@ 2013-03-26 1:12 PM (#10447 - in reply to #10446) (#10447) Top | |||||||||||||||||||
Posts: 191 Country : The Netherlands | Richard posted @ 2013-03-26 1:12 PM Notes to V2V (IV) V: Thermometers A lot of thermometers are built on a few given digits and long thermometer shapes. Not much have diagonal lines. Thermometers had to be one of the easier puzzles, since it is a well known variant, so I used pretty much givens and a straight forward route through the puzzle. R46C5 = {78} R3C3 = 8; R2C4 = 9; R7C2 = 3 R2C6 = 8; R1C7 = 9 R8C4 = 8; R9C3 = 9 R7C8 = 1; R3C5 = 5; R2C5 = 2; R8C5 = 3; R7C5 = 6; R3C8 = 2 V2V: Hot/Cold Thermometers I had the idea for this puzzle type already for a long time, it just didn’t happen to make a few puzzles out of it. In fact the type is a blend of Increasing and Thermometers, with the bulb in the middle of the line. There are a few specific moves/conclusions in this puzzle: R7C5 = 9; R5C7 = 9; R2C7 = 7; R2C8 = 8; R2C4 = 3; R2C3 = 5; R2C5 = 2; R5C8 = 1; R4C6 = 1 9 in row 3 in Block 1; 9 in C3 in Block 1 :=> R3C3 = 9; R1C3 = 2; R3C1 = 6 R5C2 = 2; R7C2 = 5; R3C79 = {12}; R3C8 = 3; R1C7 = 5; R6C8 = 5; R7C8 = 4 4 in Block 7 locked in R89C3 :=> R5C3 = 7/8; R4C2 = 6/7; R4C3 = 7/8; R4C4 = 8/9; R4C2 = 6; R6C3 = 3 R7C3 = 1; R4C5 = 5 4 in R5 locked in R5C46; R8C5 = 4 Not much troubles to be expected after this stage. | ||||||||||||||||||
@ 2013-03-26 1:18 PM (#10448 - in reply to #10447) (#10448) Top | |||||||||||||||||||
Posts: 191 Country : The Netherlands | Richard posted @ 2013-03-26 1:18 PM Notes to V2V (V) V: Rank Rank was one of my favourite puzzles in Classics vs Innovatives, so I deliberately wanted to include such a puzzle in one of my own tests. Towards the end it contains a bit of a tricky step, that can be overseen easily. R9C8 = 7; R2C5 = 9; R3C5 = 1; R3C8 = 3; R3C4 = 6; R3C7 = 5; R3C2 = 8; R3C3 = 2; R2C6 = 8; R4C6 = 9; R9C4 = 9; R8C5 = 8; R9C5 = 4; R7C3 = 5; R9C6 = 5; R8C13 = {34}; R7C2 = 9; R8C9 = 9; R5C3 = 9; R6C8 = 9; R9C9 = 6; R1C8 + R2C7 = {46}; R2C9 = 1; R1C9 = 8; R2C3 = 7; R7C8 = 1; R8C4 =1 Towards the end, placing the 7 in Block 4 is a crucial step in getting the solution logically. By that time you know that R6C4 = either 7 or 8. That means that the 7 in block 4 can’t go in R4C2, since R4C1 must be 8 and no possibility is left for R6C4 in that case. V2V: Rank killer As soon as I saw the Rank sudoku for the first time, with the dotted outlined boxes, I realised that this puzzle type could be combined with killer in a beautiful way, so that variation to a variant was settled early in the puzzle creating process for this test. R9C1 = 5; R789C2 = {123}; R9C6 = 4; R9C789 = {789}; R9C3 = 6 R9C4 = 3; R8C4 + R9C5 = {12}; R7C4 = 8; R7C56 = {56}; R5C6 = 1; R6C6 = 2; R7C1 = 7; R7C3 = 9; Maximum for R7C89 = 7 (3+4); maximum of R8C9 = 6; R6C8 = 9; R7C8 = 3; R7C9 = 4; R8C9 = 6 R8C2 = 3; R5C1 = 9; R3C5 = 8; R4C6 = 8; R5C2 = 5; R6C1 = 6; R5C4 = 4; R34C4 = {79}; R2C45 = {12}; 1 in Block 3 locked in R13C8; R6C9 = 1; R3C6 = 3; R1C5 = 4; R1C4 = 6; R6C4 = 5; R1C6 = 9; R2C6 = 5 R8C5 = 9; R8C6 = 7 From here no problems can be expected... | ||||||||||||||||||
@ 2013-03-26 3:22 PM (#10449 - in reply to #10448) (#10449) Top | |||||||||||||||||||
Posts: 191 Country : The Netherlands | Richard posted @ 2013-03-26 3:22 PM Notes to V2V (VI) V: Even Sandwich I like to explore the possibilities that certain types offer. In the case of Even Sandwich there are some interesting steps during the solve that make it a lot easier to proceed, one of them very early in the process. Also: by knowing there is an even digit and no sandwiched digit in the respective row or column there must be odd digits two cells away from the even digit. R5C3 = 2; 9 in row 4 in R4C45; R4C6 = 2; R4C4 = 9; R4C57 = {46} R35C4 = {48} 3 in C5 in R23C5; R3C6 = 1; R1C4 = 5; 5 in C6 in R56C6; R6C5 = 7; R6C4 = 3 R1C36 = {68}; R1C2 = 9; 9 in R9 only in R9C3 :=> R9C4 = 2; R9C2 = 8; R9C6 = 7 R9C78 = {16} 1 in R8 in R8C3; R8C2 = 2; 7 in Block 7 in R8C1 R5C1 = 3; R46C1 = {19} … V2V: Sum Sandwich Thinking about a variation to Even Sandwich, more ideas crossed my mind. I used two of those ideas for the Advent Calender at Logic Masters Germany, namely Consecutive Sandwich and Next to Nine. I thought the third idea would be suitable for this test: Sum Sandwich. I realised quite a lot of puzzles for the test were already on the hard side, so I decided to make this one very easy. The sums of 3 and 4 provide easy starts for this puzzle. | ||||||||||||||||||
