Dr. Sudoku Prescribes #63 – Skyscrapers
or solve online (using our beta test of Penpa-Edit tools)
Theme: On a scale of difficulty from one to five — this is “Mostly Five”
Rules: Standard Skyscrapers rules.
Answer String: Enter the 4th row from left to right, followed by a comma, followed by the 7th row from left to right.
Time Standards (highlight to view): Grandmaster = 5:45, Master = 10:00, Expert = 20:00
Solution: PDF
Editorial Note: This should go without saying, but it is advised that you pursue logic as far as you can take it instead of guessing.
I think this was easier than you previous hard Countdown puzzle. The logic in this one was far more forgiving and smooth. Therefore I also liked this one better than the earlier one. I got Roland’s Skyscraper book and this would have fit in well in that book.
I’d concur with the Editorial Note that this one can be pursued with logic all the way to the end.
This definitely gave my testers a much harder time, about 2-3 minutes worse than the Countdown, which I wouldn’t have predicted. I think this one uses the same single type of logic more consistently, so once you now why there is a 3 in a certain place in column 7, you will do a similar thing to place those kinds of digits elsewhere. This helps the flow and overall time certainly.
For my solve the 3 in C7 wasn’t really that important. It was somewhat an easy step near the end of the solve. But I see it can be deduced earlier than I did. The 3 in R6 was the key step in my solve.
It also helps to have done the example on the rules page as it has a somewhat similar opening.
I found this wasn’t as intimidating as it looked at first, especially once I got a foothold into it.
A very tasty Saturday puzzle nonetheless.
Yup, great logic in this puzzle! Probably took me around expert time.
This isn’t counting the 9 minutes I spent on this puzzle before realizing the size was 7×7 and not 8×8 🙁
46 mins!
I invariably get broken atleast once while solving skyscraper puzzles and that is what happened in this puzzle too.I almost always make pencil marks of the possible numbers in a cell and eliminate one by one as the puzzle proceeds.This strategy is not working for me.It would be better if that elusive singles which can enter only into a certain cell are spotted before.
I think i remember stating my dislike for this skyscraper puzzles before.But after solving the above puzzle i am inclined to change my preference for skyscraper puzzles.This was a super good puzzle with a nice clean logical path until the very end.
And about the solve ,other than the 2 crucial deductions stated before namely the 3 in the column 7 and in the later part of the solve the 3 in row 6 i also remember seeing an X-wing!!! of 5 in the middle of the puzzle which helped me proceed.I dont know how many would have used that logic in their solve.It was nice!
I always find it a danger when people start using terms like X-wing in puzzles like Kakuro and Skyscrapers. I think that just makes it sound much harder than it really is, because of the innate difficulty it got assigned to it through Sudoku. Because of the lack of a regional constraint these “X-wings” are much more natural and common in such puzzle types. I really don’t compare such “X-wing” like moves at all in the same difficulty as when they appear in a Sudoku.
Also, I think it’s not the best way to attack this puzzle type as an elimination type. It’s much more important to not the interactions between cells. So it’s more important to note that certain cells have to be higher or lower than others, than keeping track of pencilmarks. This puzzle type really doesn’t at all run at naked singles for me, which makes the elimination/pencilmark strategy much more difficult to use. It’s usually handier to just work these puzzles more on a global scale, working the skyscraper clues instead of the pencilmarks.
For example if you look at C7, you can note that the first square for a 5 needs to be 123 and that helps because there are 3 squares so you get a triple.
But when you look at R7 noting that the first squares are 123 for that 5, really isn’t that helpful.
In C1 again it is a bit more helpful as there is a second 5 clue that influences those numbers a bit and you can note the second must be higher than the first. So you get 12 and 23 and you can note that the digit 1 can’t be visibile for the 5 in row 2, which is an important thing to notice to solve the top of the grid.
This is good advice. I also find I typically don’t write many candidates, except when there are clear < signs that should be marked to give situations like the "12" and "23" in the first cells in C1. Those are the right numbers to mark, as well as < < or "345" "456" in the next two cells. I believe paired spot reductions are common enough in these puzzles to not need the Sudoku name of "X-Wing". But if a solver discovers one and wants to use that name because it the one that is common to them, that is fine with me. It's a fun discovery and maybe easier to spot in a type like this than a larger 9x9 puzzle. So your point that they are not as inherently difficult when in a Skyscrapers or Kakuro is well-taken.
Once I had the big buildings in rows 1 and 2, I did some combination running (with a small amount of trial and error) to limit the top two rows to two possibilities, and then stared at the puzzle for long time until I realized (duh!) what had to happen on the first column. Then it was a steady fill in from there.