Puzzle Description:

#1: Tom O'Connell (sensei69) on Jun 5, 2014 [HINT]

Check out the '3'

?

After initial LL and elementary EL

C14 5 has to go r3-7
IEL Dots r3, C5

Line logic to finish

Fun solve, Tom.

yeah, the 5 has to go up top, but you can't tell that in 2 moves ahead or less. hate to say it, but if that's your only way of solving this, it's "deep-lookahead". no guessing.

it's plenty easy enough to "see" the correct solution, but I'm not finding a moderate-lookahead way to get there. I'm not saying that there isn't one, just that I'm not seeing it.

Can't recreate it, but the 5 down creates conflicts with the 3rd row over to the left (i believe). I didn't start charting until much too late. Joe will come along and clean it up.

it makes a contradiction with the 3 in c16 if you were to put the 5 on the bottom. (3rd column to the right from the 5). Agreed. I didn't dispute that. but it's more than 2 steps look-ahead.

LL
C1-2 & C20, r7-10 placed
EL r18 dots C14-16
EL 3 C16 dots C17
EL5 R18 Dots R4-6
EL 5 R18 dots C4-6
LL
EL C14 Dots r 17
LL to finish

with ll between:
el 5 r18 =c4-6,c14-16w
el 3 r17 =c4-6w
el 2 r17 =c14-16w
el 6 c20 =r14-17w

At this point, we could use the fact that this has a unique solution to place the 5 in c14, if we were the sort to do that. (Hint: it involves the blots in r6 & r11.)

Otherwise,
del3c16(using the 5 in c14)=r3w
el3c16=r14w

Even now the 5 being in the bottom doesn't conflict with 3 in c16, forcing it into r5-7 (the blot in r6 being forced to be a 2). (Unless you go deep lookahead with the far right side. However, we could do a less deep deep-lookahead anyway.)

At this point, I ran out of moderate lookahead logic. Here's the least deep lookahead I see:

If we place c14's 5 in r11-15, then the 2 in r14 and the 2 in c15 will force the 3 in r13 to cross the 5. That forces the 1 in r13 into the far right side. However, the fact that there is only one clue left in r11 will make the far right of that row white, and thus all under that must be white, leading to a conflict.

[Alternatively, with the 5 in r11-15, the single clue in r11 makes the far right white in that row, forcing the 6 in c20 all the way up into r5. That will take up the rightmost clues in r5-6, making the 3 in c16 not fit within r4-7 due to edge logic. Meanwhile, back in r12, the 2 won't let the 3 fit in r11-13, either.]


So the 5 in c14 must be in r3-7. Then LL to finish.

Found to be solvable with deep lookahead by infrapinklizzard.

Whatever wins, however I think you can logically place the 5. I disagree.

Please give your specific steps.

To the finishing LLS steps...

EL places R1 in 8-13 and R2 in 8-``
Limits R18 to C6-13, Black 10,11
LL
Dot C20, R14
Dot C15, R14
The 4 in C5 Cannot go R3
LL to finish

Hope I notated correctly...

Joe, I never guess, and realy don't understand smile logic (think it is guessing). I solved it 3 times. The 5 solves out without guessing. The 3 in R13 is forced into the clue forcing the 5 up as it places and leaves a 1,,,

well, at least, I kept y'all busy :D

it was a fun solve, Tom! and a fun challenge to keep looking for alternate ways of solving :)

didn't realize it was a competition sgusa ;)...you can logically place the 5, just with 3 steps instead of 2. anyway, you've got your Old Style? you said you'd argue "whatever" as long as you had some....I'm leaving work and will stop for some Bud on the way home ;) (like my punnyness?)

lol thx whatever

I KNOW Tom O'Connell knows that I sometimes solve under an alt id, but does Tom King have any clue as to who I am? a challenging puzz got published today. #24686. check it out, and we can argue that one next. really, it's a tough solve and both Toms will like it. let's see what you got! jump in there too, Joe. I have a feeling this one will get a "?"...at any rate, it's infrapink-worthy.

Now I remember. It's ol' what's his name.

hahaha! N, I deleted fb a while ago also. your niece have a name? and is she still in the STL area? or just Mizzou. a couple of the younguns at work graduated MU this past May...

My niece and her daughter are both alumni. My niece worked her way up working for McGraw Hill. Not sure doing what.

did that puzzle, whatever, it is a nice solve :)

Tom - re #13

I puzzled a bit over this. Please excuse my asperger's:

The first instruction should be columns rather than rows.

The second actually limits it to c7-13 with 9-11 black. (This is a common off-by-one error)

*Why* would you dot c20r14 and c15r14? You should put in the method you used on each step.

I see that c20r14 is due to edge logic on the 6 in c20, but I don't get why you dotted c15r14.

Even dotting that place (though I don't know why) will not allow you to finish with LL.

The 3 in c16 can go in four positions, one of which does not overlap any of the others. (And that's after getting rid of r15-17 with edge logic.) So it cannot be placed.

Separately, smile logic can be looked at as a special case of edge logic.

Look at example 10 on http://webpbn.com/index.cgi?page=faq.html

Now do edge logic on the 8:
If it were to go into c1, then it would invalidate r1 (there would not be enough room for both 1s).
Therefore, it cannot go in c1 and then it solves with line logic.

Smile logic is basically recognising those situations where this sort of forcing happens. Simple smiles like examples 8 and 10 can be solved with edge logic. However, abstracting the concept allows us to apply it to a more dreadful situation like example 9.

Joe: Regarding "smile logic," I used it recently in a couple of puzzles and solved a couple using it. It is guessing. Albeit educated guessing. I accept it for what it is. If EL will solve it, I won't even attempt smile. Most of my uncompleted puzzles use smile logic, and I have worked through most of the 20x20. I'll revisit the puzzle when I have some time and truly notate it. Your third statement assumes that I was off by one. I may have been. However, I doubt three times. I'll recheck. Thanks for your time. Kurt agrees with you.

Goto next topic

You must register and log in to be able to participate in this discussion.