@ 2013-03-26 3:33 PM (#10450 - in reply to #10449) (#10450) Top | |||||||||||||||||||
Posts: 191 Country : The Netherlands | Richard posted @ 2013-03-26 3:33 PM Notes to V2V (VII) Here are the final notes to the test puzzles. Please feel free to respond, ask more clarifications, point me to other solving strategies (or typo's). 1~9 Playing with sums in sudokus is one of my favourite habits. I really liked the idea of a ‘floating sum’ somewhere in the row/column immediately when I saw it once at a Sudoku Cup. Deriving information from the grid, limiting the positions of 1 and 9, it’s all in here. R2C3 = 4; R2C4 = 9 In row 8 the digit outside the sum is 2 (33 +1+9=43), so: R8C1 = 2; R8C2 = 9; R8C9 = 9 Column 2: digits in R19C2 must be {68} Column 4: since R8C4 <>1; R9C4 = 1; R1C4 = 6 Column 3: digits outside the sum add to 8; 1 or 2 cells. Since R89C3 <>1 and 9: R1C3 = 9; R7C3 = 1; R89C3 = {35} … Frame 1~9 I had little reservations towards this type. I am not too fond of sudokus with complicated rules and I was a bit afraid that the large number of clues outside the grid would be confusing. I liked the fact that on one side you give more information about each row and column, but on the other side, by giving the clues in an increasing order, you take away information as well. So all in all, I gave it a go. Important information: no zero outside the grid means no 1 or 9 in the first and last cell, and no neighbouring 1 and 9 as well. Wheather people found all those outside clues confusing during solving or not I can’t tell, but it’s the puzzle with the highest percentage of correct entries. R28C5 = {19} R6C9 = 2; R5C9 = 1; R7C9 = 9; R9C8 = 1; R9C9 = 7; R8C9 = 3; R78C8 = {25} R9C7 = 4; R8C7 = 6; R8C8 = 5; R7C8 = 2; R9C1 = 9 9 in R4 must go in R4C4; R4C13 = {78} … | ||||||||||||||||||
@ 2013-03-26 6:25 PM (#10451 - in reply to #10230) (#10451) Top | |||||||||||||||||||
Posts: 69 Country : Germany | Realshaggy posted @ 2013-03-26 6:25 PM Hi Richard, thanks for the contest, nice Sudokus as expected. My performance felt like average for me, and had no luck too with the timing, missed the Sum Sandwich by maybe 30 seconds (but thats only 30 points). I felt I needed a break after the last contest, so I had some more or less puzzle-free weeks and no preparation, otherwise I think I would have felt "overpuzzled" in the upcoming WPF Grand Prix' and of course the German Championships. Now it's time to start training again ;-) After solving the remaining half, I'm happy I didn't do the Greater than X. Nice path and my favourite in the set, but would have cost much too much time in the contest. | ||||||||||||||||||
@ 2013-03-26 8:30 PM (#10452 - in reply to #10230) (#10452) Top | |||||||||||||||||||
Posts: 152 Country : United Kingdom | detuned posted @ 2013-03-26 8:30 PM
Enjoyed the test, albeit I had to do it in something of a rush. Thanks Richard! | ||||||||||||||||||
@ 2013-03-27 12:58 AM (#10456 - in reply to #10230) (#10456) Top | |||||||||||||||||||
Posts: 18 Country : Germany | Nola posted @ 2013-03-27 12:58 AM
When I say "too little time" means I need much more to solve all puzzles. However, the contests shouldn't be any longer (I even would prefer shorter, but I understand that it is than more difficult to have a balanced choice of puzzles). So the time is really ok. | ||||||||||||||||||
@ 2013-03-27 1:28 AM (#10457 - in reply to #10230) (#10457) Top | |||||||||||||||||||
An LMI player | An LMI player posted @ 2013-03-27 1:28 AM
| ||||||||||||||||||
@ 2013-03-27 4:17 PM (#10459 - in reply to #10230) (#10459) Top | |||||||||||||||||||
Posts: 6 Country : India | ka_bharath posted @ 2013-03-27 4:17 PM
|