Wordle thread

Praguehawk

All-Conference
Aug 1, 2003
1,025
3,471
113
John Candy Reaction GIF
Sorry, I can't help that I'm so lucky at picking first words.
Wordle 1,898 2/6

🟩⬜🟨🟩🟨
🟩🟩🟩🟩🟩
 

Nittany.Lion

All-Conference
Jul 31, 2006
516
1,378
93
For the record.

The previous attempts at statistical analysis are wildly flawed.

Wordle has an answer-only dictionary (and an approved guess list with over 12,000 words) that started at about 2300 words; it has been updated over the years, and words are now being reused. If you parse that list to exclude all words with duplicate letters and include the 10 most frequently used letters I use most of the time (aeiorstnlc), you would get a list of 75-80 words.

From AI
While the mathematical limit yields over 30,000 randomized letter strings, there are roughly 350 to 450 valid Wordle-accepted (not necessarily on the answer-only list) words that can be spelled using exactly 5 unique letters from the set a, e, i, o, s, l, n, t, r, c. [1, 2]

These letters form the absolute best foundation for the game, since E, A, O, I are the most frequent vowels, and R, T, L, S, N, C are the highest-performing strategic consonants.

When excluding plurals ending in "S" and strictly filtering out verbs, the list of valid Wordle solutions (the actual hidden answers of the day, which the New York Times separates from the raw guess list) drops significantly.

Using only 5 unique letters from the set A, E, I, O, S, L, N, T, R, C, there are roughly 75 to 80 clean singular nouns and adjectives that fit the criteria.

Ai couldn't give me an answer when I asked, what are the odds if I eliminate all the Wordle-selected words that fit the criteria from my list of possible choices?

Anyway, as should be obvious now, the odds will be orders of magnitude less than noted in previous posts.
 

longliveCS40

All-American
Oct 25, 2002
2,068
8,158
113
For the record.

The previous attempts at statistical analysis are wildly flawed.

Wordle has an answer-only dictionary (and an approved guess list with over 12,000 words) that started at about 2300 words; it has been updated over the years, and words are now being reused. If you parse that list to exclude all words with duplicate letters and include the 10 most frequently used letters I use most of the time (aeiorstnlc), you would get a list of 75-80 words.

From AI
While the mathematical limit yields over 30,000 randomized letter strings, there are roughly 350 to 450 valid Wordle-accepted (not necessarily on the answer-only list) words that can be spelled using exactly 5 unique letters from the set a, e, i, o, s, l, n, t, r, c. [1, 2]

These letters form the absolute best foundation for the game, since E, A, O, I are the most frequent vowels, and R, T, L, S, N, C are the highest-performing strategic consonants.

When excluding plurals ending in "S" and strictly filtering out verbs, the list of valid Wordle solutions (the actual hidden answers of the day, which the New York Times separates from the raw guess list) drops significantly.

Using only 5 unique letters from the set A, E, I, O, S, L, N, T, R, C, there are roughly 75 to 80 clean singular nouns and adjectives that fit the criteria.

Ai couldn't give me an answer when I asked, what are the odds if I eliminate all the Wordle-selected words that fit the criteria from my list of possible choices?

Anyway, as should be obvious now, the odds will be orders of magnitude less than noted in previous posts.
What's very obvious to me is that you're a cheater. At a game on the internet. Pathetic.
 

FLaw47

All-American
Dec 23, 2010
3,813
5,224
113
For the record.

The previous attempts at statistical analysis are wildly flawed.

Wordle has an answer-only dictionary (and an approved guess list with over 12,000 words) that started at about 2300 words; it has been updated over the years, and words are now being reused. If you parse that list to exclude all words with duplicate letters and include the 10 most frequently used letters I use most of the time (aeiorstnlc), you would get a list of 75-80 words.

From AI
While the mathematical limit yields over 30,000 randomized letter strings, there are roughly 350 to 450 valid Wordle-accepted (not necessarily on the answer-only list) words that can be spelled using exactly 5 unique letters from the set a, e, i, o, s, l, n, t, r, c. [1, 2]

These letters form the absolute best foundation for the game, since E, A, O, I are the most frequent vowels, and R, T, L, S, N, C are the highest-performing strategic consonants.

When excluding plurals ending in "S" and strictly filtering out verbs, the list of valid Wordle solutions (the actual hidden answers of the day, which the New York Times separates from the raw guess list) drops significantly.

Using only 5 unique letters from the set A, E, I, O, S, L, N, T, R, C, there are roughly 75 to 80 clean singular nouns and adjectives that fit the criteria.

