WSPC Report - WPC side

The WSPC 2021 recently concluded, and so I can post my experiences going through the 10 rounds of the WPC side. I'm writing these shortly after attempting each round, and will update with final placements/relative scores. (1st/10th/my score) Let's go!

Round 1: Welcome to WS + PC (600 points, 11 puzzles, 60 minutes)

After spending an hour dealing with technical difficulties, and delaying a day because the contest opened at the early hour of... 10 AM, I was finally able to start at around 2 AM on November 14th. Going into this round I didn't expect to finish due to the combination of 2 Diagonal Lighthouses (I'm bad at the base type AND seeing diagonals) and 2 Slitherlink variants (I'm okay at this now, but still relatively slow). Even still, I knew I had an outside shot. I started by knocking out the Star Battle, then moved to the Log In (crisscross) and solved it smoothly. Then the Anglers, which I broke by miscounting the top half a couple times- it took a couple minutes of tweaking to get things right since I knew my structure was correct, but had things slightly off. I then went to skip the Hexa Slitherlink and almost destroyed the Country Road- I instantly saw the kind of subtle break-in and the correct path for the other half, but misjudged and thought it had to be wrong. A couple minutes later I logically reached my intuited conclusion but with one small change, and that was that. The Hexa Slitherlink was then another smooth solve since it didn't look that bad- I stalled for 20-30 seconds a couple times but never felt stuck.

Top: A clean solve with no mistakes.
Bottom: A clean start with a slightly failed intuit of the bottom that was so close to right.
Fun themes on both!

And then the round went downhill.

I broke the Castle Wall and couldn't fix it (I was one very small adjustment away from the answer, too...), then recovered with a quick solve of the Double Choco. The 2-0 Slitherlink was another smooth solve and an exceptional puzzle- at so many points there seemed to be no progress, but there was always one little thing. The first Diagonal Lighthouses looked pretty easy, so I worked through that slowly with a couple errors from misreading diagonals. I made my only answer key mistake here, missing a lighthouse late in a row. At this point, I had 18 minutes left and 3 puzzles, worth a combined 175 points- barely finishing pace. Unfortunately, I'd looked at the Neighbors briefly and didn't see how to get started, so I went for the Diagonal Lighthouses. It was painful but I got a clean logical solve with exactly 8 minutes left- at this point, I was happy with my round score and was faced with a tough choice: go for the 65 point Neighbors and maybe get it, or retry the 30 point Castle Wall and probably get it. I went for the latter, broke it again, but then spotted a deduction I'd missed twice and got it with 2 minutes to spare.

I managed to solve everything but the Neighbors. Looking back, this was a solid performance- nothing exceptional, but given the technical issues going in and the mix of puzzles I'm not sure I would have been able to do much better- the Neighbors took me quite a while afterward.

727.67 (Ken Endo) / 535 / 532 (12th)

Favorite puzzles: 2-0 Slitherlink, Country Road

Round 2: In Memory of Maki Kaji (750 points, 20 puzzles, 75 minutes)

Going into this round, I expected a very strong result- all of these are established Nikoli types, and as someone who spends a lot of time solving on puzz.link/db I'm extremely quick on a lot of these types. I hoped to finish the round and see that borne out in practice. My strategy was simple: leave Masyu, Midloop and Slitherlink for last, probably in that order, as I'm relatively weaker at the selection of loop types than other types. I started at 11:15 PM on November 14th.

And immediately began to throw, breaking both Nurikabes (usually a strong type for me) and taking 7:28 for both. Then my nemesis from the Puzzle Grand Prix, the Shakashaka answer keys... where I continued to throw by misreading a pair of 2s as 3s and solving a different puzzle than what was actually there, gah. Unfortunate that it was unique that way, too. I took a good minute on just these answer keys but got them both right, then moved to Heyawake for a pair of quick solves. Broke the 2nd Norinori and had to restart due to packing 3 shaded cells in a region early, then struggled through the first Mochikoro. The rest of the round was very clean and speedy though, with the only stalls being misreading a column on the first Yajikazu, and obstinately not putting in an answer key for the second Sashigane until I solved it even though the column was determined. The Masyu, Slitherlink and Mid Loops all went well- what a bizarre round, where I did poorly where I expected to do well and vice versa. With a better start this could have been 1000 points (no breaking + less penalties) but I'm satisfied here.

