> The goal is to maximize the amount of guesses it takes to find the word
The heuristic used is not optimal for this. You want to choose the set that takes the most number of guesses to split, which is not necessarily the set with the largest number of elements.
It's a cool idea, but it'd be nice to show which letters have been eliminated like "real" Wordle. I don't think not having them makes the game harder, just more tedious.
See also qntm's Absurdle, which similarly tries to be a maximally difficult by taking your guess at each stage and using it to prune the list of possible answers, picking the largest set consistent with your guesses so far each time.
This uses different word lists than Wordle/Adversarial Wordle, but still has a 4-guess optimum, no 3-guess is possible.
My solver's found about a thousand so far, here's a good one: "abyes choup donut dingo" (note that there are a number of 'fake' words allowed for guesses).
The solution partitions the solution space exactly the way I want it to: it was designed for the site this thread is about. Obviously it doesn't work against an arbitrary adversary.
My solution is static and doesn't take any clues into account, so what you're saying just doesn't feel very relevant. The only "solution in general" to an arbitrary adversary, if you're not allowed to look at the clues it gives, is just to guess every single target word, which isn't very interesting.
Now, you could easily argue that the greedy strategy used by the site is suboptimal, and I'd agree, but AFAIK nobody's put up a site with a better one yet.
Ah yeah a friend sent that to me right after I created my version! I remembered making an “evil hangman” in class many years ago so probably a lot of folks will have the same idea.
I'm not sure if assuming malice makes sense. If the OP of this post was busy coding they'd likely have missed the earlier post. In fact I'd guess that except for some people most don't follow HN that religiously.
Side note, why might someone want to earn karma on HN? I know people on reddit sometimes do it to sell their accounts but I'd imagine it to be much more pointless here.
I wonder what is the best strategy over a long distance?
“ALONE” and “SHIRT” cover 1st to 9th and 11th most common letters in English dictionary but then, you'll be getting more yellow letters in first two words which might not be the best approach (compared to completely eliminating 10 letters altogether).
Then, for words like “_OUCH”, it might be optimal to come up with a word containing as many potential first letters as possible (while omitting letters already in use, “ouch” in this case).
So, there's another evil Wordle thread on Hacker News today (https://news.ycombinator.com/item?id=29862597) and over there, I've been refining my algorith for optimal play. Here's the current approach:
R A I S E (0 green, 0 yellow, 168 words remain)
B L U D Y (1 green, 0 yellow, 13 words remain)
C O U N T (2 green, 1 yellow, 2 words remain)
V O U C H (4 green, 0 yellow, 1 word remains)
P O U C H (5 green, 0 yellow, 0 words remain)
But what was interesting was at one point of development, it did this:
L A R E S (0 green, 0 yellow, 576 words remain)
T O N I C (1 green, 0 yellow, 50 words remain)
B O O D Y (4 green, 0 yellow, 5 words remain)
D E G U M (0 green, 2 yellow, 1 word remains)
G O O D Y (5 green, 0 yellow, 0 words remain)
See, the "BOODY" has 4 greens, but then it went and guess "DEGUM" to eliminate a lot of possibilities!
One additional feature I was looking at for a personal implementation was picking the target word- how is it being chosen from the remaining letters after the player's first guess? Are you using a word that purposely uses fewer common letters from the set of remaining words?
No, I'm looping through all possible guess and targets and grouping possible targets by matching scoring. Then I select a guess on which one has the smallest set of possible words afterwards. For example, "SERAI" seems to be a good first word. The reason why is that even if the adversary selects the worst-case scoring of (in this case, it would be no green, no yellow), there's only 168 remaining words it could be.
Paste to debug in 12 guesses. Had never heard of the original wordle before, cool game.
wwwwY Paste
wwwww Color
wwwww Print
wwwww Loops
Ywwww Bombs
RRwww Death
RRRww Debts
RRRww Debby (It should be obvious here that I didn't know how the game worked)
RRRww Debit
wYwww Junks (Questionable that this is counted as a real word imo)
YwwwY Gorge
RRRRR Debug
This is fun! Though when I dove right in I mistakenly thought that the indicator for the correct letter was also indicating the correct position for that letter. But that was not the case. I had a weird combo and had to give up before I understood that.
My kid is now trying to solve it, keeping careful notes of letters. I explained what (I'm guessing) makes it evil, but she's undeterred. (Solved in 15)
22 guesses. Seems like the trick is to select letters in their worst positions first so that it later puts them in more favorable positions (e.g. don’t guess Q in the first spot if you can put it anywhere else first.
I keep pull down refreshing the page and can't go back to what the word was. Also cataegories would be nice. Also a list of the alphabet at the bottom would cross out wrong guesses would be nice.
I think there's also the question of whether a word adopted from Danish by an international standards body in the 20th century automatically counts as English.
Even if we discard obligatory quotes, there's still the hyphen as a part of stand-alone version of this "word". So when I use "femto-" by itself in a sentence, it's at best 6 characters long.
The heuristic used is not optimal for this. You want to choose the set that takes the most number of guesses to split, which is not necessarily the set with the largest number of elements.