Average Wordle Score and Stats: What Are They and How to Find Some

You’ve been playing Wordle for weeks now, and you want to know how good you are compared to other players. But, Wordle doesn’t have a leaderboard. So, is your win streak better than average, how do you know? Well, let’s take a closer look at Wordle scores and statistics to find out.

What data does Wordle Statistics provide?

All the data you glean from Wordle statistics is related to your own progress and streak in the game (as a unit). The statistics include the total number of games played, Win %, Current Streak, Max Streak, and Guess Distribution.

  • Played is a record of the total number of games you have played. 
  • Win % shows the number of games won (W) out of the total number of games played (P) multiplied by 100; [W/P*100]. 
  • Current Streak is the record of the number of games you have been winning in a row. It is specific to the present and accounts for all the games won consecutively (without breaks) in the ongoing scenario. Current Streak is not to be confused with Max Streak.
  • Max Streak shows the maximum number of games you have won in a row (in an unbroken streak) in the game; it doesn’t necessarily be the current streak. 
  • Guess Distribution is a representation of the number of guesses you make to successfully solve each challenge. The guess distribution chart, like the other elements on the scoreboard, sheds light on your whole gaming history on Wordle and not just the current game.

The timer under NEXT WORDLE counts down to when the game will reset to give you a new challenge. The SHARE button in green lets you share your latest score.

Related: What are the Average Number of Guesses in Wordle?

What are average statistics in Wordle?

Since Wordle doesn’t have an official leaderboard or community hub to track player scores and statistics, there are no veritable means to analyze player performance on a global scale. However, unofficial and relatively small-scale data crunching from Twitter shares of Wordle scores gives significant insight into the number of guesses typically required by players to successfully finish a Wordle challenge. 

Related: What are the Average Number of Guesses in Wordle?

Wordle Stats, an unofficial source for Wordle’s Twitter score share data, has been sharing daily stats for Wordle scores by analyzing over 200k tweets every day. It is somewhere to run a rudimentary breakdown.

Their analysis shows that less than 5% of all players get the word right in 1 or 2 attempts. Players who take a total of 3 or 4 guesses to complete the challenge form the largest group on Twitter’s Wordle community. Almost all wrap up the game in 5 moves. Quite a few players manage to crack it in the final attempt. The 1% of players who fail and still share are the real bravehearts who share honest results without cheating or hiding behind the veil of pretense.

Of course, there isn’t nearly enough data to get the full picture of how players perform on Wordle, and the typical number of guesses required — the Twitter community only entails a fraction of the total number of players scattered across the globe who do not share their scores on public platforms.

That said, the unofficial numbers above could give you an inkling of your extraordinary competency if you typically solve the challenge in 3 guesses. 2 guesses (in total) translates to extraordinary in the Wordle wordle, and 1 means nothing but a thunderbolt of blind luck (or unflinching cheating skills). If you belong to the category of 4 or 5 guesses, you are an average player, maybe not extraordinary, but you are doing very well in the game. Anyone who solves it within 6 chances is a Winner, no matter!

Related: Wordle: 5-Letter Words three or four Vowels

Where to find data on your Wordle guesses?

The quickest way to track how many attempts you have been taking to solve Wordles is by checking the Guess Distribution numbers on your Wordle statistics.

If your guess distribution shows more correct guesses in 3 attempts, you’re above average. If it shows more numbers in 2 guesses, you’re in the top 5 percentile among the Wordle players. That is super good and super rare! Of course, this is all based on the data accumulated on the scores shared by the Twitter demographic.

Related: How to Enable Hard Mode in Wordle to Increase Difficulty. How Does it Work?

Should you be worried about the statistics?

In all honesty, no. Wordle is simply a game to provide you with a few lax moments away from your daily hustle and bustle. While you can certainly use the tips given above to closely analyze your score and performance to make adjustments in your future performance, there is no reason to worry about the numbers in reality.

The only person you should compete with is yourself. That is also why Wordle only shows data related to personal performance to the players — but it wouldn’t be too much of a shocker if they launched a community leaderboard, given the game’s success and mass appeal.

Can I compare my Wordle score to others?

If you’re looking for a leaderboard or other database to compare against your Wordle score, you’re out of luck. Wordle doesn’t have an official leaderboard at the moment. With that said, you can compare your statistics and wins to your friends’ by comparing your scores on a case-by-case basis.

Sharing your Wordle results on Facebook or other social media is one way to get started. Simply ask your friends to share theirs, too.



  1. Remember the stats include ONLY players posting their score on twitter. This no doubt is a small fraction of players. You will also notice that less people post their scores on the harder games. I suspect the posters for any particular puzzle have 2-3x better score than average.