I'm pretty sure my papers are cleaner than most, if not all other top solvers here. It's also really funny to think that, like most of the Japanese contingent, I have a sort of "home field advantage" with a mix like this.

1091.5 (Walker Anderson) / 799.83 / 954 (finish at 53:31, 5th)

Favorite puzzles: Heyawake 2, Yajikazu 1, Mid-Loop 2

Round 3: Made in China (1500 points, 21 puzzles, 90 minutes)

In stark contrast to the previous round, this one is all unfamiliar types. In particular Mahjong Maze just looks impossible, but it has 270 points wrapped up in it. Given the round structure, I wanted to start at the back with Triangular Dissection and Go, then probably try Muddy Creek before the mathy trio of High Wu, Elastic Sums and Magic Autumn. If this round is scaled the same way the others have been, I probably won't get through that entire list and can just ignore Mahjong Maze- I spent a good amount of time thinking about ways to approach those, and got nowhere, and a pure brute force approach won't work because there's 1716 possibilities. Though that is small enough that, if it comes down to the wire with a couple minutes left, the 110 point one is probably the best use of time... even still, for starting the round I was at the mercy of others as Matthew (my 2 year old) had fallen asleep at 6:30 PM instead of his usual 8 or later, and I didn't want to risk having to bail on what could be the most important round to attend to him. The hope was that he'd wake up and go almost right back to sleep, at which point I would feel comfortable beginning, but until then the concern would definitely affect my performance.

I started at 1:10 AM on November 15th. The round felt like a complete disaster from beginning to end, with the only part going according to plan being that I didn't have to touch the Mahjong Mazes at all. I jumped around a lot so a blow by blow doesn't make much sense, but I solved: Go 1, Go 2 (slightly miscounted, penalty), Muddy Creek 1 (broke, penalty), Go 3, Triangular Dissection 1, Muddy Creek 3, Triangular Dissection 2, Magic Autumn 1 (transposed answer keys, penalty), Magic Autumn 3, and collectively spent 20 minutes not solving the third Triangular Dissection. There was a really sneaky trick that I caught the potential for about a third of the way in and was sure I could get it, but then misremembered my math result and submitted wrong with 7 minutes left. A minute later I spot what's probably right but couldn't make it work and guessed for the second key, but guessed wrong (as expected). I also spent a bit of time on High Wu but didn't see any real progress, and a bit of time on Muddy Creek 2 which with 13 minutes left I probably should have gone for instead of the second Magic Autumn and last puzzle.

So I solved 9/21 but it looks like everyone struggled with this round to some extent, but not as much. I'd be really happy if I got the Triangular Dissection given the results but as it is, not terrible. This was definitely a high variance, low fun round for me. From solving the rest of the round afterwards, Triangular Dissection 3 is ABSURD and I'm surprised I got as close as I did, Elastic Sums 3 was free points if I didn't get scared off by the first two, Mahjong Maze 3 was actually approachable if I'd bothered to look at it, and with optimal puzzle selection this could have been a 900-1000 point round. Ouch. Genuinely, though, if I'd cut my losses on the Triangular Dissection and looked elsewhere I could have gotten another 165 points with no other changes, and of course getting that puzzle would have been 140 (which again was noticing 1 detail that I'd noted the potential for but didn't see for over 15 minutes). Getting into speculative territory, the last Muddy Creek was also potentially gettable, but I don't think I'd have gone for it next.

Three of the hardest puzzles in the round. Look how many extra notes I took!
Also Mahjong Maze 3 was literally just the honors tiles and was like a 3 minute solve after the fact, I should have at least looked. :(

Favorite puzzles: Go 2, Magic Autumn 3

Least favorite genre: High Wu, it's actually so annoying. Though I want to specially shout out Go here as a genre that has a lot of potential and I definitely want to try making some.

880.5 (apiad) / 750 / 616.5 (27th, though obviously 1 solve away from 10th. This round was a bit of a crapshoot honestly.)

Round 4: Made in India (520 points, 12 puzzles, 55 minutes)

A much friendlier mix, where I'm at least a bit familiar with all 6 types. Consecutive Quads at a total of 50 points are probably what I'll aim for last, and the rest we'll see how it goes! I ended up unable to sleep the day before and woke up around 4:30 AM on the 16th, and started this round about an hour later after getting some food and preparing.

