 fabulous grandchildren. The blog is about Sudoku solving. It covers how to start, basic solving to find candidates efficiently, and advanced solving methods in an efficient order of battle. It is about human solving methods, not computer solving.
This entry was posted in Advanced Solving, Expert Reviews, Hobiger and tagged , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s