WPC 17 - part 1 (of 2, not 17)
No way I feel like opening the email box while here in Belarus - probably too much work piling up - but I can leave a comment on the first two days of the championships.
We flew from Philly to Washington to Frankfurt to Belarus. The travel agent that booked my flights on my National Treasure Hunt prize neglected the flight number I chose, so while the rest of the US team flew the Washington-Frankfurt leg together, I was alone on an earlier flight. Flights are always tiring for me and this was no exception. The highlight happened on the trip to Washington. We had an ~ 1h flight. Wei-Hwa and I both had the "easiest" volume 1 KenKen books from the US Sudoku Championship. We thought - why not race, and try to beat both the flight (from at one gate to at the next gate) if we could, from back to front. I expected to be done in an hour. Our race started rather close, me maybe a single puzzle ahead after 15, and then the "hard" puzzles got harder - not yet very hard, but much harder than 100-85. That is when I broke away. There were some times (like when the grids went from 5x5 to 4x4, or from +/- to + only, when I had to shift thinking a bit. Still, finished the 100 puzzle book in 44'58". The best moment though was watching Wei-Hwa as he scrambled to beat the other gate end of the goal. With 3 seconds left before the unfasten seat belt ding, he finished, just over an hour and some change. Now, I'm not recommending the books since I am hardly the audience for them, but I doubt it is the next sudoku.
The hotel here is very tall, and meals are on the "22nd" floor, with a panoramic view of the surroundings. We took a walk the first night into the city a bit, sampled the McDonald's in the area which had a gyro burger of sorts, and then quickly crashed to sleep after the opening dinner, having not slept since 36 hours previously.
Tourism day was spent a bit inside the city, and then at an area in the countryside that had very developed woodcrafts and pottery. Some interesting items, but a fairly uncommon tour nonetheless. Most of the bus time was spent pre-solving puzzles and examples for the upcoming championship. Solving things like an arrows puzzle with turbulence is not recommended, and I tried my best but had to erase frequently and curse at other times when a sudden turn/bump caused chaos.
The first day began with a welcome round that was particularly unwelcoming. All the puzzles used the 32 dot WPC logo that Vlad et. al had put together for the championship - a very nice logo - but it seemed they wrote 3-4 more puzzles on it than they needed. The big 32 topology puzzle could have been - nice - if I had time to look at it - but adding in a 2 distances puzzle (basically breaking the 32 dots into 2 sets of 16 - plus 3 navigrids - plus an Easy As, plus a summing (where I lost all my time), ..., made the round unfriendly as no one was close to finished. I made a careless error closing out the beads and when writing the sequence 5-23, decided mentally to add the leading two everywhere and 25-23 was marked wrong. 12 points I should have had (55) instead of 43. The top score of 75 was hardly close to the 140 or so in the round. Welcome rounds should be more welcoming.
I did a bit better in the second round, the first of two two hour slogs, but encountered the problem I'd face all championship. I'd finished the puzzles I wanted to do, or knew a great route to do, and then had to make choices. I spent a lot of time on the numbers in squares puzzles and got neither. Others untouched were the Minimax patience which took too much time in the example and the Domino Castle which also seemed to be hard on first look to break into with 3 branch points on the bottom row seeming to be the best. 301 was ok, but without solving anything really in the last 30 minutes, I lost some ground to those who broke into the 330 range.
The first team round involved assembling 6 sets of 9 squares in different puzzle ideas. I started on the square location (with circles representing vertices in colors on the nine pieces). After 30 minutes our team had solved 5 of the puzzles, all but mine, and I would contribute nothing to the team's results. It was though, as received by other teams, one of the hardest if not the hardest of the set. Still, I didn't start lunch until I saw a solution.
The second team round after lunch was the Weakest Link round. This round type fails ALL THE TIME if the individual puzzles are too hard. The only real flaw I remember from WPC Eger - my favorite WPC still - was the impossible darts puzzle which kept teams from being able to solve as teams. This round had a hard individual part, which Roger solved first for our team, then me with 10 minutes left (50 minutes having been spent cracking the 4 individual puzzles, including using back-solving data from puzzle 4 to help get into puzzle 3 which should not be required. Wei-Hwa reached the team desk with 10 seconds remaining, which was hardly enough time to do anything, let alone finish the puzzle (which seemed to need all 4 team member’s clues, even though it was claimed it could be solved with no clues in some amount of time (“days” Vlad joked).