    1. This was my first thought as well. Definitely skews the data. My guess is the real average is 4-5, rather than 3-4 as shown here.

    2. The only time I’ve ever tweeted my Wordle results was today, because I got it in two.

      I have never gotten it in two before.

      And the idea that 1% of people get it in one is total BS.

      1. You’re right. That is statistically impossible. Weird they didn’t point that out.

        1. I’ve changed my mind slightly about the odds of getting a ‘hole in one’. Wordle has a list of 2315 solutions – but there is also a much longer list of 5 letter words which it will accept as legitimate but which are never going to be a solution. Some words are on that second list because they are excessively obscure (vinic) and others because they are swear words I think. Enough of the excluded words are familiar to me that I think the chance of a correct first guess is actually less than 1/2315 and the chance of doing it twice in a row well under one in 5 million.

          1. Simple plurals too

      2. Quite right. I think there area around 2500 words in their dictionary. So a simple calculation suggests the chance of a “1” is 1/2500 = 0.025%. You could probably make it 0.03% if soeone kepta record of all teh words used to date. However, what the 1% is useful for is giving you an idea of the “weighting” that needs to be applied to the self-reported scores…

        1. There are 12972 words in their dictionary, but only 2309 are listed as answers. All the rest are accepted words when trying to determine the answer.

      3. I agree. It’s blind luck. By the way, is the average wordle score based on the weighted average of wins only, or does it include games you didn’t win? My take on calculating the average on all games played would be the weighted win average divided by your win percentage (i.e if weighted win is 3.90 and your win % is 96%, it would be 3.9 divided by .96, or 4.06). Other than that, I don’t know how to value games lost.

  2. The only way to get the word on the first try is pure luck/guessing. Depending how many correct letters you guessed in the first word, your second might be skill or also mostly luck. It’s only/mostly when you get to the 3rd try that skill plays a significant role.

  3. Agree that the scores of those who specifically post them to show off will naturally be skewed upward. Add to that the fact that Twitter users are natural self-promoters and it seems fair to assume that great scores will naturally bubble onto Twitter while “average” score achievers will keep it to themselves.

    1. Since there is a large list of words accepted as guesses, but which are never going to be the answer, it seems likely that the wider your vocabulary, the more tries you are likely to take, on average, as you have more tries available in your mental locker.

      1. Probably true, and we don’t know quite how Josh compiled the solution list but you soon start to learn by experience that there are probably no plurals of 4 letter words (boots) and no ‘rude’ words (penis). I have a sneaking suspicion that Josh took a list of every 5 letter word from a standard dictionary and hand picked the ones he was going to include as those that he recognised and didn’t want to exclude for the above reasons, so that you are effectively matching your vocabulary against his. I’ve just seen some one post a performance go 6 second guesses out of 35 attempts on twitter which seems highly improbable. I might try to calculate the odds later.

        1. My scores are similar to the one you saw – I’ve played 27 times, solved it in 2 guesses 4 times, in 3 guesses 9 more times. (All of those are fair, but of course you don’t have to believe me). Two caveats: I always play with my 14-year old son and we work together (one of those 2-bangers he came up with both the first and second guesses). All of those 2-guess solutions involve both luck (a fortuitous first guess that gave at least 1 green and 1 red), and some deductive reasoning. And yes, an informed but lucky stab at the second guess.

          1. 4/27 is probably quite a bit more likely than 6/35 to be honest. As the length of your streak goes up you will tend to ‘regress to the mean’. I’ve come to the conclusion that one and two guess solutions are about 100% and 90% pure luck. There are very few first word results that will really narrow down the choice for the second word to only one choice and not many that even leave it a 50/50 guess even when you get 4 greens – a classic example was ‘shake’ when people got “sha_e” but still only had a 1 in 5 chance of getting the second guess right. I think tactics can influence your distribution of 3s 4s and 5s though. People with a good vocabulary who make the most of the information available in the first two guesses should probably score more 3s than 4s, especially if they have been playing for a while and can remember quite a few of the words that have already ‘gone’. If you take an interest in wordle it’s also pretty hard to avoid subtle hints – there’s a strong argument for doing it first thing on waking before talking to anyone else or checking your phone if you want your results to be wholly unbiassed.

          2. I dont believe you UNLESS your stats got more realistic real quick after that great LUCKY beginning. I have played about 160 games ( hard mode actually smarter mode to use) and Ive only solved it on second try 6 times and thats only because of the luck of the draw ( first word choice LUCKY)

  4. Posted results are skewed.
    Most people only post their very best (and sometimes surprising) results (at least the ones who would like to keep their friends). Others cheat. Only Wordle can provide accurate stats.