In short: I finished this round clean! This entire round was lovely from start to finish, though I struggled a bit with the Place by Product and Heterocut. Canal View, Rassi Silai, Balance Loop and the first Place by Product fell in a bit over 20 minutes, and then I misunderstood the center of the big Place by Product briefly- and overly narrowed the possibility space for the 24s, which caused some confusion later that I was fortunately able to correct and trace back. The first Heterocut was easy, and then the second looked hard so I did the Consecutive Quads, finishing those with 15 minutes left on the clock. The last Heterocut then took nearly 12 minutes, as I spent 3 minutes finding the break-in and then proceeded to do it wrong. A few minutes later I find the correct start (intuitively, from trying to find any other arrangement) and then run into a problem with the regions of size 4 and needing 6 of them. Wait, that's a 5 pointing at a 5! Fix that! Now there's a 4 pointing into a 6 and that 4 can be a 5 and it just works!

The eraser marks on this grid can honestly tell a lot of my process for solving it on their own.

I definitely could have been 5-10 minutes quicker here with smoother solves on the big ones, but this was good.

Favorite puzzles: Canal View 1, Balance Loop 2, Place by Product 2

755.33 (Freddie Hand) / 520.33 / 557 (finish at 51:18, 6th)

Round 5: Multi-Skyscrapers (350 points, "3" puzzles, 35 minutes)

I suck at Skyscrapers. Here's the entire plan: panic (and probably leave the regular one alone to maximize my chance of getting the 100 point 2nd puzzle and the entire 225 point combined one). I expect a pretty poor result here. I went straight into the round after writing the above recap, at 6:35 AM on November 16th.

I spent 16 minutes on the 100 point puzzle to start and then irreparably broke it. I then got the single puzzle and 5 of the subgrids in the last one, for a dismal score of 147.5 (1 penalty). I hate Skyscrapers. At least this is a "low" scoring round... I think if I'd gone straight for the 225 point puzzle I could have gotten it, but the partial points only being available on that one made it too tempting to leave for later.

Favorite puzzles: being done with this (the huge linked grid is neat)

Least favorite: wasting literally half the time for no points due to breaking a puzzle- I feel like for a 20 minute round this would be a decent score though. It's just that it's 35 minutes, so it's subpar.
525.5 (Roland Voigt) / 371 / 147.5 (tied 104th, oof)

On the left: 17.5 minutes of effort for 125 points and about 7-8 minutes away from 100 more. (if you look closely, you can find an error I haven't caught yet!) On the right: 17.5 minutes of effort for a broken puzzle and 25 points. I've learned my lesson for similar structures in the future: if you stall 8 minutes in with a ton to go, just cut your losses and bail. It looks like my mistake was taking a very deep bash and then mistakenly thinking it contradicted, then running with the other side which put a BUNCH of stuff into the grid that was wrong and nigh-impossible to trace.

So at the halfway point, I've has what I would consider 2 good rounds (2 and 4), 2 bad rounds (3 and 5) and one okay round that's probably good but I still feel like it could have been better (1). I wonder where that puts me in the standings- probably like 15th at this stage, given the abysmal Skyscrapers. I did solve the second puzzle afterwards without pencilmarks as practice and it's pretty fair- just sad that this round was 1 mistake away from being decent.

Round 6: All Stars (770 points, 14 puzzles, 75 minutes)

This round, in contrast, looks like a fun mix. As far as number placement goes when it involves math I'm usually fine at it, so Kakuro and Terra XX shouldn't be too bad. The only thing that I'm scared of is the 3D Gokigen, because tracking loops on that is hard enough as it is. I'll probably solve with the paper constantly being rotated, honestly! I ended up waking up at 5:30 AM, and just like the day before, started nearly an hour later.

The first page. You can tell I started careful and then decided that drawing numbers/drawing borders was for people who don't want to go fast. The Hebi solve was also mostly intuitive.

