#108 - 24HPC 2025 recap and (practice) puzzles

I participated in the 24 Hour Puzzle Championship again! It took place right after WSPC, in Eger, Hungary. If you've not had a chance to take a look at the puzzles yet, they are available here. Definitely check them out!

This was the 22nd edition of the event, and it was my third time participating. I ranked 23rd this time, down from 13th last year. At first glance that seems like a significant drop, but with some 100 more participants this year, many of which are typically stronger solvers than myself, I'm pretty content with my result. I'd say I would have done better if I had prepared more for the event (which I hadn't much, because of WSPC), but I think that holds true for almost everyone there. 

As always, the event was fantastic and tons of fun. I especially loved doing it with this many more people. I hope we see many of them return for future editions. Unlike previous years, this year's edition came directly after a week of already exhausting WSPC solving, making the challenge of staying awake and solving at your best for 24 hours even harder. I'm glad it wasn't my first time participating - I'm sure I would have done much, much worse or potentially even fallen asleep halfway through if it was. 

I had no significant changes to my non-puzzle preparation this time around and brought largely the same food and supplies (except for energy drinks, I'm never touching those during 24HPC again) as last year. I did forget to bring my glasses to Hungary, so I was very worried about having to wear contacts for over 24 hours straight and getting very dry and sore eyes, but it didn't turn out to be much of a problem. My solving was rather different from previous years - instead of trying to go full speed on the rounds I was most looking forward to/feeling good about and pacing myself more on the remaining rounds, I went more for steady and consistent solving throughout, making sure not to wear myself out on any round. Not really by choice, but more because I didn't have it in me to up the pace anymore. I found myself wishing for the end around the 10th round already, earlier than in prior years, so I think this was a good decision. I also did not care about rankings at all throughout the tournament - I don't think I took a proper look at them until after the end. During the tournament I had only a vague sense of how I was doing.

I also authored a round this year for the first time, together with Bram and Mark, two excellent fellow Dutch authors. I'll go into more detail about our round below the tournament recap, for now I'd just like to mention that writing for 24HPC and hearing all the positive feedback on our round was an extremely rewarding experience. I hope you all enjoyed it!  

Up next is a recap of my performance in each round. If you're just looking for puzzles I wrote (both for our round as well as practice puzzles for other rounds), those are towards the bottom of the post. Feel free to skip ahead!

--------------

Round recap 

Round 1 - puzzles by Nikola Zivanovic: 390/1000 (highest score 975 / 40,0 tournament points / tied 42nd out of 134)
My opening round this year was rather weak. Last year I opened decently strong on Nikola's round with over half of the points and a top 20 ranking, so I'm not really sure what went wrong. Maybe I took a while to warm up this time, or perhaps I strategized wrong. I did have some mishaps during the round - I tried to vibe-solve the Vama, failing miserably and not giving up soon enough, and I had to erase a bunch in the Trampoline Sudoku. I enjoyed the round though - great puzzles as per usual.

Favorite puzzle: The 2nd Statue Park. Filling in the middle was still a little bit of a struggle, but the opening step was really cool.

I think Nikola does a certain type of Fillomino really well - the one this year was great fun again, as usual. 

Round 2 (continuous round) - Puzzlers Club: 655/1000 (100 / 65,5 / tied 13th)

Round 2 was our round, so I solved the Puzzlers Club continuous set instead. Last year the tournament closed with the PC sets, so seeing it this early on felt unfamiliar. I performed well, with the round being my 4th best round by ranking. I had one major hiccup during the round, repeatedly breaking the Isowatari before eventually giving up on it. As always, the round was fantastic, filled with nicely themed puzzles and interesting types and variants.

Favorite puzzle: I thought most of the puzzles in this round were excellent. If I had to choose, I'd say Circles and Two Squares and Pass Squares (Masyu). 

Notation is always a bit of a problem/time sink for me in Bunnyhop, but I'm still always happy to see it appear. Tricky puzzle this year!