  5. This approach suffers from bias of exclusion. Individuals who guess the word correctly are likely to post about it and even then, the likelihood of share increases with their perception of how well they did guessing. This misses the bulk of the low guesses and missed guesses. So the Twitter analysis is going to heavily bias towards the low end.

  6. Oh come on! First try=no skill only luck.
    Also sometimes there are multiple possibilities that you have to exhaust and you still might get it wrong. For example. I had this recently for the first three letters.
    C R A
    Had it on the 2nd try and lost!
    Crank, crack, cramp, crazy, NOPE!
    It was crass. All my guesses were available letters (of course). So according to you a wordless super genius would have gotten it in two tries or immediately would have known it was ‘crass’ ?! c’mon

    1. Once you tried crank and the k was grey, the guess of crack was partially a waste since the k wasn’t in the word.

      1. I had similar.
        U ER on second guess. Couldn’t finish.
        UNDER -2

        1. I agree that was a tough one but what was your first word?

        2. You don’t need to test each possibility. You pick a different word which combines many of the possible letters so you can eliminate 2 or 3 words.

          1. Jeff Hammersley

            Does not work in ‘Hard’ mode.


        3. Dumps or similar should have been your next guess. No chance of finishing in 3 but a much higher chance of finishing in 4. My stats are 0 1 10 17 5 4.

          1. As long as you’re not playing in ‘hard’ mode, which forces you to re-use letters you have got right. In that situation there are words which theoretically can take 14 guesses to get right, though that involves a lot of very obscure words from the scrabble dictionary which most people would correctly assume are not in the wordle solution list.

        4. But surely your best tactic after the U ER was to try a word like “SHRUG” or “PANGS” or similar that would have cut down the options?

    2. Then you played that wrong, once you had the first 3 letters, your next guesses should have been random letters to utilize as many remaining letters , stomp,drink etc

    3. I have played wordle 33 times and once got 3 green slots on the first round, viz. S?A?E. It also ended up the game with my highest score, 6, and even then I consider myself lucky because even on the 6th guess, I had multiple viable options. It turned out trying to find the remaining 2 letters while keeping the correct 3 was the wrong strategy, because there are hundreds of words that fitted the missing 2 letters. The second time I got lucky with 3 greens on the first round, I went for a different strategy, by first abandoning the 3 correct letters and trying to find words with up to 4 letters of the potential missing letters, e.g P, H, M, L, T, D, V, G, R,C, …, e.g. CHUMP, GUILT, DROVE – I then got it in 5 and would have been unlucky to get it in 6….

    4. Why would you have guessed “crack” when your choice before eliminated the K?
      Looks like your sort of beat yourself!

    5. Your second choice should have been smack. Strategy!

    6. This is my thought exactly!!!

  7. Great writeup. I couldn’t believe there existed no ranking tool out there to compare your Wordle stats to others, so I took some (read: way to much) time last week and built one. I encourage people to check it out when they can. There is some sample data in there currently, but we need as much real-world data submitted to ensure more accuracy later. Thanks!

  8. Having a leaderboard is pointless as it’s far too easy to cheat.
    The twitter stats are self selecting and therefore skewed, not surprising that no on is advertising their failures.

  9. Useful, however very wrong at one point.
    All players will have a distribution of scores. As in the data this will be around 3,4 and 5.
    If your Average is 2 then you are cheating… not in the top 8-10%, or even the top 5%.
    Depending on how many you have done you can analyse whether your average is significantly better than the average of the large number of reported scores, I suspect that most average around 4 and less than 3.5 probably puts you in the top 10% and more than 4.5 in the bottom 10%…

  10. If win % is 100 how are my streaks not equal to my games played?

    1. You may have missed a day. So that game wasn’t played so your streak goes back to 0 but you have still scored 100% on the games played

      1. THANK YOU! I have been beating my head against the wall to figure out why my 100% isn’t in my streak and you have provided the explanation! It would have been very helpful if the writer of the article ha provided that info in the “definition” section. So – thanks!

        1. It’s a useless stat. “How many days in a row have I played?” Moreover, who wants to see how “addicted” they are?! “You’ve taken methamphetamine 128 days in a row. Congratulations!”

          1. Not totally. Someone may have given up when they reckoned they would end up with a high score for the day… Ie distorting their results by not entering a score on a tactical basis

          2. Alan’s response leads me to an alternative rule. Maybe they break the streak if you miss a day but attempted to play. I missed a day a few weeks ago, never even turned on my device that day–they allowed my streak to continue.

        2. It was pointed out to me that if you didn’t care about your consecutive streak, that you could let a game lapse if you were faced with a dicey 6th 50/50 guess or worse. How about in place of the useless Streak stats which takes TWO places on the stat line

          PLAYED – WIN% – CURRENT STREAK – MAX STREAK (useless), how about:

          PLAYED – WIN% – INCOMPLETE (for games started but not finished – a dose of TRUTH!)