Part 5 was timed too long for me, particularly when I have an arrows weakness. In each round, EVERY NUMBER, as a clue in a minesweeper or an outside clue in a battleship, was +/- of the correct value in the answer. I solved 8 of the puzzles in the round efficiently, even using my experience constructing 9x9 battleship puzzles from my first book of puzzles to know how to break in with a 54 clue in consectuive rows. However, after 8 puzzles, I had 8 minutes on the clock and no idea how to solve the arrows. I have never learned to do this type well, so to do a +/- 1 variation where there are no simple work ins (like an 8 in the grid becoming a 7 which forces a lot), I had too much data with no path through it. I felt like saying finished then and there as I was “finished”, just not in the good way. Still, only Ulrich finished the round so not losing ground to the rest of the field.
Part 6 generally had potential, but was super hard after a long day where every round was hard. In this championship, the number 17 is being used a ton. In this round, there were 5 17x17 puzzles. They are easily the hardest different neighbours I ever solved. The every second turn was ok. I did the Tents which had some good massive sum constraints to break in (most often discovered after I had an error and needed to erase a whole section), and I was racing the clock at the end to do the star battle. I get everything in but notice with 30 seconds left I have a region with 2 next to a region with 4, no trivial way to resolve this problem, and no time - having wasted 5 minutes on the other tough Pills puzzle - to get those points. Still, 3 of the 5 was a solid showing.
At the end of the day (although scores weren’t released until the morning), Hideaki Jo of Japan was 1st with 554 and Pal Madarassy and I were tied with 553, exactly one point behind. Alphabetically I lost the tie-breaker and was third. Ulrich was just behind his teammate Michael Ley in 6th/7th, about 20 points behind, having only done one of the 17x17 puzzles in Part 6 after getting very broken in the star battle and not being able to recover. So, it looked like a close fight for the finals (which are crazy, and I will discuss later). The Germans, who finished the first team part (and were the only team to do so, no team finishing the Weakest Link) had an ~80 point edge on the US in second and then Japan.
Day two began with a sprint round - in practice by far the easiest of the rounds. I had a long mental cramp on the crypted square, spending 5 minutes on it after 25 on the other 16 puzzles combined (17 puzzles again I think) and was the only one to say finished. I’m still wondering if I remembered to get all the A’s in the AYDA, but I trust I am clean.
Round eight followed, another brain-busting two hour round. Now, long rounds need not be brain busting, but this championship has so few types we’ve seen before and many many types that seem to just require exploring a very deep search space which is the kind of puzzle I hate generally and am not as efficient at particularly. I did my best - papers not turned back yet - but Ulrich was quite happy afterwards and with 60 more points than me at least, had jumped ahead of me by 30 or so.
Round nine was a team round that defines why Wei-Hwa must be on our team. It was a manipulative round, building “diamonds” from assembling pieces that were cut to have 3/4/5/6 cube bits of a 3x3 cube when assembled. Some have called it like a Soma cube - I mean to look this up including spelling - but that was the kind of assembly. We had 30 pieces, five pairs of six representing the diamonds you could make. We marked the distinct faces, arranged the blocks by size, and then tried to get one together. Roger found the first, which greatly reduced our search space, but Wei-Hwa got all the rest, and clearly was the strongest of us at this exercise. We finished 3rd as a team, but with Germany only getting 3 of the cubes, it now looked like a race between us and Japan (who finished behind us, but not by a lot time bonus-wise) in the round.
Round ten was a “chrono part” team round with an ordering the pictures puzzle Zack and Roger did which was no problem, and a very challenging chain puzzle which Wei-Hwa and I worked on. We mapped the 8 constellations of the 17 double-sided pieces representing the flags of the nation countries and cities of the first 17 WPCs. Then we tried to string this whole chain to sit in the grid. Based on some unlikely constraints, we eliminated 4 of the choices, proved 2 trivially were unsolvable, and then spent all the time with the other 2. Wei-Hwa eventually went off paper exploring one route, but somehow in his copying of data and my checking, we missed an error. Zack/Roger/myself could not solve state A and WH could not do state B so we did not finish. Germany did, with 10 minutes left, and one other team, but we had still gained more in the first team round of the second day.
