|
mdemone posted:It's just all these loving ones mdemone posted:I refuse to not get No Assist
|
# ? Jan 15, 2022 22:40 |
|
|
# ? Jun 13, 2024 05:48 |
|
I didn't bother with this on S1 for whatever reason. I had just finished Mario Picross which is hardcore and I needed a break, but I'm doing Picross on Sega-cross with No Assist. Probably not Mega-Picross because I still barely get it.American McGay posted:Yeah Picross is the ultimate "fresh eyes" game. Sometimes if I'm stuck on a puzzle I'll take a screenshot and then post it on Twitter and look at it on my phone. You'll start seeing stuff in 20 seconds that you were stuck on for 20 minutes. I love those puzzles where there's only one path forward, but when you finally find it, it's like dominoes tumbling down. It's rather satisfying. mdemone posted:It's just all these loving ones I can't visualize things well so I use tons of 'x's to get things done. With tons of ones here may be some that are actually already a whole row. You may not know the number combos just yet but you'll get there.
|
# ? Jan 15, 2022 22:47 |
|
mdemone posted:I refuse to not get No Assist You can always go back and redo them later without assist to get that clear marker, but I understand that variety of brain worms. In that case, the most important thing is to learn to count. As an example if you’re playing a 10x10 and you have two rows 5 1 1 and 1 1 1 2 1: 5 (1) 1 (1) 1 = 9, because each one must have a space between. Mentally I call the remainder the variance in my head, but call it whatever. You’ve got a variance of 1 - this means everything can be shifted 1 square forward and still be solvable. With that information you know that 4 of the blocks in that 5 have to be in positions 2-5. No matter how you place it, the only valid solutions require those 4 to be filled, so they’re guaranteed. The second row 1 (1) 1 (1) 1 (1) 2 (1) 1 = 10, this row can be completely solved. So finally if you’ve got a row that’s just 6, we have a variance of 4 - start counting from the first block of the row. Ignore the squares up to 4, that’s your variance. Blocks 5 and 6 must be filled no matter what. Other useful stuff - if you’ve got a line that starts with 1 and the second block is filled, you can x the first block. Seems obvious, but the same thing applies to bigger numbers. If your row is 6 2 and the 7th block is filled your line is solved - the first block can’t be filled because it would cause you to fill 7 in a row, so it has to be x 6 x 2. Eventually you’ll start to see a lot of the other patterns that indicate something absolutely has to be filled or absolutely has to be X, but that should be enough to get started on getting concrete information right out of the gate.
|
# ? Jan 15, 2022 22:49 |
I finally got it (it's #14, the ocarina), and now I'm so mad I gotta cool off. Edit: okay maybe just one more mdemone fucked around with this message at 23:27 on Jan 15, 2022 |
|
# ? Jan 15, 2022 23:22 |
|
Vegastar posted:You can always go back and redo them later without assist to get that clear marker, but I understand that variety of brain worms. It's weird how this is exactly the way I do it, and is likely how all other Picross players do it yet the method always at least feels just my own. Maybe because I initially logic'd that method out on my own?
|
# ? Jan 15, 2022 23:30 |
Rupert Buttermilk posted:It's weird how this is exactly the way I do it, and is likely how all other Picross players do it yet the method always at least feels just my own. Yeah this is very similar to the logic trees in sudoku, which is something I'm very familiar with, so it's scratching that same itch.
|
|
# ? Jan 15, 2022 23:37 |
|
American McGay posted:) which is basically like a form of trial and error. I am calm.
|
# ? Jan 15, 2022 23:48 |
|
How many boards are in these Picross S versions?
|
# ? Jan 16, 2022 00:39 |
|
KingSlime posted:I did not realize you could quit halfway thru a puzzle and it'd save your progress...then again I'm still working through the first one lmao Someone already said yes, but I want to point out there is one exception. If you suspend a puzzle in Clip Picross, it won't let you do any other Clip Picross puzzles until you finish the one you suspended. It's only just that mode though, I assume it's because of how the puzzle selection screen in Clip Picross is laid out.