Ai couldn't give me an answer when I asked, what are the odds if I eliminate all the Wordle-selected words that fit the criteria from my list of possible choices?

Anyway, as should be obvious now, the odds will be orders of magnitude less than noted in previous posts.

Look, I'm interested and trying to be charitable to you here. Is your perspective basically that "all of this information is publicly available and it'd be foolish to not take advantage of all of the information available to me?" Because I can understand that.

I do have a very difficult time understanding, though. how you can beat optimal play in a very large sample size.
 
Dec 4, 2001
5,498
17,947
113
Look, I'm interested and trying to be charitable to you here. Is your perspective basically that "all of this information is publicly available and it'd be foolish to not take advantage of all of the information available to me?" Because I can understand that.

I do have a very difficult time understanding, though. how you can beat optimal play in a very large sample size.
The thing is that this is supposed to be fun and we have always had a gentleman's/gentlewoman's agreement of no cheating. I mean, what is the point of this other than to have fun? To that end, I like seeing how well other people can do and seeing how I can improve my strategy, which is clearly not as good as others. Here is how I do it:

  1. I always start with TASER on hard mode.
  2. If I have no letters yellow or green, I guess "DOUGH". Otherwise, I try to pick a second word with as many unique common letters that fit with the results of guess 1.
  3. If I somehow whiff on having any letters after two guesses, I pick "CLIMB". Otherwise, I try to pick a third word with as many unique common letters that fit with the results of guesses 1 and 2.
  4. For the par shot, this is typically where I will double up on letters such as DD in BUDDY, if it makes sense.
  5. For bogey, I usually kick one of my dogs first, and then make the best pick available.
  6. For double bogey, I kick my other dog first, and then make my pick and hope my streak does not end.
 

FLaw47

All-American
Dec 23, 2010
3,813
5,224
113
The thing is that this is supposed to be fun and we have always had a gentleman's/gentlewoman's agreement of no cheating. I mean, what is the point of this other than to have fun? To that end, I like seeing how well other people can do and seeing how I can improve my strategy, which is clearly not as good as others. Here is how I do it:

  1. I always start with TASER on hard mode.
  2. If I have no letters yellow or green, I guess "DOUGH". Otherwise, I try to pick a second word with as many common letters that fit with the results of guess 1.
  3. If I somehow whiff on having any letters after two guesses, I pick "CLIMB". Otherwise, I try to pick a third word with as many common letters that fit with the results of guesses 1 and 2.
  4. For the par shot, this is typically where I will double up on letters such as DD in BUDDY, if it makes sense.
  5. For bogey, I usually kick one of my dogs first, and then make the best pick available.
  6. For double bogey, I kick my other dog first, and then make my pick and hope my streak does not end.

I use
SLATE
for every start. I am a competitive jerk, though, so after the last tourney I paid for a NYT games subscription and have been checking with the bot and playing the archive to try and improve my play.
 
Dec 4, 2001
5,498
17,947
113
I use
SLATE
for every start. I am a competitive jerk, though, so after the last tourney I paid for a NYT games subscription and have been checking with the bot and playing the archive to try and improve my play.
I don't like using a starting word that ends with E, on hard mode, because that can lock you into a death spiral if you get a pattern like _ _ A _ E. But, maybe I need to rethink that.
 

FLaw47

All-American
Dec 23, 2010
3,813
5,224
113
I don't like using a starting word that ends with E, on hard mode, because that can lock you into a death spiral if you get a pattern like _ _ A _ E. But, maybe I need to rethink that.

I do "hard mode" voluntarily until its obviously going to be a risk to my streak. The word from yesterday, I think it was, would have been a real problem for me.
 

longliveCS40

All-American
Oct 25, 2002
2,068
8,158
113
I don't like using a starting word that ends with E, on hard mode, because that can lock you into a death spiral if you get a pattern like _ _ A _ E. But, maybe I need to rethink that.
Interesting. I use a start word ending in E and play hard mode exclusively. I've definitely put myself into the situation you describe many times. Might be time to change starting words.
 
Tiger Illustrated

Latest

  1. 01 Tiger Illustrated What the Clemson - LSU game may come down to
  2. 02 Tiger Illustrated TUESDAY NEWS CONFERENCE: Swinney names Vizzina QB starter, Lane Kiffin, LSU - NFL additions, team info & more
  3. 03 Tiger Illustrated Talking season is in its final hours in Clemson
  4. 04 Tiger Illustrated VIDEO: Tom Allen, Will Heldt, Vizzina and Olsen Patt-Henry
  5. 05 On3 What happens next for LSU, the SEC and former NFL players trying to play in college?