As expected, this round was really fun! I'd have done relatively better on a 70 or 80 minute round due to how things shook out, but I'm really happy with how I did. I opened with the Kurotto (very clever!) in 3 minutes, then moved to the Snake (2 minutes), Kakuro (5 minutes + a bit of starting the Terra XX before moving on), Tetroscope (2 minutes), Hebi (2 minutes), and Nanro Cave (5 minutes). Every single one of these puzzles was great, and on the back half I started on the Clone Yajilin and stalled. I took a couple bashes, broke it, and did the Koburin and Yajilin Castle in this time. Then the highlight of the round for me- the Heyawacky and Heyawake 3D, both of which are exceptional puzzles that I knocked out in a combined 7 minutes for 140 points. The Nurikabe LITS tricked me a bit with the top right, but I got it out with 26 minutes to go. I burned half of my remaining time solving the Terra XX (oops I'm still bad at this) and took a penalty here for misreading a note I put in the grid (a "G" for outside grid notes) as a 6 in a solution row. Oops! Then I knocked out the Clone Yajilin for real, and then got two thirds of the way through the Gokigen 3D before time ran round despite having 8:30 left on the clock. So if I didn't break the Clone Yajilin, I probably finish the round. If it's an 80 minute round, I probably finish the round. If it's a 70 minute round, other peoples scores go down and mine stays the same. You see what I mean about how it shook out? I think all of my decisions on what to go for were correct and this result is nearly ideal for me though.

Favorite puzzles: Literally half the set, but especially the Kurotto, Nanro Cave and Heyawacky.

936.33 (Ken Endo) / 590 / 663.5 (6th)

Round 7: 20 by 20 (600 points, 5 puzzles, 60 minutes)

Another high variance round, albeit with mostly comfortable types. The gimmick is obvious: 5 puzzles at 400 cells each. I'll probably aim to start with the Index Yajilin (I'm pretty good at basic Yajilin techniques, and the variant seems workable) and Cave, and then go from there depending on what the Spiral Galaxies looks like. If I don't break anything horribly this should be a clean finish?

I started this round at 12:40 on November 17th, and what a rush- I ended up starting with the Akari since it looked really easy and had it within 4 minutes, though putting in the answer key took a minute and I still used the wrong row once by accident. I then jumped straight into the Cave which took just under 10 minutes (including keys), and then tackled the Index Yajilin. I missed some easy deductions of the form of impossible indeces for a while but got through it in 20 minutes. I definitely could have been 5 to 10 minutes quicker on this if I came better prepared, but the example puzzle being only 6x6 and the type being novel combined to obfuscate the critical ideas until I was solving the main puzzle. The Masyu then took 7:40 and the Spiral Galaxies 6:40 to finish the round in 48:32. Neither of those solves were too eventful, though I did spot a uniqueness trick in the Masyu that I was only mostly sure of and then traced the other side to a relatively short but subtle contradiction.

Left: The worst puzzle from the round- every step was trivial and the puzzle didn't even take advantage of the size- there totally could have been a couple big galaxies!
Right: The best puzzle from the round. Great logic the whole way through and a fun theme.

Favorite puzzle: the Cave, definitely. What a slick theme.

809.83 (Ken Endo) / 495 / 706.67 (finish at 48:32, 4th)

Round 8: Night and Day (700 points, 8 puzzles, 70 minutes)
Oh no crossing loops oh no oh no multiple crossing loops

I don't know how to deal with this round yet, the variant rules are very wacky and hard to follow. I'll probably work mostly in order despite the Round Trips being at the back and worth the most points because it's definitely my weakest base type in the round.

I ended up sleeping from about 7:30 PM to 3 AM that night (for reference, I tend to end up on a schedule of either sleeping from 6 PM to 2 AM or 5 AM to 2 PM when left to my own devices), and if you work it out yes I did switch from the latter to the former during this week. Insomnia can be rough. I spent a while thinking about how to notate the 2 loops, watched a couple videos, and jumped into the round just past 4.

I broke every single puzzle I attempted multiple times but still managed to correct and get both Masyu, both Tapa-Like Loops, and the first Railroad Tracks. At this point I had 12 minutes left and tried the second Railroad Tracks, getting about halfway through. I did not enjoy this round at all- crossing loops are already a headache, multiple loops are already a headache, and combining them with extra rules on when they can cross is a double headache. This definitely seemed like a round where making some guesses could save a lot of time, but I was too afraid of breaking things irreparably! In hindsight, I think starting with the Masyu was a mistake, and I should have opened with the Tapa-Like Loop and Railroad Tracks, then gone to the Masyu for cleanup. Getting Railroad Tracks 2 instead of both Masyu would have been a gain of 25 points, and I would have gotten it. Without factoring in potentially solving the first Masyu after, too, which likely would have happened for a gain of 50. Not a big mistake though, and this was always going to be an uncomfortable round that felt tailor-made to target my weaknesses. Having finished the rest of the round, the second Round Trip was very approachable and I probably could have gotten it in place of anything except the first Masyu. It was definitely intimidating though! In hindsight, Round Trip was the least impacted by the added rules, since it already had free crossing.