|
# ? Jan 16, 2022 00:42 |
|
I don't like doing constraint satisfaction puzzles (sudoku, nonograms, etc.) by hand. I'd much rather write a program to solve a kind of problem and consider it a solved problem (nevermind that sudoku and nonogram solvers have also existed for a long time). I wrote a fairly efficient sudoku solver 15 years ago and got it pretty good on the first try--it could complete 5x5 puzzles in a couple of seconds on the computers of the time. Mechanically, sudokus aren't really that complicated so any puzzle that has a very large search tree simply isn't fun for humans to do, so most sudokus you encounter in the wild are fairly trivial to solve on a computer. The last time there was a few pages of Picross discussion in this thread I decided to write a nonogram solver, and I've been working on it here-and-there for the past month or so. It turns out that nonograms are much more difficult than sudoku to solve efficiently. For one, the search space is a function of the size of the grid and the number of hints, so you can have some fairly straightforward 15x15 puzzles but ridiculous 10x10 ones, just based on how many hints there are and the overall complexity of the resulting image. For nonograms it's also not enough to just eliminate candidates as you go along once you know they will cause contradictions, to be efficient about it you need to prune branches of the search tree that eventually lead to invalid solutions as early as possible--meaning there's a lot of room for optimization. Put another way, good nonograms have a lot of interplay between hints and the fact that you're building towards a recognizable image helps humans intuit the search better than a computer--with the consequence that if you do make a mistake it'll wreck you for a good while. That Twitter puzzle on the last page (#197, which I reduced to black and white) took just under an hour for my current solver to solve. Interestingly enough, I've not yet implemented this particular optimization regarding overlapping candidates: Vegastar posted:\With that information you know that 4 of the blocks in that 5 have to be in positions 2-5. No matter how you place it, the only valid solutions require those 4 to be filled, so they’re guaranteed. 100YrsofAttitude posted:I love those puzzles where there's only one path forward, but when you finally find it, it's like dominoes tumbling down. It's rather satisfying. There's definitely sudokus where you have to "guess" at some point though, or at least use advanced techniques that allow you to do things like evaluate two candidates between two spots, etc. Those are the "harder" ones, with increasingly difficult sudokus making you guess more and more times for the optimal solution pathway. Edit: Oh, sometimes nonograms can have more than one valid solution. That's always fun. ExcessBLarg! fucked around with this message at 00:59 on Jan 16, 2022 |
# ? Jan 16, 2022 00:57 |
|
Waltzing Along posted:How many boards are in these Picross S versions? 150 regular, 150 mega (same puzzles, different clues), and then however many are available for Clip and Colour Picross on top of that. So, a lot.
|
# ? Jan 16, 2022 01:08 |
|
b_d posted:i thought the elemental immunities in hell were just preset per monster type. but it's been 1,000 years so i may be wrong. farewell As I recall monster types have static resistances for each difficulty, but then uniques and similar special monsters can roll some random modifiers (cold enchanted, ghostly or similar) that stack additional resistances on top of the innate monster resistances and this can cause extra immunities. I think those ones specifically roll fire->cold->lightning or possibly cold->fire->lightning and only two immunities are allowed on any given monster, so the lightning ends up being skipped if the unique monster already rolled enough immunities before it gets there. It would just make the lightning immunity slightly less common than the other two.
|
# ? Jan 16, 2022 01:15 |
|
Rupert Buttermilk posted:150 regular, 150 mega (same puzzles, different clues), and then however many are available for Clip and Colour Picross on top of that. That's not bad for $10 at all.
|
# ? Jan 16, 2022 01:29 |
|
ExcessBLarg! posted:
I’m a little surprised that it’s able to solve at all without that - it’s so critical to the human method of solving them at least. That’s really interesting either way - please post the results, I’d be fascinated to see how it works out.
|
# ? Jan 16, 2022 01:33 |
|
That's literally the only strat I know. Look for long strings of numbers and then lay them out and check for overlaps.
|
# ? Jan 16, 2022 01:45 |
|
Vegastar posted:I’m a little surprised that it’s able to solve at all without that - it’s so critical to the human method of solving them at least. Whereas when you solve them by hand you want to delay making a "guess" for as long as possible because unwinding from a wrong guess is much more difficult. Still, it would be beneficial to do that in a computerized solver because once you reach the point when (if) you're forced to guess, you're not putting in much extra work only to find an invalid solution.