Round 3 - Christian Konig & Silke Berendes: 430/1000 (645 / 66,7 / tied 11th)
We were warned ahead of time that this round would be the hardest this year, and that scoring 400 would be respectable. I did quite well, achieving my 2nd best round this year. I optimized for the classic types and only tried a few of the hybrids, figuring I would likely not get good PPM on the high-pointers in this round. The puzzles were really good, but man, they were hard! Even some of the cheapest puzzles took me a good while to get through. I think I enjoyed the post-solve more than the actual competition round because of it - I was able to appreciate the puzzles a lot more without the pressure of the clock. One minor remark I have for this round is that most of the puzzles were printed uncomfortably large - for future years I think it would be better to make them a tad smaller.

Favorite puzzle: Both Rollercoasters. I post-solved both - I did not feel comfortable going for them in the round because the example for the classic felt very hard, but the competition puzzles were very smooth and enjoyable.

Some puzzles required keeping track of several different types of notation. I was worried about using a permanent marker during the competition, but fortunately I didn't make any errors. 

Round 4 - Team Japan: 620/1000 (1000 / 62,0 / 23rd)
This round, prepared by members of the UTokyo Puzzle Club, had a really interesting variety going on. I wasn't sure how much I would like it, but it turned out to be very enjoyable with cool puzzles throughout. I was rather amused to see Stitches make an appearance. It popped up in the Dutch championship as well this year, but I think besides that puzzle and the one on my blog I've never seen any handcrafted ones. Performance-wise this was a decent round with clean solving and no mistakes for me.

Favorite puzzle: The Tren (during the round) and the Countries (from post-solving).

Another example of several notations, with unshaded connectivity causing trouble in this case. I wasn't confident enough to use a marker in this one, and didn't find another way of notating that didn't confuse me, so I was praying I'd be able to solve the puzzle without marking unshaded cells. Thankfully it was small enough to manage.

Round 5 - Wei-Hwa Huang: 510/1000 (985 / 51,8 / 19th)
Oh how I was dreading this round. I figured I'd probably do well if I just went slow and steady, but it still scared me mightily. I started with a few of the low-point classic non-Sudoku types, then spent the rest of the round attacking the Not-Quite-Sudokus since I thought it would be easier to get momentum on those. I solved all except the 2 Nonconsecutives, but unfortunately made an error in the large Antiknight, which also took me the longest. During the round, my strategy devolved into just immediately guessing whether rows/columns/boxes would be false since it turned out much faster than trying to do it logically. It did lead to me erasing/restarting quite frequently. I don't think this was one of my favourite rounds, but I also didn't dislike it like I thought I would. Same minor remark as round 3 - the non-Sudoku types were printed way too large for me.

Favorite puzzle: The big XV.

60 points and a lot of time wasted on this puzzle, unfortunately.

Round 6 - Tawan Sunathvanichkul: 600/1000 (985 / 60,9 / 30th)
Tawan's set is always one I'm looking forward to. I was glad to see a Word Search again, but was less excited about the lower points. Still though, as usual I started my round with it. I also timed myself again and was quite pleased to register a time of 3.36. Unfortunately the rest of the round did not go so well - I restarted the 100-point Mastermind Sudoku halfway through and still submitted with an error, took forever on the 2nd Tumbleweed Loop and could not for the life of me find the eighth difference in the Twins. Nonetheless, undoubtedly one of my favourite rounds once again.

Favorite puzzle: Of course the Word Search did not disappoint. I also unexpectedly really liked the Pentominous XLs.

Quite a few of the puzzles this round had a 3D element to them, the Word Search included. Really cool spin to put on word puzzles!

Round 7 - Team China: 500/1000 (745 / 67,1 / 11th)
Much like the Japanese round, the Chinese round also had an interesting spread and a bunch of very unusual puzzle types. I scored exactly half of the points, which was enough to get me 11th place, my 2nd highest ranking. Pretty good, despite wasting a large amount of time failing to solve both the Tontonbeya (last-minute guess gone wrong) and the Hazy Simple Loop (gave up after failing to vibe-solve and erasing the grid for what felt like the millionth time). Thankfully, much like in WPC 2024, the number placement puzzles felt a bit overvalued, allowing me to recover decently. Overall a very solid round, filled with many great puzzles. 