The Railroad Tracks from the round. I had the left half of the second puzzle during the time limit, and honestly these were both huge struggles with lots of erasing. I wish I hadn't been scared off by the idea of Round Trip with 2 loops...

Favorite puzzle: Tapa-Like Loop 2. It had the fairest solve, I think.

961 (Ken Endo) / 609.5 / 335 (64th, oof)

Round 9: Triple Jump (950 points, 18 puzzles, 80 minutes)

I'm looking forward to this round, most of the puzzles seem pretty approachable except the Outside Hashi, but maybe the scale is relatively small. The fact that ALL islands in a row or column are given is definitely helpful (empty rows/columns are clued as such), there might be a gimmick to the puzzle too. I'll probably skip the Kropki and work my way through, and decide what to spend the rest of the time on then. It's not that I'm bad at Kropki, it's just that it's probably slower than the rest of the round.

It ended up feeling a bit more like hurdles than a triple jump, due to two very stupid errors on my part. But let's start at the beginning- it'll be interesting to explain why I think my result was the best I could reasonably expect.

I opened with both LITS puzzles within 4 minutes, then the LITZ in about 6. I kept missing sneaky possibilities for the T and Z but eventually got used to the mix. Then I destroyed both Fivecells in 1:20 and 3:00... except I had a small error in the corner of the second one. This really threw me and I spent a minute checking my answer keys and the solution, and didn't find anything wrong. I then messaged Prasanna quickly to ask what I should do, while having a nagging feeling that it had to be my fault because no one else had reported this issue. Sure enough, I find my mistake and correct it, but at the cost of 3 minutes + 8 points.

I then go through the Tapas (2:00, 3:30, 6:15) with the Alternative Tapa being very nice. I caught what the break-in had to be almost immediately but still struggled to find what the details were- extremely subtle. Then the Hashi, where I got the first in 2:20 (with an answer key error- I noted a dot couldn't be passed in one direction and wrote a 0 down for later, but it was crossed later. Oops!) and then... disaster. I really struggled with the second Hashi, breaking it several times. The problem? I misread a 3 as a 4 and the version I was trying to do actually was impossible. I don't know how that happened, but it apparently took me over 10 minutes to catch. It's hard to say exactly, because I alternated between the Outside Hashi (!!!) and this for nearly a half hour. I did manage to get both at least, with the Outside Hashi being a much fairer solve than I expected. If I read the second Hashi right, probably 10 minutes saved here. But with 13 extra minutes, how could I not score extra?

Well, with 17:30 left on the clock and 6 puzzles to go, I went for the Tents. The first one took me a while and I didn't see much progress, so I jumped to the second one and knocked it out. 13 minutes left. I then returned to the first and solved it with some uniqueness arguments. 11 minutes left. Kropki 1 looked small and approachable, so I did a guess and check on R6C2, which eventually worked. 8:47 left on the clock and I have Kropki 2, Kropki Pool and Big Tents left. I elected to start with the Kropki Pool and made some progress towards determining the digit set but not enough- there were only 4 minutes left. So I made the call to try to intuit and tweak the Big Tents, which I actually managed to do with less than a minute to spare!!! And that's why I think this is about the best I could have hoped for- without the specific set of circumstances leading into the ending puzzle selection, I don't think I would have tried to intuit the Big Tents as a last ditch effort to get any points, and though I may have gotten the 2nd Kropki, that would have been roughly the same amount of points that I got. I still don't know how to do the Big Tents logically, by the way :P

Left: A careful, methodical solve to make sure there were no errors.
Right: A frenzied, intuitive solve to try to get points at the last minute. You can tell how much of my starting guesses were correct from the lack of erasing on half the grid. I think getting this solve was my favorite moment from the competition.