Round eleven was the highlight of the championship so far, but also another sign of how being tired leads to some slowness where I shouldn’t have it. It had 6 puzzles with letters replacing numbers in a fences, sudoku, four winds, skyscraper, snake, and clouds, and because the letters used between puzzles were shared you could figure constraints on some of them (F not equal to E, both in fences and sudoku so two of 1,2,3 - but look - not a 3 - or whatever). It was the only sudoku in this year’s WPC so far, and I got 4 puzzles done very efficiently, the skyscrapers was tough, but the clouds was very difficult. There was one as yet unassigned letter. I COULD have counted the row and column to figure it out, but forgot this was possible. So I solved without an easy 0. It took 15 minutes. Ulrich was likely behind entering the clouds (he would never take 15 minutes on a clouds and would know to find that 0) but finished 4.5 minutes before me. Still, 2nd with over 16 minutes left in the round should give me good placement and time bonus presuming I am clean.
The last round, cards, defeated me for several reasons. Some might be too much puzzle experience. The puzzle involved 6 grids of constraints on cards in rows/columns with suits not touching and all rows/columns summing up to a certain number. By including 2 jokers, a whole set of 54 cards was used over all 6 puzzles. Now, the 6 constraint sets were listed on the left side of a page. On the right, were 6 grids to fill (with a suit or rank given) were there to assist. Seeing no reason to separate clues from grids if they naturally went together by location (you would space grid+clue together as a 3x2 set), I figured they weren’t directly assignable by location. Mishearing a comment in the instruction meeting further compounded this confusion. I wasted 10 minutes trying to solve a square without the grid clues and eventually forced the fifth one to be reasonably solved with ranks and some suits, say it could only fit one place (the fifth one), and figured out the correlation. Many asked proctors about this and got clarification. If its not type-set well, and 4+ questions about it get asked, you should generally make an announcement to the room to update people about this fact. Anyway, still 50 minutes to work through it. I got 3.9 done (one card in the fifth grid, the one I had done first, was a JQK of diamonds but I did not know which. I spent the last 6-8 minutes trying to finish the 6th set and had the answer but had a JQdiamonds unforced which added to the other part to tell me I had a non-unique answer. I was not absolutely confident of other assignments, so I tried to relax constraints and then resolve the 6th set another way to find the correct answer. Turns out the puzzle has 2 solutions, which should totally have been constrained down as possible by changing the identity of the clues. So, encountering non-uniqueness cost me the 5th set, and likely the 6th if I had time to turn to it. So, while others finished, two sets of puzzle presentation + non-uniqueness sticking points spoiled an ok round.
Overall, the individual puzzles have been tough. Rounds have a lot of new designs, but not in the better way to present a new design (a round of new type XXX, 7 puzzles say, with an easy, medium, hard, very hard, etc. so people learn what to do) but instead 10-12 new types in a 2 hour round. Also, having so many puzzles (often 2 of each type), just makes the work much harder. I know there is a WSC now, but sudoku are puzzles too. Not really represented here. No observation puzzles (identify similar pairs, word search, counting, spot the differences - frankly, a “screen test” round) hurts too. There are few word logic puzzles (fill-ins, etc.) which I’m good at, but Zack and Roger are very good at, so we are a bit hurt by that too. I’m expecting a placement after general qualifier somewhere in the top 5, with Ulrich 1, but we’ll see when results are posted.
Other things - been playing Tichu this year, a fun game, with the US team. Some RftG (Race for the Galaxy) but not as much although Wei-Hwa brought a set for Georgi so I expect to join them once I am done here. The testing room is tiny, and hot. I expected to need warm layers, instead I’m running outside between rounds to cool off. Oh, and the finals use a format I decidedly dislike.
17 people make it, and it will be a ladder system. the lowest 5 start (13-17) with 3 puzzles, and the top 2 from that run-off will advance (10-12+2 finishers) and this continues until 1-3 + two finishers face off in a longer 5 puzzle final. There is some time bonus based on results, which simplifies the process, but the problem is as someone waiting awhile to actually start puzzles, I will not know what I will be solving. This is because there were 17 puzzles written for the finals, and 13-15 get to choose the 3 they want in their run-off before it starts. Then 10-12 choose the 3 they want in their run-off. 5 puzzles will remain for the grand championship. Given smart competitors will choose types they are familiar with first, I worry that there will be either very new types or very difficult to solve types left before the finals. Certainly no sudoku or bridges will last that long.... Anyway, I’ll spell-check later but thought people would want some WPC news.
