WPF Sudoku GP

Information and discussion on Puzzle Competitions
Post Reply
detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

WPF Sudoku GP

Post by detuned » Fri 04 Jan, 2019 12:10 am

The 2019 season starts off this weekend, with Dutch authors taking their traditional 1st round spot.

http://gp.worldpuzzle.org

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Wed 09 Jan, 2019 8:45 pm

Preliminary results for round 1 are out, top 3 UK are as follows:

31. Tom Collyer 576
62. Sam Cappleman-Lynes 456
64. Mark Goodliffe 452

http://gp.worldpuzzle.org/content/preli ... ?sort=stat

I was about a minute away from finishing a last classic, but couldn't quite manage it. Look's like a strong showing to start things off!

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Tue 05 Feb, 2019 9:48 am

Top 3 UK Round 2 according to the preliminary results are:

45. Sam Cappleman-Lynes 491
96. Mark Goodliffe 382
101. Heather Golding 371

Only 12 all-correct finishers this time, compared to 28 for Round 1, so a step up in difficulty! And 19 solvers (including myself) who were able to solve everything except the Sum-Product Killer...

I'd left myself with about 20 minutes at the end to solve the Sum-Product Killer and the first Classic; made a lot of progress with the Killer, but after 15 minutes realised I had no way to finish it without breaking the constraint on at least one cage. It's a neat puzzle but not a variant I particularly enjoyed.

Did anyone else find that the points values didn't match (at all!) their solving times for the puzzles? I thought that the Braille was severely overvalued, the hardest Classic was severely undervalued, the Irregular was much harder than 43 points and some of the other mid-range puzzles weren't deserving of quite as many points as had been given them. Maybe it's just me, but normally my times don't vary *quite* this much from what I'd expect given the points.

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Wed 06 Feb, 2019 11:21 am

I knew Classic 5 was familiar to me! For those who have access, take a look at the final Classic from WSC 2017 Rd 5. Almost identical clue-layout, and the break-in step is the same as in that puzzle too, although less obvious here and not telegraphed as clearly.

In that puzzle, the crucial digits all fall on the main diagonal of givens, and here none of them do. They do, however, still have the following two properties:

1. They appear as givens three times, in three 3x3 boxes, of which no two are aligned either horizontally or vertically with each other
2. Among the three appearances as givens, two of them occur in corresponding cells of the 3x3 box.

With this clue layout, that is sufficient as a general rule to force a fishy friend to appear.

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Sun 10 Feb, 2019 6:42 pm

I had a go at this earlier today. I think I'd have been amongst those who got all but the sum-product killer out. The repeating numbers within cages wasn't particularly nice to see.

Re the hard classic - making a well placed guess to get going wasn't particularly difficult. I guess it's a matter of taste about these appearing in competition, but I can understand it not having more points than it did as ultimately I think it's possible to solve these puzzles very quickly. Would be interesting to see Mark give it a "Cracking the Cryptic" treatment potentially!

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Thu 07 Mar, 2019 7:47 pm

Top 3 UK round 3:

35 Sam Cappleman-Lynes 516
45 Mark Goodliffe 483
49 Jeremy Kong 482

Jeremy Kong I think is a new name name for me, but a top 50 result is very creditable!

On a personal note, after I clicked go on this I had a good 15 minute battle with my printer and the wifi here before I even had anything in front of me to solve, and even then it seemed pot luck as to whether the codes were being submitted properly.

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Sun 31 Mar, 2019 1:08 am

Round 4 is happening this weekend

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Tue 02 Apr, 2019 7:00 am

Preliminary results should be ignored for the time being, as the points values showing as being awarded for each puzzle are still the points values from GP 3.

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Tue 02 Apr, 2019 9:56 pm

Looks like the issue has been fixed for the preliminary results. Top 3 UK:

28. Sam Cappleman-Lynes 508
44. Tom Collyer 454
45. Mark Goodliffe 451

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Sun 28 Apr, 2019 10:43 am

Round 5 happening this weekend!

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Wed 01 May, 2019 8:14 pm

Tough round I thought, and I didn't think I'd done particularly well:

42. Tom Collyer 457
51. Jeremy Kong 429
73. Mark Goodliffe 390

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Wed 29 May, 2019 2:18 pm

For round 6, I did not get the hidden palindrome logic at all... perhaps I’m missing something obvious??