Favorite puzzles: LITZ, Fivecells 1 (very smooth fun solve), Alternative Tapa, Outside Hashi

1214.17 (Ken Endo) / 795.5 / 817 (8th)

Round 10: Mitosis (600 points, 14 puzzles, 60 minutes)

The final round, and one with a very math-heavy mix. I don't expect to finish this medley of puzzles- Hundred is very fiddly and slow, Dice Skyscrapers is a terrifying concept especially given round 5, and Pento Fences is something I can do but not quickly. Oh and a 60 point Battleships is scary too. Most people would probably be afraid of a 70 point Curve Data but I think that sounds like fun puzzle. We'll see how this one goes, I could see myself getting anywhere from 400 to 700 points on this round.

I started this final round at 4:25 AM on November 19th. Instantly, disaster struck as the printer jammed and refused to print! I had already knocked out the Hundred (which I practiced for!!!) in under a minute while listening for the printer to finish, so I went to the Dim Sum. I miscounted something, as I ended up with 2 values of 27*. One of them was a lot more prone to a miscount than the other, so I went with that as my answer and it was right. I still burned 4 minutes here, and the printer still hadn't printed. I made the call to go with a digital paint solve of as much of this round as I could rather than waste the round on technical issues outside of my control. The Digital Kakuro was a gimmick puzzle but in a way I hadn't expected, which took me a minute and a half to bash through. I then got the Arrow Maze (penalty for reading a "Z" I put in the grid to go before an "A" as a 1 next to the 6), my answer was right though) and the Figure Doppelblock in about 2 minutes each. 11 minutes, 5 puzzles, things are going great right?

Well, I skipped the Dice Skyscrapers and the NC Easy As and did the Mathrax slowly and painfully in paint, I definitely lost a couple minutes here. Then I tried the Compass for a long time without getting the final resolution correct before knocking out the Kurodoko quickly. Again though, paper would have been quicker there. I then returned to the Compass and finally figured out my mistake, and then spent 12 minutes on the Curve Data unable to resolve the top right corner. When I finally figured out what was going on there, the problem moved to the bottom right, then the top left, then the center, and I finally got it right. With 8:30 left on the clock, I had 347 points and wasn't too happy. I'd briefly tried the Battleships earlier and figured I had the choice between the 50 point Fillomino, the 55 point Pentomino Fences (which I'd also briefly started and moved on from for the Curve Data) and the 60 point Battleships. I went for the Fillomino as I knew I could get it, and did with 4 minutes left. The Battleships was my only remaining hope of points, and I managed to intuit my way to a solution with a minute to spare again! If the printer worked, I would have had time to get the Pentomino Fences too, and maybe one other puzzle. Instead, I got an okay performance. Anywhere from 400 to 700 was right though, on a good day this could easily have been a finish if I understood the Curve Data quicker- it was actually very logically fair.

The three puzzles I didn't get during the round. I definitely would have gotten 2 if the printer had just printed.

Favorite puzzles: Fillomino, Curve Data, the theme of each grid being an NxN square

773.33 (Ken Endo) / 500 / 457 (18th)

*counting continues to be the hardest thing in puzzles, as my digital counts of 26, 22, 20, 25, 27, 25, 27 don't line up with my later counts of 29, 24, 25, 28, 27, 25, 27. I blame the technical issues. But also, the puzzle is extremely unclear that the dim sum on top of the baskets are, themselves, not to be counted. So the true counts are 25, 19, 20, 22, 18, 21, 26 and I finally had a case in a competition where I was wrong but got lucky about it.

So with all 10 rounds done, I can finally see combined scores and gauge where I'm likely to end up. I can also assess how many rounds I think went well or poorly, and the answer is pretty simple: the 3 rounds I expected to go badly went badly, and the rest went well except for technical difficulties outside my control, that I still feel like I handled about as well as I could. No really exceptional performances, but nothing else under my expectations. And really, even the rounds that went badly were reasonably close to what I expected. Gripes aside, I had a lot of fun with this and would like to thank the organizing team for putting on such a fantastic event. I look forward to attending a WPC in the future, hopefully next year.

With a couple days left for some final results to trickle in, I think it's safe for me to estimate that I will end up in 6th place, behind Ken Endo, Walker Anderson, Freddie Hand (duh, the obvious top 3 participants), Hideaki Jo (who may stumble and end up in 5th, but I think my results were worse by enough that I'll lose the spot by about 100 points) and one other person. Some other people may post a set of fantastic ending results- Kota Morinishi, Phillipp Weiss, Michael Mosshammer, Roytaro Chiba, and nyoroppyi, for instance, have a shot- but I'm virtually assured of a top 10 spot here.