We flew from Philly to Washington to Frankfurt to Belarus. The travel agent that booked my flights on my National Treasure Hunt prize neglected the flight number I chose, so while the rest of the US team flew the Washington-Frankfurt leg together, I was alone on an earlier flight. Flights are always tiring for me and this was no exception. The highlight happened on the trip to Washington. We had an ~ 1h flight. Wei-Hwa and I both had the "easiest" volume 1 KenKen books from the US Sudoku Championship. We thought - why not race, and try to beat both the flight (from at one gate to at the next gate) if we could, from back to front. I expected to be done in an hour. Our race started rather close, me maybe a single puzzle ahead after 15, and then the "hard" puzzles got harder - not yet very hard, but much harder than 100-85. That is when I broke away. There were some times (like when the grids went from 5x5 to 4x4, or from +/- to + only, when I had to shift thinking a bit. Still, finished the 100 puzzle book in 44'58". The best moment though was watching Wei-Hwa as he scrambled to beat the other gate end of the goal. With 3 seconds left before the unfasten seat belt ding, he finished, just over an hour and some change. Now, I'm not recommending the books since I am hardly the audience for them, but I doubt it is the next sudoku.
The hotel here is very tall, and meals are on the "22nd" floor, with a panoramic view of the surroundings. We took a walk the first night into the city a bit, sampled the McDonald's in the area which had a gyro burger of sorts, and then quickly crashed to sleep after the opening dinner, having not slept since 36 hours previously.
Tourism day was spent a bit inside the city, and then at an area in the countryside that had very developed woodcrafts and pottery. Some interesting items, but a fairly uncommon tour nonetheless. Most of the bus time was spent pre-solving puzzles and examples for the upcoming championship. Solving things like an arrows puzzle with turbulence is not recommended, and I tried my best but had to erase frequently and curse at other times when a sudden turn/bump caused chaos.
The first day began with a welcome round that was particularly unwelcoming. All the puzzles used the 32 dot WPC logo that Vlad et. al had put together for the championship - a very nice logo - but it seemed they wrote 3-4 more puzzles on it than they needed. The big 32 topology puzzle could have been - nice - if I had time to look at it - but adding in a 2 distances puzzle (basically breaking the 32 dots into 2 sets of 16 - plus 3 navigrids - plus an Easy As, plus a summing (where I lost all my time), ..., made the round unfriendly as no one was close to finished. I made a careless error closing out the beads and when writing the sequence 5-23, decided mentally to add the leading two everywhere and 25-23 was marked wrong. 12 points I should have had (55) instead of 43. The top score of 75 was hardly close to the 140 or so in the round. Welcome rounds should be more welcoming.
I did a bit better in the second round, the first of two two hour slogs, but encountered the problem I'd face all championship. I'd finished the puzzles I wanted to do, or knew a great route to do, and then had to make choices. I spent a lot of time on the numbers in squares puzzles and got neither. Others untouched were the Minimax patience which took too much time in the example and the Domino Castle which also seemed to be hard on first look to break into with 3 branch points on the bottom row seeming to be the best. 301 was ok, but without solving anything really in the last 30 minutes, I lost some ground to those who broke into the 330 range.
The first team round involved assembling 6 sets of 9 squares in different puzzle ideas. I started on the square location (with circles representing vertices in colors on the nine pieces). After 30 minutes our team had solved 5 of the puzzles, all but mine, and I would contribute nothing to the team's results. It was though, as received by other teams, one of the hardest if not the hardest of the set. Still, I didn't start lunch until I saw a solution.
The second team round after lunch was the Weakest Link round. This round type fails ALL THE TIME if the individual puzzles are too hard. The only real flaw I remember from WPC Eger - my favorite WPC still - was the impossible darts puzzle which kept teams from being able to solve as teams. This round had a hard individual part, which Roger solved first for our team, then me with 10 minutes left (50 minutes having been spent cracking the 4 individual puzzles, including using back-solving data from puzzle 4 to help get into puzzle 3 which should not be required. Wei-Hwa reached the team desk with 10 seconds remaining, which was hardly enough time to do anything, let alone finish the puzzle (which seemed to need all 4 team member’s clues, even though it was claimed it could be solved with no clues in some amount of time (“days” Vlad joked).