          1. I like that proposal. I’d even go so far as to have the Win % be measured as “games started but not completed in 6 guesses”, so the incompletes are treated as losses. It would be irritating for someone to post that they have a 100% win rate, but omit that they bailed out of 15% of their games.

    2. If you miss playing one day your current streak will start over. My played is 66, win 100%, current streak 6, max streak 53. I just forgot to play on some days

  11. It’s an interesting statistical problem isn’t it. Clearly the comment in the original post that “if you are consistently getting the word in two tries you are above average” is something of an understatement, the key to this being the word ‘consistently’. No-one, over the course of several days will get it in two repeatedly. Anyone who claims this is cheating. The wordle dictionary is said to be 2315 words so you have a roughly 1/2315 chance of a ‘hole in one’ and a roughly 1/5 million chance of getting it right first time twice in a row… IF you don’t ,like most people, have a favourite starting word. If you do always start with the same word then we have to guess whether Josh’s algorithm allows the game to pick the same word twice (haven’t seen it so far). If not then your chance of a double first guess win with that strategy is zero. I say ‘roughly’ in those figures because any individual’s vocabulary of 5 letter words will not exactly match the wordle dictionary, which in turn, is not a complete list of every 5 letter word known to English. Only the first guess is amenable to this sort of classical probability calculation unfortunately.

    The wordlestats twitter tool provides lots of data but, as others have noted, suffers from serious sampling error because of the tendency for people to tweet their ‘better’ scores. Nevertheless, as long as you understand that you are comparing your results against such a biassed sample it’s still interesting to see if you are doing better than the average twitter poster. Looking at the spread of scores for a single word allows you to see whether you did better than average on that word, but of course there is such an element of luck in this that you can’t judge your performance from any single attempt. It’s easy enough to convert those wordlestats numbers into an average score for each word though which gives us an empirical measure of how ‘difficult’ each word is. Of recent ones, PROXY seems to have been the toughest and POINT the easiest. Pooling a series of 30 or so of these scores gives an average difficulty score of 4.15 tries. If your average number of tries in a streak of 30 or so attempts in wordle is less than 4.15 then you are probably better than the average twitteratus.

    How much better is much harder to quantify. The mean difficulty scores are not normally distributed and you would have to apply some non-parametric statistics to a larger data sample I think to arrive at a proper estimate. I would be curious to know what methodology the earlier poster used to build their wordle rank calculator. If its only based on the data submitted by users then it suffers from the same sampling bias as the twitter stats.

  12. I don’t get the stats in the first example in this article: They say this player has won 100% of her 3 games, but her current streak is 1 and her longest streak is 2. Isn’t a “streak” a series of consecutive wins? If she has won 100% of the 3 games she played, isn’t her longest streak 3, and her current streak 3?

    1. Yes, they’re not describing the stat clearly. I’ve played 44 times, winning 100%, and my current streak is 31. Using their definition (“the record of the number of games you have been winning in a row”), my streak should be 44. But streak requires **uninterrupted days of play**, so the actual definition of streak is “Number of consecutive wins without missing any days.”

  13. First, you have to know the total number of people who start the game, then the %who solve it in 1-6 guesse, and, thus, wha % who don’t get the word.

  14. Not a chance in hell that these stats make sense. Consider guessing on first attempt. Pure chance. Can’t be 1/100 correct. How many 5 letter word combos out there? Two guess, also has to be rare. 1/20. No way. 3 guess is excellent and still unlikely. I played this game for years as Jotto: Achieved with three guesses is excellent but not uncommon. 4 is still excellent but people can skillfully narrow down options in first three attempts and get 4 regularly. My guess is the actual stats are more like 0%, 1%, 12%, 30%, 40% 12%, not discovered 4%.

    1. The twitter stats are presumably rounded up – so anyone getting it first time will make it appear like 1% even if its only one out of 300k. The actual chance of getting it right first time is about 1 in 2315 so it should be about 0.05%. It will be a bit higher when it turns out to be a ‘common’ word of the type many people will choose as their starting word and probably highest of all when Wordle chooses one of the words that have been widely suggested on the net as good starting words. Conversely it should be lower for words which people are unlikely to choose as a random guess – like ‘Wrung’ – wordle 225. Of course all the numbers on wordlestats will be biassed by the tendency for people to selectively post their ‘best’ results, as noted by several people here. Getting it on the second attempt is much harder to calculate the odds for but on average people with a vocabulary of words similar to the wordle dictionary are going to get it more often than 1/2314 (having eliminated one word with the first guess) because in many cases their first guess will have markedly reduced the list of possibles. We’ve had 3 second guess hits out of 92 attempts. As this is still largely down to luck, just with a reduced pool of possible words to choose from, I can’t at the moment think of a better way of estimating the odds for guess two. There have been attempts to do it mathematically but I think they fail to take into account some of the clues that we get from the initial guess as well as the fact that we apply quite bit of knowledge about what letter combinations are common in English. So at present I would guess that there is about a 2-4% of a second guess hit overall.