|
# ? Jan 16, 2022 02:24 |
|
American McGay posted:That's literally the only strat I know. Look for long strings of numbers and then lay them out and check for overlaps. Yeah, same. 10x10 and you've got a line of a single 6? I know I have to assume 4 blanks on each side, but everything else is guaranteed. I literally don't know how else to logic that out.
|
# ? Jan 16, 2022 02:25 |
Rupert Buttermilk posted:Yeah, same. 10x10 and you've got a line of a single 6? I know I have to assume 4 blanks on each side, but everything else is guaranteed. Middle two are fills. No possible string of 6 fits in 10x10 otherwise. Then look up and down from there
|
|
# ? Jan 16, 2022 02:39 |
|
ExcessBLarg! posted:Edit: Oh, sometimes nonograms can have more than one valid solution. That's always fun. One of the final puzzles in Murder By Numbers had a second valid solution, but it of course doesn't count for how the game wants you to solve it. It makes sense that it has to be a certain way, but I feel like they probably should have made sure that it only had one intended solution. Anyway in general you rarely have to just guess in the Picross S games. There's practically always a logical flow and a way to figure out what goes where, even if it's really hard to see (like the puzzle I posted about earlier in S4). Sometimes the way to logic it out isn't directly, though, you have to think ahead a couple of moves (which is why they have the button that marks a square without filling it in).
|
# ? Jan 16, 2022 04:17 |
|
mdemone posted:Middle two are fills. No possible string of 6 fits in 10x10 otherwise. Then look up and down from there ah, thanks.
|
# ? Jan 16, 2022 04:20 |
|
ExcessBLarg! posted:So I do the "counting" thing where you add up the size of the hints and the required blanks to figure out which hints on the board have the least number of potential placements--say two "variants". At that point, with a computerized solver it's pretty cheap to make a copy of the board and walk done the solution branches for both variants, and just throw away the work for the invalid branch once you've determined that it's invalid. The problem is that many subsequent operations will be the same in both branches so you end up duplicating the effort even though only one of them will yield a valid solution. Compound that with many guesses and the work gets pretty inefficient quickly. I'm pretty sure Jupiter's puzzles are designed to be solvable without ever guessing, so a solver should be able to work through them sequentially without backtracking if it knows all the tricks I only played S1 but IIRC guessing is never required in that one
|
# ? Jan 16, 2022 04:27 |
repiv posted:I'm pretty sure Jupiter's puzzles are designed to be solvable without ever guessing, so a solver should be able to work through them sequentially without backtracking if it knows all the tricks I'm pretty sure (but I just started today), because I'm very familiar with these kind of algorithmic games, that all of them must necessarily be solvable by logic trees alone, otherwise they couldn't have been produced by the game...producer. Now, it may well be that the trees aren't easily grasped in the most difficult cases, but I still haven't seen one that doesn't conform to this type. Very similar to sudoku, in the way it is created - reduced - mapped. Edit: I'm saying that I agree with you. Sorry if that wasn't clear.
|
|
# ? Jan 16, 2022 04:35 |
The best example I can give is the following: Any sudoku, now fill in "pencil possibilities". Search all column/row intersections for uniques. Iterate until solved. All these puzzles must be created by reversing these steps without getting to a point where degenerate solutions exist.
|
|
# ? Jan 16, 2022 04:41 |
|
I want to put all my carts in some kind of book or case instead using individual cases. What's a good option for something that can hold a few dozen carts and isn't gigantic?
|
# ? Jan 16, 2022 04:57 |
|
DR FRASIER KRANG posted:I want to put all my carts in some kind of book or case instead using individual cases. What's a good option for something that can hold a few dozen carts and isn't gigantic? Store them in your cheeks, like a chipmunk.
