Still good stuff after more than two years! Is it fair to say that Almost Locked Sets is a sort of "double Y-wing" ? Y-wing := 3 cells with two candidates being subset of 3 numbers spread over two different rows or columns and two different boxes; ALS: = 4 cells with two candidates being subset of 4 numbers spread over 2 different rows or columns and 3 different cells, with the two cells in one box sharing one common number , and the two cells in same row or column sharing the each another number.
Simon seems to suggest that either method breaks the puzzle open. Perhaps I am missing something obvious, but after either method yields a 9 in r8c9, I cannot see how to get even one more cell. Could anyone point out why getting the 9 helps one to proceed further?
For this to work you need first exclude 9 from r8c2 (you can do it using xy chain (r5c2 (29) - r2c2 - r1c1 - r1c6 - r1c9 - r8c9 (39)) but is it worth?) . Or you still need to spot Finned x wing. With 9 possibility in r8cd you can use almost locked set method, but it is only rule out 3 from r8c8 (you end up with 93 pair r8c2 and r8c8), and it gives nothing. Simon seems to forgot about this 9 when did pencil marks, so this als in case of this puzzle is NOT alternative cuz you still need to use xy chain and f. xwing.
In declaring the finned X-wing, you have precluded the possibility of 3 in r7c1. If 3 can appear in r7c1, r8c9 and r9c8, can you still use the X-wing method? Also, in the second part, you have precluded the possibility of 9 in r8c2. Is this preclusion dependent on the above X-wing for the number 3 (i.e. then r8c9 must be 9)?
The possibility of 3 in r7c1 doesn't affect the X-wing at all; the finned X-wing is concerned only with the location of 3 in r1 and r9, and Simon's logic didn't require the absence of a 3 in r7c1.
Although I really do not want to sound at all critical, I have a lot of difficulty following your logic because I am so distracted by your always-whizzing cursor, which can be hypnotic. I get stuck watching that cursor fly around, sometimes clicking back and forth and back and forth between two squares, and do not hear what you are saying, or cannot concentrate on what you are saying. That disappoints me because you are teaching techniques I very much want to learn. Unfortunately, at the end, I find I have no idea what you did. I suppose my ability to focus has declined over the years, as I used to be one of those people who could concentrate to the point of not hearing what was going on around me. Not so anymore, as I age. And my short-term memory decline is no help. I continue to marvel how you guys remember so much.
I think the confusing aspect may be the pencil marks. r9c8 is not ONLY limited to 3 or 7. It can also take a 4. I can't remember if I used hybrid pencil-marks in this video or not but, in general, any pencil marks I make in these videos will indicate that a number can only go in two positions in the 3x3 box. Sometimes when I'm solving slightly faster I'll be more flexible and either a) make pencil marks where a cell can only contain exactly 2 numbers or b) go even further and fully pencil-mark a box for all the possibilities it can contain. But option b) here will be relatively rare. In this example, r9c8 is definitely not suitable as it does contain the extra unknown.
It appears to me that the almost locked method is a version of the two string kite method. The bottom middle block is the body of the kite and row 8 and row 9 are the strings which means that cells in row 8 col 2 and row 9 col 9 cannot both be false.
@@CrackingTheCryptic I realise I'm very late to the party with this one, but doesn't the possibility of a 9 in r8c2 damage the almost locked set approach?
Thank you Mr Simon for explaining so lucidly the difficult concept of Almost Locked Sets which occur in Expert level puzzles. warm regards
Still good stuff after more than two years! Is it fair to say that Almost Locked Sets is a sort of "double Y-wing" ?
Y-wing := 3 cells with two candidates being subset of 3 numbers spread over two different rows or columns and two different boxes;
ALS: = 4 cells with two candidates being subset of 4 numbers spread over 2 different rows or columns and 3 different cells, with the two cells in one box sharing one common number , and the two cells in same row or column sharing the each another number.
I like it! Thanks!
Simon seems to suggest that either method breaks the puzzle open. Perhaps I am missing something obvious, but after either method yields a 9 in r8c9, I cannot see how to get even one more cell. Could anyone point out why getting the 9 helps one to proceed further?
For this to work you need first exclude 9 from r8c2 (you can do it using xy chain (r5c2 (29) - r2c2 - r1c1 - r1c6 - r1c9 - r8c9 (39)) but is it worth?) . Or you still need to spot Finned x wing. With 9 possibility in r8cd you can use almost locked set method, but it is only rule out 3 from r8c8 (you end up with 93 pair r8c2 and r8c8), and it gives nothing.
Simon seems to forgot about this 9 when did pencil marks, so this als in case of this puzzle is NOT alternative cuz you still need to use xy chain and f. xwing.
In declaring the finned X-wing, you have precluded the possibility of 3 in r7c1. If 3 can appear in r7c1, r8c9 and r9c8, can you still use the X-wing method?
Also, in the second part, you have precluded the possibility of 9 in r8c2. Is this preclusion dependent on the above X-wing for the number 3 (i.e. then r8c9 must be 9)?
The possibility of 3 in r7c1 doesn't affect the X-wing at all; the finned X-wing is concerned only with the location of 3 in r1 and r9, and Simon's logic didn't require the absence of a 3 in r7c1.
I abort most puzzles because of the almost locked sets
good stuff
Although I really do not want to sound at all critical, I have a lot of difficulty following your logic because I am so distracted by your always-whizzing cursor, which can be hypnotic. I get stuck watching that cursor fly around, sometimes clicking back and forth and back and forth between two squares, and do not hear what you are saying, or cannot concentrate on what you are saying. That disappoints me because you are teaching techniques I very much want to learn. Unfortunately, at the end, I find I have no idea what you did. I suppose my ability to focus has declined over the years, as I used to be one of those people who could concentrate to the point of not hearing what was going on around me. Not so anymore, as I age. And my short-term memory decline is no help. I continue to marvel how you guys remember so much.
Why couldn't r9c8 form the almost locked set in r9 rather than r9c9?
I think the confusing aspect may be the pencil marks. r9c8 is not ONLY limited to 3 or 7. It can also take a 4. I can't remember if I used hybrid pencil-marks in this video or not but, in general, any pencil marks I make in these videos will indicate that a number can only go in two positions in the 3x3 box. Sometimes when I'm solving slightly faster I'll be more flexible and either a) make pencil marks where a cell can only contain exactly 2 numbers or b) go even further and fully pencil-mark a box for all the possibilities it can contain. But option b) here will be relatively rare. In this example, r9c8 is definitely not suitable as it does contain the extra unknown.
@@CrackingTheCryptic Thanks for that - much appreciated. All clear now.
It appears to me that the almost locked method is a version of the two string kite method. The bottom middle block is the body of the kite and row 8 and row 9 are the strings which means that cells in row 8 col 2 and row 9 col 9 cannot both be false.
@@CrackingTheCryptic I realise I'm very late to the party with this one, but doesn't the possibility of a 9 in r8c2 damage the almost locked set approach?
@@Alex_Meadows I have the same question--did you figure out an answer?