  15. Wordle is a three letter word…pure joy.

  16. A further thought – I don’t think Wordle repeats words, at least not for now, so if you have a good memory for the words you have already solved the pool of possibilities will be decreasing every day and thus improving your chance of getting quick solutions somewhat.

  17. The ACTUAL average of your score requires calculation of the decimal point to give a graduated rating. This equates to your total attempts ‘A’, divided by the total games played ‘N’ : Average = A/N.
    Just a little more indicative than rounding to the nearest integer!

  18. I would have thought that the real measure of success is not how many guesses are required to solve each puzzle but rather the overall percentage achieved. That cannot be hidden or skewed in any way. To my mind is a more accurate statistic to follow.

  19. Wordle is not a team sport as it is an opportunity to share a pastime with friends on a level playing field. For me, I like to keep my personal average score, like your school GPA, to accurately demonstrate where I stand on any day. I tally my total points, divide by number of games played and then try a bit harder to bring that average down, slowly it seems. I’ve yet to exceed six guesses and if I do, it will be marked as six points that day for scoring average purposes. In my 25 games, I have totaled 93 total points for a 3.72 average. Earlier on I had a 3.5 average but my solo sixer along with a couple fivers blew that up to a 3.85. Focusing more, rethinking my starter words, is actually bringing that number down to today’s average. Love the challenge, enjoy the exercise. Phil M.

  20. True of course, but still very subject to reporting bias and cheating. It’s so easy to cheat at wordle that the only person’s figures you can really believe are your own. Only you know whether you had any hints that steered you in the right direction – even a friend saying “Oh it’s a tough wordle today” will influence your choice of words somewhat. I try to do it first thing in the morning before anyone can inadvertently tell me the answer but no-one else can verify that I really do that, except perhaps my wife, and even she would not know if I had secretly checked the wordle source code to find out in advance what word was coming up. For anyone who hasn’t twigged this yet, there is a predetermined list of solutions, at least until such time as the NYT decide to change it.

  21. Phil – I’m currently running an average of 3.82 over a 50 streak but I think a fail should be counted as some number higher than 6. 7 is probably the most obvious approximation as I think most people who are not just choosing random words on every guess have probably got it narrowed down to one or two options by guess 5. Pooling all the available data from @wordlestats currently gives an average score of 4.23 but that’s markedly biassed towards better performance. I’ve seen one starting guess with 4 letters in the right place which still left 6 more alternatives – that person must have been really annoyed when the right one turned out to be their 6th guess.

  22. Would be interesting to know how the stats change when there is a word with two of the same letter in it.

  23. Graham – words with two letters the same do tend to be harder, at least judging by the twitter stats, only ‘robot’ has recently had a below average mean reported score. Words ending in e also seem to be tougher – I think because there are so many of them hence ‘shake’ a few days ago, you could have sha_e and it could still be shade/shake/shale/shame/shape/share/shave – so anyone who started with any of those words other than shake still had a 1/6 chance of failing despite having 4 correct letters on a first guess

  24. Update on that last comment. There is a strategy for solving the shade/shake etc problem as long as you are not in hard mode, you can guarantee to get it in 5 or less after the second guess

    1. Wasn’t is “SHALE” that day?

      1. It was ‘Shake’ but the problem is the same with any of the 7 common words that fit that pattern. I’ve just discovered a mathematical/computational approach which demonstrates that any word from the 12972 possibilities in the original wordle ‘dictionary’ can always be solved in 6 guesses, but not necessarily in 5 so Josh Wardle’s original choice of 6 seems to have been exactly right, just enough that there is always a solution (at least in standard mode – in hard mode it can take up to 14 attempts to arrive at the right answer) but no more than is necessary. The same source tells me that the NYT deleted 30 words from the original solution list – thus making it slightly easier to get a first guess right. I’m not sure whether 4/6 letter wordle would be harder or easier – the statistics of it is surprisingly complicated.

        1. Without looking at statistics, I assume 6-letter words would be easier, and 4-letter words slightly harder than 5-letter words. Furthermore, 7-letter even easier, and 3-letter harder than 4-letter.

          1. I’m not sure – there’s a more limited pool of 3 letter words to choose from (1065 according to the scrabble dictionary, and 3996 four letter ones). So the reduced number of opportunities to rule out letters by getting them wrong might be counterbalanced by the smaller pool of solutions.By 5 letters we are up to about 13k words and by 6 letters something like 22-24k.

