Exocet 2014_09 Discussion #2

Advanced methods and approaches for solving Sudoku puzzles

Exocet 2014_09 Discussion #2

I'm behind on following the responses in Discussion #1, so I decided to start a separate discussion on a specific puzzle anomaly.

I tried to compare my solver's solutions to those found by champagne in 03_E1_exo_je.txt. There were instances where I found three Exocet patterns, but champagne listed (at most) two JExocet patterns. No big deal. However, I ran across several other instances where I found two Exocet patterns and champagne listed one JExocet pattern. Again, no big deal ... until I examined one of those patterns.

Code: Select all
`98.7..6..75.....8...6......4...3.5....56...9......2..3..78...5.....2.1.......1..4;13461;GP;kz0;1;r3c1 r3c2 r2c4 r1c8 1234;; +--------------------------------------------------------------------------------+ |  9       8       1234    |  7      B145    B345     |  6       1234    125     | |  7       5      Q1234    |  12349   1469    3469    |  2349    8       129     | |  123     1234    6       |  123459  14589   34589   |  23479   12347  R12579   | |--------------------------+--------------------------+--------------------------| |  4       12679   1289    |  19      3       789     |  5       1267    12678   | |  1238    1237    5       |  6       1478    478     |  2478    9       1278    | |  168     1679    189     |  1459    145789  2       |  478     1467    3       | |--------------------------+--------------------------+--------------------------| |  1236    123469  7       |  8       469     3469    |  239     5       269     | |  3568    3469    3489    |  3459    2       345679  |  1       367     6789    | |  23568   2369    2389    |  359     5679    1       |  23789   2367    4       | +--------------------------------------------------------------------------------+ # 175 eliminations remain ### -1234- QExocet   Base = r3c12   Target = r1c8,r2c4==r1c9 ### -1345- QExocet   Base = r1c56   Target = r2c3,r3c9 r1c5  = 1; -1r2c3 =  X-Wing c34\r46 - r8c46   = r3c8 - r3c12 = 1r2c3 r1c6  = 3; -3r2c3 =  X-Wing c34\r89 - r8c89   = r3c8 - r3c12 = 3r2c3 r1c56 = 4; -4r2c3 = r8c3 - r8c4 = r6c4 - r6c8 = r3c8 - r3c12 = 4r2c3             5r1c56 - r1c9 = 5r3c9`

If any of <134> is true in the base cells, then it must be true in r2c3. However, if <5> is true in the base cells, then it must be true in r3c9. So, my solver included the second QExocet above. Now, this pattern isn't a JExocet, but it does raise a flag in my tired/sleepy head. Your thoughts!!!

There are 5,227 puzzles where my results differ from champagne's results; and, I'm not sure how often the explanation is simply that I found three patterns and champagne only reported two of his patterns. However, it is possible that the above observation is not a single anomaly.

_
daj95376
2014 Supporter

Posts: 2624
Joined: 15 May 2006

Re: Exocet 2014_09 Discussion #2

Regarding that one, my JE code recognize only one exocet, but in the "bande" search, I got these 3

98.7..6..75.....8...6......4...3.5....56...9......2..3..78...5.....2.1.......1..4;
3;
r1c5 r1c6 r2c3 r3c9 ;
r3c1 r3c2 r1c8 r1c9 ;
r3c1 r3c2 r1c8 r2c4 ;;;

I'll check later to see why

I also run the file 14_09 after changes in the preliminary eliminations and got that table

Code: Select all
`            lev1     lev2  lev3   lev4    totalJE       1006135     1096   188 182428  1189847Bande      74761      237   161  15559    90718other        135       17    40      206     1398no exo                                   247636                                        1529599`

level 1 to 3 contradiction single digit
level 4 contradiction full expansion of a true in base false in target

here, some puzzles are duplicated, but the tendency is there.
most of the exocets appear at the very beginning or in extended mode;
A huge majority is qualified as JE
The ratio of puzzles having an exocet is close to 84%

level 1 to 3 contradiction single digit
level 4 contradiction full expansion of a true in base false in target
champagne
2017 Supporter

Posts: 6711
Joined: 02 August 2007
Location: France Brittany

Return to Advanced solving techniques