That was fun! It's like a golden age for online word games.
You probably want some feedback. I guess most of the other one-a-day games are a little smaller/less daunting. (I wonder if it would make sense to start this game with a smaller grid, and then offer a larger grid if you solve the smaller one, or not have the bottom rows appear until you solve the higher rows.) Also wondering if you can write a program to auto-generate these puzzles for you.
I eventually got all the words. (Although there was one word that punctured my confidence just a bit because it was hard to figure out where it started!) Maybe I needed to read the instructions more carefully.
But I was glad all the words were normal words. I think one thing that makes Wordle better than some of its clones is they really do restrict themselves to very commonly used words -- so even when you're stumped, you have faith the final word will in fact be a word you've heard of.
This seems like a game I'd be good at, but not really. Could only manage to find 4-letter words, which made me think I'll get to a point after 20-30 minutes where things don't add up.
After playing around for a couple minutes... I don't quite get it. Are 100% of the letters in the grid supposed to be used in words? Or is the goal to find the solution where the maximum number of letters are used in words, but some letters will remain unused? Or just some arbitrary pre-selected set of words?
Each daily puzzle has a single solution that uses 100% of the letters in the grid. It will let you select dictionary words that are not part of that solution, and then un-select them later.
The final goal is indeed to use all letters. If you just go around choosing words, let's say, at random, you'll end up with places where you cannot use some letters. Then you will have to unmark some "wrong" (valid but not the ones in the searched solution) words and try to find some other combinations that don't leave unused cells.
Note that a "score" is calculated when you reach the 100% filled solution. This score penalizes marking "wrong" words and going back and unmarking them. So if you want to not only solve the puzzle but get a good score, you need to be careful not to go too fast adding just whatever word you find. Instead, try to build words which don't leave isolated letters, e.g. start in a corner and build from there, or from the bottom, etc.
I just added this to my list of daily word games. It's a short list; most of the onslaught in the past couple months haven't had staying power, but this one is fun!
Intimidating at first as you struggle to get a foothold, but after a couple words are found, it becomes a fast middle game, then a difficult endgame where I hunt for variations of the claimed territories to fit the final word(s).
If you think it's too hard, just keep at it for another 10 minutes or so. I suspect after playing a few of these boards I'll get better at spotting.
(Also, you may want to work 'backwards', starting at the bottom right and figuring out what words can possibly end in that corner)
When I revealed the solution to Puzzle 1 it had "PUNT" in a non-contiguous region. (https://imgur.com/a/GqeTUIq) That isn't selectable, and consequently the board was impossible to solve.
EDIT: Ah, I see the problem. The word is 'puncture', but the solution is highlighting the tiles in a way that makes it look broken. It is solvable, but it just looks wrong.
This one has nailed a strong 'hook' for puzzle games, I think--the game appearing at first more challenging than it becomes after the initial learning.
This is a nice puzzle. I wonder, when generating the puzzle, how do you prove the solution is unique? If you take the naive approach of just fitting words to the space, there is a chance that alternative solutions could become possible by chance.
Simple example: you can never have the following contiguous string:
WORDSAMPLE
Because the solution could legitimately be WORDS+AMPLE or WORD+SAMPLE.
A brute force of all possible contiguous regions, to hunt for rogue alternative solutions, is probably feasible in this small grid, but is there a better way?
> you can never have the following contiguous string: WORDSAMPLE
This kind of situation actually does occur, and you can then use the knowledge that the solution is unique to conclude that neither alternative (WORDS+AMPLE or WORD+SAMPLE) can be part of the correct solution.
Because puzzles have unique solutions. This is often (but, admittedly, not always) part of the definition of a puzzle, as a specific form of a problem or a game.
If the fact that a solution is unique is given as part of the rules, then it can be used directly to rule out some classes of solution. That seems to be the case in Cell Tower.
The popularity of these types of word games around developer circles and my total disinterest in them (partially because I'm so bad at them) makes me wonder if I have a differnet brain setup than the average coder.
It's a bit slow to use with a touchpad. I'd suggest being able to drag from any highlighted letter to an adjacent letter to add it (and more) to that word and maybe being able to drag from any highlighted letter (not vital for connectivity) 'inwards' to remove it (and more) from the set (where sensible). The whole delete-and-select menu is a bit clunky, in comparison. You can always highlight invalid words differently.
The example looked nice, but I bounced almost as soon as I saw the huge puzzle. I don't have time to work that out. There should be smaller ones. Half that size (or less) as an option.
28 comments
[ 2.1 ms ] story [ 61.7 ms ] threadYou probably want some feedback. I guess most of the other one-a-day games are a little smaller/less daunting. (I wonder if it would make sense to start this game with a smaller grid, and then offer a larger grid if you solve the smaller one, or not have the bottom rows appear until you solve the higher rows.) Also wondering if you can write a program to auto-generate these puzzles for you.
I eventually got all the words. (Although there was one word that punctured my confidence just a bit because it was hard to figure out where it started!) Maybe I needed to read the instructions more carefully.
But I was glad all the words were normal words. I think one thing that makes Wordle better than some of its clones is they really do restrict themselves to very commonly used words -- so even when you're stumped, you have faith the final word will in fact be a word you've heard of.
A smaller first puzzle would be nice.
This seems like a game I'd be good at, but not really. Could only manage to find 4-letter words, which made me think I'll get to a point after 20-30 minutes where things don't add up.
The final goal is indeed to use all letters. If you just go around choosing words, let's say, at random, you'll end up with places where you cannot use some letters. Then you will have to unmark some "wrong" (valid but not the ones in the searched solution) words and try to find some other combinations that don't leave unused cells.
Note that a "score" is calculated when you reach the 100% filled solution. This score penalizes marking "wrong" words and going back and unmarking them. So if you want to not only solve the puzzle but get a good score, you need to be careful not to go too fast adding just whatever word you find. Instead, try to build words which don't leave isolated letters, e.g. start in a corner and build from there, or from the bottom, etc.
Intimidating at first as you struggle to get a foothold, but after a couple words are found, it becomes a fast middle game, then a difficult endgame where I hunt for variations of the claimed territories to fit the final word(s).
If you think it's too hard, just keep at it for another 10 minutes or so. I suspect after playing a few of these boards I'll get better at spotting.
(Also, you may want to work 'backwards', starting at the bottom right and figuring out what words can possibly end in that corner)
Wordle, octordle.com, wafflegame.net, and now this one.
EDIT: Ah, I see the problem. The word is 'puncture', but the solution is highlighting the tiles in a way that makes it look broken. It is solvable, but it just looks wrong.
Simple example: you can never have the following contiguous string:
WORDSAMPLE
Because the solution could legitimately be WORDS+AMPLE or WORD+SAMPLE.
A brute force of all possible contiguous regions, to hunt for rogue alternative solutions, is probably feasible in this small grid, but is there a better way?
This kind of situation actually does occur, and you can then use the knowledge that the solution is unique to conclude that neither alternative (WORDS+AMPLE or WORD+SAMPLE) can be part of the correct solution.
If the fact that a solution is unique is given as part of the rules, then it can be used directly to rule out some classes of solution. That seems to be the case in Cell Tower.