Favorite puzzle: I have a few - the NIKOJI was really well made, I thought I would very much dislike Intermingle but I loved the competition puzzle, and (I never imagined I would say this) the U-Bahn was also really good. 

Yin-Yang using Stained Glass type clues (like the popular genre recently Kurarin) was really nice, I'd love to see more of these.

Round 8 - Puzzle Duel: 565/1000 (910 / 62,1 / 20th)

In terms of selected genres, this round looked like it might be quite enjoyable. This turned out to be mostly the case, except for a few puzzles that missed the mark. The first Sudoku Minus One seemed to have no reasonable logic, and I needed a deep bifurcation to get through the large Full Masyu. Other than that the puzzles were fine, and I had a good time working through them both during the round and during my post-solve, though fewer Arithmetic Squares would have been nice. My performance definitely dropped from not managing to secure any points for the first Sudoku Minus One despite spending well over 15 minutes on it, other than that I think I did alright.

Favorite puzzle: Probably the classic Kurodoko.

After flipping to this page, I doubted for a while if I wanted to try the puzzle - it looked like it might give me a headache. I'm glad I did, the solve was smooth and brought me some good PPM.

Round 9 - Team PAKO: 535/1000 (1000 / 53,5 / 32nd)

The first of the Turkish rounds. Overall I found this round pretty enjoyable. I had an error in the 45-point Star Battle after putting 3 stars in one row, but even with those points my result would have still been pretty unremarkable. I guess at this point exhaustion was catching up with me, even though we were only just past midnight at this point. I also started feeling like solving stuff I wouldn't usually go for, which in this round turned out to be the word puzzles. They bagged me a very welcome 140 points, but probably took way too long to solve.

Favorite puzzle: The O'utcast and the first Araf.

I'm grateful the top right cell did not get marked as an error, despite the checkers circling it with a green pen, indicating they did see it.

Not so lucky with this solve, though. The pen/pencil combination suggests I guessed the end and didn't check carefully enough.

Round 10 - Sinisa Hrga: 375/1000 (840 / 44,6 / 48th)

Even though this is the most predictable round and therefore should be one of the easiest to practice for, I have to date not put in the effort to try and do better on the Sinisa round, which once again shows from my ranking. The fact that the round came at a time when the urge to sleep is the strongest for me also didn't help. Usually the number placement puzzles allow me to salvage some points, but this time I wasn't able to make much of a dent in the Diagonal Irregular Sudoku and gave up on the Big Bands Sudoku. Maybe I'll practice a bit for next time, but honestly, probably not.

Favorite puzzle: The Zig Zags were fun.

I wasted some time counting the dominoes in the bank and the cells in the grid after noticing 6-6 was missing from the bank. I guess this is technically a broken puzzle?

Round 11 - Team Türkiye: 610/1000 (845 / 72,2 / 10th)
The 2nd Turkish round turned out to be my best this year, by ranking. It definitely didn't feel that way at the time. I guess in the night rounds performance drops across the board. I'm certain mine also did as I was feeling spent at this point, but maybe the experience of having been through this a few times now still played in my favour. Having an efficient bash through the 75-point Sudoku also helped. I did fail to solve the valuable Star Battle again, just like in the first Turkish round. 

A little anecdote: there was a large Scrabble with a bunch of puzzlers' first names as the word bank, my own included. After the booklets were returned, Ken walked up to me and laughingly said "Sorry for forgetting you". He pointed at the Scrabble, which was correctly solved apart from having forgotten to put in my name. Sorry I cost you those points!

Favorite puzzle: The 2nd Skyscrapers (Thermo).

I made a wrong deduction in this Futoshiki using pen and had lost my correction tape, so I decided to redraw the grid. I have no idea what I was thinking redrawing the grid as individual boxes rather than just drawing a standard 6x6 grid. I'm glad the checkers didn't mind.