The remainder of this post was written after results were finalized. I ended up in 8th.

0. First place of all rounds (8675.16)
1. Ken Endo (8369.2)

2. Walker Anderson (7647.8)

3. Freddie Hand (6837.8)

"3.5". Finish all rounds, no bonus (7340) (6720.5 with Made in China set to top score)

4. Hideaki Jo (6024.8)

"?". Tenth place of all rounds (5966.16)

5. nyoroppyi (5916.2)

X. Jeffrey Bardon, but with a working printer, worst case estimate (5827- +Dice Skyscrapers, rest the same)

6. Ryotaro Chiba (5819)

7. apiad (5796.8)

8. Jeffrey Bardon (5787)

9. Kota Morinishi (5691.7)

10. Neil Zussman (5427.3)

My estimate was a bit off, but not by that much- apiad absolutely killed round 3 with the highest score on it to barely nip 7th, and Ryotaro also barely managed it by a single puzzle with multiple strong results. nyoroppyi hadn't done enough for me to be sure, but I definitely expected one of that group to pass me.

It turns out that I did lose 2 spots to my printer issue on round 10, as even if I only got one more puzzle out that would have made the difference for those spots. That's really kind of annoying. I could also point to any number of other things that I did suboptimally and could have gotten a better result, but that would miss the point- my score is the sum of the decisions I made and how well I performed. Losing points to something fully outside of my control stings a bit, but losing a placement in a concrete way* really stings. How can I be so sure that the printer failure actually cost two places? Well, I spent some time slowly working on the Pentomino Fences in paint, and got about halfway through the logic. I never made my way back to it. I also lost over a minute doing the Mathrax in paint compared to pencil, and briefly tried the Dice Skyscrapers since it looked approachable but aborted because I could tell a paint solve would not have been worth it. And then, of course, the overhead of setting up paint and putting puzzles into it vs. just having the papers! All of that adds up to about 5 extra minutes that were directly lost from dealing with paint. My post-competition solving of the other 3 puzzles I didn't get, all of which were greater valued than the score gap, all took less than 5 minutes. The Pentomino Fences especially was my most likely target, and being able to clearly and quickly mark the outside cells made the next step clear at every point. My point is then that it doesn't really matter how the decisions I would have made during the would have changed- I have time for another puzzle, I get another puzzle, and that's the difference. (You could even make the argument that I decide not to go for the Curve Data, and then get all 3 for about +60. In no world does the printer alone cost 5th though, but I do find it interesting how if I'd just looked at every puzzle in round 3 that would have done it. It's been so long since I've had to try to plan for a round that's not even close to finishable that I forgot that taking a couple seconds to assess everything is well worth it, which is a mistake I won't be making in the future.)

*The Puzzle Grand Prix 2020 had most of its rounds extremely poorly timed with the rest of my life at the time and I performed relatively poorly on about half of them as a result. However, I can't know to what extent that affected me, only that it did, so I don't see any reasonable "what if" scenario the same way I can with the Shakashaka answer keys this year (cost me a placement, but my fault) or the printer issue for this event.

Honestly though top 10 is extremely good and I should be more focused on that than the potential for top 5, or the few things that went awry. I suspect that removed from the specific events, I'll look back on this competition in that way. Especially because I also beat every other US competitor except Walker, who's just insane anyway. (Thomas Snyder and Palmer Mebane did not compete- I've actually never competed with Palmer, and I did beat Thomas in the GP this year). But at least right now it's hard to look past that one random event that has nothing to do with puzzles and how it impacted things.

Results aside, I've looked back through the entire set several times at this point and nearly every puzzle is extremely high quality, with either an excellent theme or logical path, or even both! I have a lot of plans on my plate but I'd definitely like to try to make a best of WSPC series to post later- unfortunately probably not until January though. Would people even be interested then? The conceit of each round was also either excellent or at the very least interesting, even if it didn't land for me. The overall convention experience was also a huge hit, and hanging out in the Zoom chatroom after the closing ceremonies was a blast even if I was restricted to just typing for most of it. Thank you again to the entire organizing team for a wonderful event.

See you in 2022.