Part 5 was timed too long for me, particularly when I have an arrows weakness. In each round, EVERY NUMBER, as a clue in a minesweeper or an outside clue in a battleship, was +/- of the correct value in the answer. I solved 8 of the puzzles in the round efficiently, even using my experience constructing 9x9 battleship puzzles from my first book of puzzles to know how to break in with a 54 clue in consectuive rows. However, after 8 puzzles, I had 8 minutes on the clock and no idea how to solve the arrows. I have never learned to do this type well, so to do a +/- 1 variation where there are no simple work ins (like an 8 in the grid becoming a 7 which forces a lot), I had too much data with no path through it. I felt like saying finished then and there as I was “finished”, just not in the good way. Still, only Ulrich finished the round so not losing ground to the rest of the field.
Part 6 generally had potential, but was super hard after a long day where every round was hard. In this championship, the number 17 is being used a ton. In this round, there were 5 17x17 puzzles. They are easily the hardest different neighbours I ever solved. The every second turn was ok. I did the Tents which had some good massive sum constraints to break in (most often discovered after I had an error and needed to erase a whole section), and I was racing the clock at the end to do the star battle. I get everything in but notice with 30 seconds left I have a region with 2 next to a region with 4, no trivial way to resolve this problem, and no time - having wasted 5 minutes on the other tough Pills puzzle - to get those points. Still, 3 of the 5 was a solid showing.
At the end of the day (although scores weren’t released until the morning), Hideaki Jo of Japan was 1st with 554 and Pal Madarassy and I were tied with 553, exactly one point behind. Alphabetically I lost the tie-breaker and was third. Ulrich was just behind his teammate Michael Ley in 6th/7th, about 20 points behind, having only done one of the 17x17 puzzles in Part 6 after getting very broken in the star battle and not being able to recover. So, it looked like a close fight for the finals (which are crazy, and I will discuss later). The Germans, who finished the first team part (and were the only team to do so, no team finishing the Weakest Link) had an ~80 point edge on the US in second and then Japan.
Day two began with a sprint round - in practice by far the easiest of the rounds. I had a long mental cramp on the crypted square, spending 5 minutes on it after 25 on the other 16 puzzles combined (17 puzzles again I think) and was the only one to say finished. I’m still wondering if I remembered to get all the A’s in the AYDA, but I trust I am clean.
Round eight followed, another brain-busting two hour round. Now, long rounds need not be brain busting, but this championship has so few types we’ve seen before and many many types that seem to just require exploring a very deep search space which is the kind of puzzle I hate generally and am not as efficient at particularly. I did my best - papers not turned back yet - but Ulrich was quite happy afterwards and with 60 more points than me at least, had jumped ahead of me by 30 or so.
Round nine was a team round that defines why Wei-Hwa must be on our team. It was a manipulative round, building “diamonds” from assembling pieces that were cut to have 3/4/5/6 cube bits of a 3x3 cube when assembled. Some have called it like a Soma cube - I mean to look this up including spelling - but that was the kind of assembly. We had 30 pieces, five pairs of six representing the diamonds you could make. We marked the distinct faces, arranged the blocks by size, and then tried to get one together. Roger found the first, which greatly reduced our search space, but Wei-Hwa got all the rest, and clearly was the strongest of us at this exercise. We finished 3rd as a team, but with Germany only getting 3 of the cubes, it now looked like a race between us and Japan (who finished behind us, but not by a lot time bonus-wise) in the round.
Round ten was a “chrono part” team round with an ordering the pictures puzzle Zack and Roger did which was no problem, and a very challenging chain puzzle which Wei-Hwa and I worked on. We mapped the 8 constellations of the 17 double-sided pieces representing the flags of the nation countries and cities of the first 17 WPCs. Then we tried to string this whole chain to sit in the grid. Based on some unlikely constraints, we eliminated 4 of the choices, proved 2 trivially were unsolvable, and then spent all the time with the other 2. Wei-Hwa eventually went off paper exploring one route, but somehow in his copying of data and my checking, we missed an error. Zack/Roger/myself could not solve state A and WH could not do state B so we did not finish. Germany did, with 10 minutes left, and one other team, but we had still gained more in the first team round of the second day.