Round 12 - Matej Uher: 590/1000 (1000 / 59,0 / 27th)

I thought this round would be one of my favorites going in - full of genres I like as well as mashups which are always good fun. Unfortunately two of the number placement puzzles turned out to be broken, and I spent a very long time on one of them. I wasn't sure if it was my sleep-deprived brain or the puzzle though, so I eventually wrote "Is this broken?" and moved on. When I got my booklet back, tragically "Sorry" was written next to it. I could have definitely accumulated more points in that time, but I'm sure that goes for more solvers. The rest of the round was stellar though, and I especially loved all the mashups. Finally realizing the trick to the Fuzuli during post-solving after not being able to make a dent in it during the round was also a rewarding moment.

Favorite puzzle: From the combos, the Region Division Combo. From the standalone puzzles, the FiveCells.

I erased this puzzle so many times eventually the ink started fading.

Round 13 - Prasanna Seshadri: 535/1000 (1000 / 53,5 / tied 15th) 

I always look forward to Prasanna's puzzles, and they once again were great this year. I particularly enjoyed the Heavy Dots and Border Blocks. There's just something fun for me about going for funky genres in the last rounds, when my energy level is at its lowest. I also wanted to try the big mashup, since Prasanna usually writes one and I always regret not trying it during the round. I did, but I wasn't able to solve it (not even close) despite looking at it for quite some time, so it probably wasn't the best choice. The post-solve was lovely though.

Favorite puzzle: The 2nd Heavy Dots and the big Loop mashup.

Exactly the kind of weird stuff I want to solve after 22 hours. Delightful.  

And that's that! Plenty of stuff to be happy about, but also a lot that could have gone better. I consider this year a bit of an outlier due to the packed schedule, so I don't think there's much point in analysing my result much further. As always, a big thanks to the organizers, authors and participants for putting on a great event again! I'm already looking forward to next year, hopefully I can return as both a solver and an author!

----------------

Dutch round

As already mentioned, we also authored a round for this year's tournament. Our round consisted of 30 puzzles following a classic+variant structure, with the variant focusing on enlarging some elements of the puzzle, whether cells, clues or parts of the solution. I loved hearing what people thought of the set (overwhelmingly positive reactions!) - thank you to everyone that took the time to share their thoughts with me. 

We started working on our set in April, and after a short brainstorm quickly landed on this theme. Our main reasoning was we wanted to keep variations and rules simple to understand, considering how packed the schedule already would be with WSPC and how unlikely it was that people would be very well prepared for 24HPC. It's also a very versatile theme - the "Large X" variant can be applied to many different genres in many different ways, allowing us plenty of flexibility with what we chose to write. We also liked that in some cases the variant serves as a natural extension of the rules and doesn't change all that much, while in others it completely changes the way a genre solves. Lastly, inspiration was also not hard to come by, with puzzles with enlarged elements having appeared in many different places and tournaments in the past.

We initially wrote 16 puzzles, after which we evaluated if we wanted to do something else for the remainder (like smaller cells as in Tight Fit puzzles, or a different theme altogether). We decided to continue with the main theme, eventually completing our set with a few days to spare. The round ended up not having any real high-pointers - this was not as much a deliberate choice as it was a consequence of what we ended up writing, but I do like the way that turned out. We were a bit worried about the length of our round, especially as our testing also suggested that it might be finishable for the top solvers (which it wasn't supposed to be), but the maximum score turned out to be 825 - perfectly calibrated. All in all, I'm really proud of the set and think it turned out great.

On an editorial note: I definitely have a newfound respect for editors of puzzle competitions - I poured an endless amount of hours into writing the booklets, formatting the puzzles and calibrating the points, including learning two new types of software. Still though, seeing how well the set was received, I'll gladly do it all again next time. 

Below are the puzzles I wrote for the round, along with some comments on them. Note that some are presented differently in Penpa+ from the booklet. Do definitely check out the full round (using the link at the top of the page), the puzzles written by Mark and Bram are really really good!

Linesweeper - 15 points - https://tinyurl.com/2yj67rwj

Written to be a gentle opening low-pointer. For a long time this was to be the first puzzle in the round, until the Akari was written, which we deemed a better fit.