          2. It’s official. Wordle2 (6-letter Wordle) is easier to solve.
            – – – – – – – – – – – – –
            My Wordle stats:
            3 tries – 17
            4 tries – 17
            5 tries – 17
            – – – – – – – – – – – – –
            My Wordle2 stats:
            3 tries – 8
            4 tries – 3
            5 tries – 1

            Smaller sample size, but I’m confident the ratio will stay about constant.

            I aver that Wordle 4-Letter would be MORE DIFFICULT than 5-Letter.

  25. The nervousness if you have 100% success, then get greedy one day but end up with a 6th guess!!! The Streak stat is useless, all it points out is that you forgot to play one day. QUESTION: Would a 6-letter Wordle be easier or harder? Sunday Wordle should be 6 letters. Moreover, would a 4-letter Wordle be easier or harder? And Wordle success stat would be more meaningful if only 5 guesses were allowed. 6 guesses = too many 100% players. Whoa, did I just jinx myself?!

    1. I have a significantly better average on Wordle2 than Wordle, so maybe 6, at least for some people, is easier. My son is the other way round, we do have different thinking styles, I tend to a more analytical approach, he to a more visual spacial one.

      1. Thank you Robert! I didn’t know there was a Wordle2. I can see doing today’s that 6-letters can require slightly different strategy than the 5-letter.

      2. 6-letter is easier. Eventually, your son will improve on 6-letter with repetition.

  26. My stats have 1 recorded on the first try because I already knew the word. I played the Wordle on my phone on the mobile site and got the word in 3 or 4 tries. When I read the instructions on how to save your Wordle in your app list, it reset my stats, but I played the word I knew to get started. I was ignorant as to what was happening and how the game stats worked. This was only my second time playing Wordle. The first was at a meeting ice breaker with other people. One could assume a lot of the first try stats came from experiences similar to mine where they knew the word when they first saved the app link on their device.

  27. It’s official. Wordle2 (6-letter Wordle) is easier to solve.
    – – – – – – – – – – – – –
    My Wordle stats:
    3 tries – 17
    4 tries – 17
    5 tries – 17
    – – – – – – – – – – – – –
    My Wordle2 stats:
    3 tries – 8
    4 tries – 3
    5 tries – 1

    Smaller sample size, but I’m confident the ratio will stay about constant.

    I aver that Wordle 4-Letter would be MORE DIFFICULT than 5-Letter.

  28. The Win percentage calculation program isn’t factored correctly. The equation is percentage played. Skipping games does not affect your percentage. There is a glitch in the program.

  29. Anybody who gets it in 1 or 2 guesses is simply, lucky. I like to do it against time. I try to finish within one minute, and the majority of time, I can do it. Of course, this means using more guesses, but there is also skill in being to do it quickly, athinkjng about the letters you have uncovered, whilst typing in another word.

  30. I, too, had concluded the Streak was the number of consecutive days of wins, and if you failed to play for a day, your streak would end. But this past week I was on a plane throughout the day, did not cross the Date Line, and did not play Wordle. The following day I played and my streak continued as though I hadn’t missed a day. Maybe they offer a one-day forgiveness period? Since they tracked what time zone I was in, possibly they offer a mulligan, knowing air travel would eat up most of the day. I have no answer, but my Streak continues.

    1. I did an international trip across the date line – my streak ended when gong east even though I did the puzzle early in the day. On the return, I did both (once on each end), and only got credit for one so the streak ended again. I think it has problems with the date line.

  31. I disagree that solving in 2 is simply lucky. Solving in 1 is virtually blind luck, thereafter luck plays less and less a role with each attempt. It’s not just the letters you choose in your guess; it’s also the order of the letters that give bang for the buck. You could use cored or coder or decor, but one of those is a waste of letter placement. That said, I don’t play against time as Paul does. I may spend an hour analyzing results and options. It pays to be retired.

  32. Wordle in under a minute?! Then what else are you gonna do while drinking the rest of your coffee or sipping a glass of wine? Slow down Paul. That being said, I do the Jumble in under a minute!

    1. It does raise an interesting point – what your aim is. The Aim could be:
      – the sortest time (as with Paul)
      – the lowest average (what most do? but could result in people “abandoning” any game where it would tale 4 or more to complete)
      – the maximum chance of doing it in 2 words (in which case you would try to optimise the first word you choose rather than do what I do get the optimal two opening words)
      – the approach that takes the least brain power without losing (this favours fixing the operning two words rather than just one)
      – never losing
      – the maximum streak (never miss a day, make sure you never lose)
      No doubt there could be other AIms I haven’t considered. And who’s to say which AIm is right?