I’ll post the uk results up later

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Wed 29 May, 2019 3:53 pm

The main thing is to narrow down the cells that can/cannot be the removed cell along a palindrome line. Work out which cells are the possible "mirror" cells along a line based on which cell is removed, and narrow down the possibilities based on which pairs can't contain equal values. That becomes easier once you already have some values on the line, obviously!

To actually get started, the way I found was that, by normal Sudoku logic, R6C6 (edited to have the correct coordinates) is a 6, and since a 6 cannot go anywhere else along that line, the remaining cells on that line must be palindromic.
Last edited by Feadoor on Wed 29 May, 2019 6:25 pm, edited 1 time in total.

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Wed 29 May, 2019 5:56 pm

Top 3 UK:

38. Sam Cappleman-Lynes 501
55. Tom Collyer 426
58. Jeremy Kong 416

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Wed 29 May, 2019 9:55 pm

Feadoor wrote:
Wed 29 May, 2019 3:53 pm
The main thing is to narrow down the cells that can/cannot be the removed cell along a palindrome line. Work out which cells are the possible "mirror" cells along a line based on which cell is removed, and narrow down the possibilities based on which pairs can't contain equal values. That becomes easier once you already have some values on the line, obviously!

To actually get started, the way I found was that, by normal Sudoku logic, R6C6 (edited to have the correct coordinates) is a 6, and since a 6 cannot go anywhere else along that line, the remaining cells on that line must be palindromic.
I was fixated in the top left for a while, made a speculative guess that didn’t work and was left stumped.

I’ve solved it now, it was still a bit fiddly after that and I can’t really see how the constraint would give you something nicer. I wouldn’t be too unhappy to never see that particular type again that’s for sure.

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Wed 26 Jun, 2019 6:36 pm

Round 7 happened at the weekend. Top 3 UK:

42. Sam Cappleman-Lynes 449
72. Tom Collyer 367
94. Matthew White 335

I thought that was a hard round, especially the Sum by X and Renban puzzles.

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Wed 10 Jul, 2019 11:24 pm

Since I've just done it for the puzzle GP, I'll list out the overall GP standings with 1 round yet to play:

37. Sam Cappleman-Lynes 2921 - 6 scoring rounds
49. Tom Collyer 2646 - 6 scoring rounds
63. Mark Goodliffe 2455 - 6 scoring rounds
66. Jeremy Kong 2394 - 6 scoring rounds
130. Matthew White 1786 - 6 scoring rounds
155. Ned Walker 1637 - 6 scoring rounds
182. Sam Boden 1507 - 6 scoring rounds
190. Howard Wilkinson 1470 - 6 scoring rounds

A bit more spread out for the Sudoku GP then. Beating the previous top UK ranking of 36th looks in reach for Sam, albeit he has a higher points total already in what I feel has been a pretty tough competition this time round. Breaking the 3000 point barrier would be an excellent achievement!

Tom has a low score of 366 to work out of his set of results - but this would still mean finishing the last round with bonus time to get close to Sam. Given the results so far, an improvement of 100-200 points on this seems more realistic, leaving him outside the (current) top 40.

Mark and Jeremy are pretty close together - I haven't checked for low scores, but both are looking at improvements between 150-250 points to start thinking about the top 50.

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Thu 11 Jul, 2019 6:34 am

A second-worst round of 449 means I'd need to attain a new high score of 528 on Round 8 to break 3000 points. It's possible - I've had 3 scores of over 500 points already.

Feadoor
Posts: 78
Joined: Thu 17 May, 2018 7:08 am

Re: WPF Sudoku GP

Post by Feadoor » Sat 13 Jul, 2019 5:09 pm

Feadoor wrote:
Thu 11 Jul, 2019 6:34 am
I'd need to attain a new high score of 528 on Round 8 to break 3000 points.
So that means I can leave the awkward-looking symmetric product sum puzzle alone and "just" do everything else.

Shame there's also a medium nonconsecutive or I'd give this plan a go!

detuned
Posts: 1825
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: WPF Sudoku GP

Post by detuned » Sat 13 Jul, 2019 11:57 pm

What a horrible looking variant.

I can only hope the instruction booklet is a bluff and that the competition puzzle only has 2x2 regions. Even in that case I feel like you are stretching the idea of what makes a good sudoku variant pretty far. The fact this needs a 2nd attempt within the IB to explain what is going on does this variant no favours whatsoever in my eyes...

Post Reply