2026 UK GP Results

[pinkagape]

Sudoku GP Round 7 sees two relative newcomers in the top 10, both of whom played one previous round. Chaerin Lee scored 496, and Savi Shienhmar scored 595 last time out. The other newer visitor to the top 10 is Zbigniew Zarzycki, with his first score over 900 during his second visit to the top 10, having finished 9th in 2025 Round 2, and being his 23rd visit to the top 50.

In the UK, there was less success than usual, though Zoe Ashford got close to her personal best score with her 10th score over 400.

[pinkagape]

Freddie took a comprehensive victory in the delayed results of Puzzle GP 7 with a lot of Americans also liking this slightly different round, even dropping 2 puzzles. This was his third GP win, and only his seventh time over 1000 - it's not easy to do, despite the way he makes it look.

Meanwhile, Peter and Todd continue to solidify their credentials to challenge Neil and James as they both beat James and Peter also beat Neil. This was Neil's 80th time in the top 50 in 92 appearances, while Peter made his 4th showing. Todd was only 6 points off his best ever points tally as he came just outside the top 50.

[pinkagape]

This time with the promptly released Sudoku GP 8, Freddie took his 3rd top 10 finish. 莉緒 藤本 is a new name at the top of the leaderboard and seems suspiciously good, but Kota Morinishi's 100th Sudoku GP Is one to celebrate as he finishes 4th with his 82nd top 10 finish.

In other UK news, Calum Mailer posted his 50th Sudoku GP with his 22nd top 100 and 22nd score over 400. Amber Pease delivered her best placing and first time in the top 300.

[detuned]

Haven't done the top 3 run down for a while, here goes nothing

Top 3 UK 2026 Puzzle GP Round 5:
2 Freddie Hand 1116
43 Peter Roder 564
50 Neil Zussman 527

Honourable top 100 mentions also to Aleksei Daniliuk, Todd Davies, Adam Bissett, Maurice Blount and James McGowan

Top 3 UK 2026 Puzzle GP Round 6:
2 Freddie Hand 1027
35 James McGowan 595
37 Neil Zussman 569

Peter only 5 points behind Neil, with Todd also making the top 50. Further honourable mentions to Aleksei, Matthew, Calum, Maurice, Desirée and Adam - 11 Uk finishes in the top 100 is pretty remarkable!

Top 3 UK 2026 Puzzle GP Round 7:
1 Freddie Hand 1280
38 Peter Roder 595
46 Neil Zussman 555

Honourable mentions to Todd, James, Aleksei, Jeremy Kong and Calum also inside the top 100.

Well done everyone!

[detuned]

And rounding out Sudoku:

Top 3 UK 2026 Sudoku GP Round 5:
21 Freddie Hand 753.1
29 Calum Mailer 691
91 Mark Goodliffe 470

"Tim Tom" and Tom Collyer were both just outside the top 100

Top 3 UK 2026 Sudoku GP Round 6:
7 Freddie Hand 878.8
63 Calum Mailer 540
113 Adam Bissett 429

Top 3 UK 2026 Sudoku GP Round 7:
23 Freddie Hand 776.5
55 Calum Mailer 587
132 Zoe Ashford 422

Top 3 UK 2026 Sudoku GP Round 8:
10 Freddie Hand 803.8 (probably 9th given likely cheater)
57 Tom Collyer 499
75 Calum Mailer 445

We also had Zoe and Adam lurking just outside the top 100.


Overall, it looks like Freddie's form this year will put him as the highest ever GP finisher for the UK at 16th for the series - possibly with half a chance of the play-offs depending on who does and doesn't come to India. Calum looks like he'll have finished 27th for the series as well, testament to a really impressively consistent season.

[pinkagape]

The final Puzzle GP results for the year are out, and Ken edged Freddie by four minutes. Points were relatively obtainable as Naoki Sekiya made his best ever score in his 50th appearance and wasn't the only one in the top 10 to do so.

The UK also saw three players make their best ever scores: Peter Roder's first score over 600, Aleksei Daniliuk celebrating his 25th appearance, and Maurice Blount with his first score over 500. Further down the table, Jack Smith topped his previous best by 17 in his 57th appearance.

Congratulations everyone!

[pinkagape]

I don't normally look at the overall scores for the year, but I think this makes really positive viewing for the UK (Puzzle GP overall standings 2026):

#2 Freddie Hand
#37 Peter Roder
#44 Neil Zussman
#52 James McGowan
#65 Aleksei Daniliuk
#68 Todd Davies
#82 Maurice Blount
#83 Adam Bissett
#103 Calum Mailer
#110 Desirée Wilson
#118 Matthew White

[Umnik]

Where did you take these overall places from? This page gives slightly different positions: https://gp.worldpuzzle.org/content/wpf- ... sults-2026

[pinkagape]

That’s weird. That is where I was looking. Not sure what happened there!

[Mblount]

It looks like there's two pages of results - the other one is here:

https://gp.worldpuzzle.org/content/wpf- ... -results-0

The difference seems to be that the above link makes some attempt at controlling for difficulty of rounds, with a "PGP score" for each round which is scaled on the 100th ranking score for that round (according to the rules).

[pinkagape]

I think that’s a good way of thinking! I wonder if that’s motivated by that weird German dice round.

[detuned]

I missed the rules change here as well. I think it was in place at the start of this years series however:

Playoff Qualifying Points
Each participant, for each round, will receive Playoff Qualifying Points (PQP). PQP scores are used to determine which participants are invited to the Playoffs (see below).

The 100th-place competitor in the round receives 100 PQP.

For all other competitors, they receive PQP in proportion to how their roundscore compares to the round score of the 100th-place competitor. For example, a competitor who has twice as many points would get 200 PQP, and a competitor who has half the number of points would get 50 PQP.

(Note that PQP is calculated after any points from time bonuses are awarded. The organizers realize that this means that the "strength" of time bonuses can be weighted differently, but we expect the effect to be relatively small and not worth the complexity of a multi-stage calculation that would otherwise balance this more fairly.)

There have been a few normalisation discussions over the years... most obvious example is the 24hpc, which pegs to the round winner, but also pretty similar to the idea used for the 2013 WPC around the world in 80 puzzles (4x rounds of 20 puzzles, but solvers only solved 3 of them), which pegged to 10th place. Pegging points to the solve times of a panel of the same test solvers is also a way of benchmarking as it means the points between rounds should be comparable - this is what the Sudoku GP is doing in contrast to the Puzzle GP - clearly the panel of Puzzle GP testers hasn't proved to be collectively consistent hence the change.

A while ago I had a vague theory that if you do something like take the log of points per minute (or simply time taken, with an individual puzzle) then you end up with a normal distribution - from there you can also normalise by taking the inverse to get a percentile on that distribution. Obviously a bit more complicated than any of that, and not as intuitive to calculate. Doing a proper distribution does have the benefit of not totally breaking down at the edges of the sample - i.e. the round winners - where there's a bit more interest in what the scores are!