  33. I think your average score depends a hell of a lot on your strategy and your tactics (as opposed to just your intelligence or vocabulary). As an example, I used to ALWAYS use “Royal” and “Unite” as an opening two words, and had an average score of 4.2 (with 14% 3s and 45% 4s). I changed (based on an analysis I read) to “Cones” and “Trial” and averaged 3.7 (with 47% 3s and 37% 4s). Same me, very different results.

  34. You’ve exchanged the two least-used vowels (u,y) to two commonly used consonants (s,c). I never bought into the strategy of loading vowels at the front end. I have five sets of three words, using the u/y in the third word, but rarely needing to use the third word at all. And btw, I’d suggest you switch trial to trail– ai is more common than ia.

  35. I remain puzzled by the problem of 2nd guess solutions and how statistically likely they are. Now that quite a few people have posted 100 streaks you can see that a few people are posting what looks like unlikely numbers of second guess hits (10/100 for example). My guess is that people with significant numbers of second guess hits are either consciously or unconsciously using hints – there are enough of them about – including one twitter spoilsport I saw who posted the solution for the next day’s puzzle under the hashtag for “today’s” – fortunately I only saw that twitter posting two days later otherwise it would have ruined a days play for me. We’ve had 4/218 between the two of us which I suspect is a fair estimate of the odds but I still haven’t got round to trying to calculate it statistically.

    1. Jeremy, note today’s (Saturday, 4/23/22) word. I’ll share something with you tomorrow.

  36. It never occurred to me to abandon a game before a sixth and potentially failing guess to keep my 100%! Which I would never do, but, duh. If I were a ‘100% with an asterisk’ type of person, I should’ve done that with SCOUT, SHOUT, SPOUT, STOUT, SNOUT from a few weeks ago.

    Follow up question: That SNOUT train wreck made me 72 out of 73 = 99% (but 98.63% rounded up). Do you think that when I get to 200 out of 201 = 99.502%, it’ll get rounded back up to 100% ???

    Another follow up: My WordHurdle (6-letter), guess distribution is virtually the same as my 5-letter disti.

  37. Jeremy, regarding yesterday’s (4/23) game, I solved it in 2, my 7th 2 in 85 games. I don’t cheat and never bailed out of a game in order to avoid a loss. My streak is only 71 because in the first month of play I didn’t realize that I had to play every day to avoid breaking the streak. So someone hitting it in 2 ten out of 100 times is not at all odd. Regarding probability, you can’t really measure it cuz your first guess could be chockful of good letters or it could be “kayak”. If I’m forced to follow my strategy through two words, I’ve covered the top 6 consonants and top 4 vowels in the English language. But I strategically choose my first word based on recent history in the game. So here was yesterday’s game colors for my first guess: black-green-yellow-black-green (it seems I can’t copy/paste the color scheme here). With that data, I reviewed every letter permutation and came up with 3 possible solutions: 1 skateboard trick that you’d only know if you skateboarded or did crossword puzzles; one noun that is nearly always used as a 6-letter plural; and finally a vegetable. The veggie was thus a no-brainer. Moving on to today’s game, I’m proud to announce I hit 2 again! I’ve hit 8 out of 86 on the second try. Four yellows on my first guess, knew which letters would likely come up. Choose your first word, and if necessary your second, wisely.

  38. Actually if you know something about the code underlying wordle you can estimate the probability of getting it in two if you make one approximation. I just haven’t got around to finding the time to write the computer code to calculate it. The approximation is how closely your own vocabulary of 5 letter words matches the 2300 or so chosen by Josh as solutions to the puzzle. There are actually many more than 2300 which will be accepted by the puzzle as legitimate guesses but for practical purposes I think we can assume that most people do not know that ‘vinic’ is a word and that ‘phpht’ is only of interest to scrabble players. So we have about 2300 or so first guess possibilities. It’s then a fairly minor, though repetitive, task to take every possible first word and solution (2300 x 2300 combinations) and comb the list again for how many possible solutions would fit the ‘rules’ generated by the first word combined with the solution. For example, if the first guess is ‘Shape’ and the solution is ‘Shake’ then the rules generated by the first guess are that the solution is “SHA_E” and the possible solutions after the first guess are Shade, Shake, Shale, Shame, Share, Shave. You thus have a 1 in 6 chance of getting the solution in two with that opening word/solution pair. Give this combination a ‘score’ of 6. If you repeat this exercise for every possible starter/solution pair you will arrive at three things:

    a) it will be apparent which is the best starting word if you want to solve in two – it is the starting word where the sum of the scores for all 2300 solution words is the least. My guess incidentally is that even with the optimum starting word the sum score for all 2300 solutions will be > 23000 making an average performance of 10% second guess wins increasingly unlikely as the sum score gets bigger. Bear in mind that Shape/Shake, though infuriating for anyone in hard mode, is actually quite a low score. If you think about starter Shape, solution Lowly there are going to be hundreds of possible second guesses – essentially every 5 letter word without the letters SHAPE

    b) The average score for a given starting word gives the overall probability of the player getting a solution in two using that particular word

    c) The average of all the average scores for all starting words is the overall probability of getting it in two with ‘any’ starting word