Linesweeper (Large Clues) - 25 points - https://tinyurl.com/2debabkw

The first variant puzzle I wrote. Make sure to count the overlaps of adjacent clues correctly!

Pentominous - 25 points - https://tinyurl.com/223qspzy

My most strongly themed puzzle this year. I think it came out really nice. It's the only puzzle on a non-square grid, so I did have to mess with the page layout a lot to fit it in.

Pentominous (Large Cells) - 25 points - https://tinyurl.com/2yzwbyog

One of the last puzzles I wrote, and I think if I had more time I would have probably done something else with it. I was looking for a way to theme a grid around the number 25. It worked out, but it's barely noticeable (and I am against highlighting purely for aesthetic reasons). I also had a wrong step which made the puzzle a lot harder than it was intended to be. I ended up adding two extra clues (R2C5 and R9C6) as I was out of time to rewrite it entirely, but that had the opposite effect and made it easier than I wanted it to be. 

Anglers - 35 points - https://tinyurl.com/26h5a7us

I mostly focused on a simple topology idea. It worked out, but can be bypassed. Still, I'm pretty happy with this puzzle. 

Anglers (Large Fish) - 35 points - https://tinyurl.com/2affgqg5

The Anglers were last minute additions to the round. Mark had written the Akari with large bulbs, Bram the Star Battle with large stars, so I was looking for another variation on the enlarged elements. I suppose technically this can be listed under Large Clues as well though. I love how different this solves from normal Anglers, with parity becoming much less important. Small detail: I knew this puzzle would require SVG editing so I made the conscious choice of not including any vertically oriented fish in the puzzle, to avoid the trouble of having to find/make a suitable icon. I know, I know, I could have just made the cells large and kept the fish normal size, but that wouldn't have stayed true to the theme :)

Arrow Flow - 25 points - https://tinyurl.com/256zzt7k

 Not too much to say about the classic Arrow Flow - I think it's a cute puzzle, written around a break-in idea I had lying around for a while. 

Arrow Flow (Large Clues) - 50 points - https://tinyurl.com/287wgmc6

My main reason for going for Arrow Flow was wanting to have a genre that seemed very incompatible with large clues at first glance. As a consequence, making the puzzle unique took a very long time across many failed attempts, but it's probably one of my favorites of the set.

Canal View - 40 points - https://tinyurl.com/yq85q4fc

I focused more on having an atypical break-in, after which connectivity drives most of the solve. Also one of my favorite puzzles of the set.

Canal View (Large Clues) - 50 points - https://tinyurl.com/2bquu9zb

One of my first genre choices, although I don't think the Large Clues variation worked quite as well as I had hoped. It felt like it could easily turn in to a tedious counting puzzle, but to avoid that many of the large clues get completed before you even have to think about them. 

Slovak Sums - 30 points - https://tinyurl.com/22b7mfsz

I didn't want a break-in revolving around minimality/maximality, which I feel usually is the case in Slovak Sums. Instead I focused on determining which cells are used and unused, and I think it makes for an interesting little puzzle.

Slovak Sums (Large Clues) - 65 points - https://tinyurl.com/2673xnr5

I needed a number placement set in my genre selection and was debating for a while to write a Sudoku, but wasn't feeling any of the variants I could find that I thought might go well with enlarged elements. I landed on Slovak Sums instead, which worked perfectly. This is the most valuable puzzle in the set, though I don't necessarily think it is the most difficult.

--------------

Practice puzzles

To close out the post, here are a few practice puzzles. I didn't think I was going to write any since all my preparation time went into WSPC, but I found a few hours of free time on the day between WPC and 24HPC and managed to write 3 puzzles. For how hastily they were written, I think they came out alright. I particularly like the Bhai Bahan. Rules are in the links or can be found in the IB.

R7 - Inequality - diff. 2/5
https://tinyurl.com/27cumbra

R6 - Pentominous XL - diff. 2/5
https://tinyurl.com/26epotwl

R13 - Bhai Bahan - diff. 2.5/5
https://tinyurl.com/5995cmxc