Round eleven was the highlight of the championship so far, but also another sign of how being tired leads to some slowness where I shouldn’t have it. It had 6 puzzles with letters replacing numbers in a fences, sudoku, four winds, skyscraper, snake, and clouds, and because the letters used between puzzles were shared you could figure constraints on some of them (F not equal to E, both in fences and sudoku so two of 1,2,3 - but look - not a 3 - or whatever). It was the only sudoku in this year’s WPC so far, and I got 4 puzzles done very efficiently, the skyscrapers was tough, but the clouds was very difficult. There was one as yet unassigned letter. I COULD have counted the row and column to figure it out, but forgot this was possible. So I solved without an easy 0. It took 15 minutes. Ulrich was likely behind entering the clouds (he would never take 15 minutes on a clouds and would know to find that 0) but finished 4.5 minutes before me. Still, 2nd with over 16 minutes left in the round should give me good placement and time bonus presuming I am clean.
The last round, cards, defeated me for several reasons. Some might be too much puzzle experience. The puzzle involved 6 grids of constraints on cards in rows/columns with suits not touching and all rows/columns summing up to a certain number. By including 2 jokers, a whole set of 54 cards was used over all 6 puzzles. Now, the 6 constraint sets were listed on the left side of a page. On the right, were 6 grids to fill (with a suit or rank given) were there to assist. Seeing no reason to separate clues from grids if they naturally went together by location (you would space grid+clue together as a 3x2 set), I figured they weren’t directly assignable by location. Mishearing a comment in the instruction meeting further compounded this confusion. I wasted 10 minutes trying to solve a square without the grid clues and eventually forced the fifth one to be reasonably solved with ranks and some suits, say it could only fit one place (the fifth one), and figured out the correlation. Many asked proctors about this and got clarification. If its not type-set well, and 4+ questions about it get asked, you should generally make an announcement to the room to update people about this fact. Anyway, still 50 minutes to work through it. I got 3.9 done (one card in the fifth grid, the one I had done first, was a JQK of diamonds but I did not know which. I spent the last 6-8 minutes trying to finish the 6th set and had the answer but had a JQdiamonds unforced which added to the other part to tell me I had a non-unique answer. I was not absolutely confident of other assignments, so I tried to relax constraints and then resolve the 6th set another way to find the correct answer. Turns out the puzzle has 2 solutions, which should totally have been constrained down as possible by changing the identity of the clues. So, encountering non-uniqueness cost me the 5th set, and likely the 6th if I had time to turn to it. So, while others finished, two sets of puzzle presentation + non-uniqueness sticking points spoiled an ok round.
Overall, the individual puzzles have been tough. Rounds have a lot of new designs, but not in the better way to present a new design (a round of new type XXX, 7 puzzles say, with an easy, medium, hard, very hard, etc. so people learn what to do) but instead 10-12 new types in a 2 hour round. Also, having so many puzzles (often 2 of each type), just makes the work much harder. I know there is a WSC now, but sudoku are puzzles too. Not really represented here. No observation puzzles (identify similar pairs, word search, counting, spot the differences - frankly, a “screen test” round) hurts too. There are few word logic puzzles (fill-ins, etc.) which I’m good at, but Zack and Roger are very good at, so we are a bit hurt by that too. I’m expecting a placement after general qualifier somewhere in the top 5, with Ulrich 1, but we’ll see when results are posted.
Other things - been playing Tichu this year, a fun game, with the US team. Some RftG (Race for the Galaxy) but not as much although Wei-Hwa brought a set for Georgi so I expect to join them once I am done here. The testing room is tiny, and hot. I expected to need warm layers, instead I’m running outside between rounds to cool off. Oh, and the finals use a format I decidedly dislike.
17 people make it, and it will be a ladder system. the lowest 5 start (13-17) with 3 puzzles, and the top 2 from that run-off will advance (10-12+2 finishers) and this continues until 1-3 + two finishers face off in a longer 5 puzzle final. There is some time bonus based on results, which simplifies the process, but the problem is as someone waiting awhile to actually start puzzles, I will not know what I will be solving. This is because there were 17 puzzles written for the finals, and 13-15 get to choose the 3 they want in their run-off before it starts. Then 10-12 choose the 3 they want in their run-off. 5 puzzles will remain for the grand championship. Given smart competitors will choose types they are familiar with first, I worry that there will be either very new types or very difficult to solve types left before the finals. Certainly no sudoku or bridges will last that long.... Anyway, I’ll spell-check later but thought people would want some WPC news.