|
# ? Jan 16, 2022 05:09 |
|
DR FRASIER KRANG posted:I want to put all my carts in some kind of book or case instead using individual cases. What's a good option for something that can hold a few dozen carts and isn't gigantic? Superglue some little pouches to the back of your switch like that one guy did to his 3ds. Had to go find the picture. Vegastar fucked around with this message at 05:19 on Jan 16, 2022 |
# ? Jan 16, 2022 05:15 |
|
I do not want to blow up a cow, though. Is it still a good plan if that's the case?
|
# ? Jan 16, 2022 05:22 |
|
If you drill in the right spot you can put them all on a keyring.
|
# ? Jan 16, 2022 05:23 |
|
DR FRASIER KRANG posted:I want to put all my carts in some kind of book or case instead using individual cases. What's a good option for something that can hold a few dozen carts and isn't gigantic?
|
# ? Jan 16, 2022 05:25 |
|
Vegastar posted:please post the results, I’d be fascinated to see how it works out. repiv posted:I'm pretty sure Jupiter's puzzles are designed to be solvable without ever guessing, so a solver should be able to work through them sequentially without backtracking if it knows all the tricks The biggest challenge I have with the solver is that it doesn't match up hint placements from a row (or column) with potential placements for adjacent columns (or rows) which is obviously what you want to do if you're doing it by hand. It will exclude placements that are otherwise guaranteed to introduce a contradiction (mostly blanks), but sometimes a specific box could potentially be filled by more than one hint and which one actually does isn't determined until later. Instead, what it does is wait until all placements for a row or column is complete, mark anything leftover as blank, and then check to make sure that all filled boxes were actually filled by placements for that row (or column) and if not then it knows it's reached an invalid solution. Probably what I'm missing is a step where, whenever it places a hint for (say) a row, it checks to see if any column has only one hint that could satisfy it, and in that case it places that next.
|
# ? Jan 16, 2022 05:26 |
|
This is more because my kids keep putting them back in the wrong cases and leaving them all over the place. I figure if they had a lil book of games it'd make it easier for them to keep them in one spot.
|
# ? Jan 16, 2022 05:26 |
|
DR FRASIER KRANG posted:I want to put all my carts in some kind of book or case instead using individual cases. What's a good option for something that can hold a few dozen carts and isn't gigantic?
|
# ? Jan 16, 2022 05:28 |
|
This is not a passive aggressive dig or anything: it’s pretty fascinating (and kind of impressive) that you’ve been programming a solver for picross despite apparently having no idea how to solve picross.
|
# ? Jan 16, 2022 05:47 |
|
Get a bunch of these and string them on a necklace https://www.amiami.com/eng/detail/?gcode=GAME-0019650
|
# ? Jan 16, 2022 05:50 |
|
You can solve every single puzzle in Picross S1 and S2 without guessing or thinking any more in advance than the current move. I assume the rest are the same (currently working through S3).
|
# ? Jan 16, 2022 05:58 |
|
Holy HECK did this thread not do its job. Sure y'all said Tetris Effect was a good VR experience and now it had a Switch port, but you did NOT say how engrossing the music, visuals, and audiovisual gameplay was. I spent the first half trash talking, the second half being trash talked, and the whole experience having my vision black out around the TV screen. You also didn't tell me that Colin Mocherie is old but Wayne Brady is the same age, or that Puyo Puyo Tetris 2 is out. For shame, thread. For shame.
|
# ? Jan 16, 2022 06:03 |
|
My partner got a new Switch and we went to import her account to it She didn't import her authenticator when she got a new phone The backup codes she has a screenshot for don't work Customer service doesn't open until 9am
|
# ? Jan 16, 2022 06:08 |
|
Sounds like you'll be waiting til 9am? Hope you'll be ok?
|
# ? Jan 16, 2022 06:12 |
|
Alan_Shore posted:Sounds like you'll be waiting til 9am? Hope you'll be ok? I mean of course I'll be fine. I'm not the one with no patience and a shiny new Switch they can't import their Stardew save to
|
# ? Jan 16, 2022 06:19 |
|
|
# ? Jun 13, 2024 05:48 |
|
Len posted:My partner got a new Switch and we went to import her account to it That sucks, fingers crossed for y'all.
|
# ? Jan 16, 2022 06:23 |