    That’s how to calculate it – you need to download the solution list to write a program to do it and I’ve resisted doing that because it would risk spoiling my own enjoyment

    There isn’t really much strategy involved in choosing a first word based on recent history. The only advantage given by recent history is that we know that wordle does not repeat solutions so you can mentally eliminate as many words as you can remember as a first guess, and also as a second guess solution. I can’t remember all of my 110 solutions to recite them but many will feel familiar when I think about them as a possibility – I will have an inkling that “We’ve had that before”. Out of a list of 2300 though, knowing that a 100 are so are not solutions today does not lower the odds much. There is no other predictability to the solutions, they are just a predetermined random word list and there are no hidden ‘rules’ such as – it’s not going to start with c two days in a row.

    I’m pretty sure my choice of starting word, which I made without very much analysis when I started playing, is not the optimum one for two guess solutions, but I’m sticking with it because I know that sooner or later it is going to be the solution and I would like a hole in one (I’m just keeping my fingers crossed that it was not a very early solution – again I don’t want to scan the solution list to check). Once it is then I will pick a new one, and at that point I will be doing some mild cheating in that I will exclude all the solutions I’ve already seen, thus increasing my chances of another 1st guess solution

    1. It’s a bit harder than that to calculate the odds of getting it in two. You’re making the assumption that people are only trying to get it in two, whereas most players are probably aiming to solve it in as few tries as possible, and if they get it in two, then that’s a bit of good luck.

      On your second guess, most of the time you will be playing a word to get a few more letters, rather than going for the actual solution.

      The times when I’ve got it in two are:
      – when I think my first guess has narrowed the word down to one possible solution
      – when I think my first guess has narrowed the word down to two possible solutions, and I get a bit lucky with a 50/50
      – when I think my first guess has narrowed the word down to three (or more) possible solutions, but by guessing one of the solutions, I can reduce the remaining solution set to one
      Other than these scenarios, I’ll probably play a letter-reducing word to reduce the possible letter combinations.

      Of course, sometimes I’m wrong about the number of possible remaining words in the solution set!

      I’ve got a ~10% record of getting it in 2 (6/60), so the stats don’t seem unreasonable. Average is 3.7.

      1. I don’t think there are many first guess/solution pairs that really narrow down the options to only 2 or three possible words. Again I can’t put figures to that without doing quite a lot of number crunching but it can be done – eventually I’ll get bored with my wordle streak and do the math I guess. The strategy being played will influence the raw odds to some extent but I don’t think it can improve on them – only reduce your chances of a solution in two. For example choosing two starter words to cover all the commonest vowels and consonants and using the same two every day will distinctly reduce your chances of a 2 solution, reducing it to zero once the second word has appeared as a solution. Some people choose unusual objectives for wordle. Someone I saw the other day had the aim of producing a symmetrical display by the time they had the solution! Pooling all the data available on twitter @wordlestats shows that people overall are claiming just about 6% second guess solutions but given the presumed severe bias in favour of people posting their ‘best’ results I’m pretty sure that the chances of a second guess solution with varied starter words are a good deal less than 6%, and 10%, over a long run, is looking distinctly like an outlier. It’s possible that some optimal starter words might push the figure higher than that if used repeatedly – again an interesting calculation.

  39. I would like to see Wordle provide (in the summary) the average number of tries for a given word. For example, today’s word was “trash”, and it took me three guesses. I’d like to know what percentage of players got in two, three, four, …. guesses.

  40. Please do not announce the answer to today’s puzzle. Some people may not yet have played.

  41. I cannot believe what I am seeing with Wordle. It is a kids’ game, embarrassingly easy. The world is just so dumbed down these days. I played Word Mastermind 40- 35 years ago as a kid, but you were not told which letters were right or wrong, you had to work that our with logic. Worder.ly is available on line and is the real way to play.

    1. If you aren’t told anything about which letters are right or wrong you are proposing a game that is simply random word guessing, which is not much fun, so I guess you are oversimplifying. That is not the point of wordle however, it’s not supposed to be a really difficult challenge, just a few minutes diversion that doesn’t swallow up your whole day. Curiously the design is almost perfectly judged – every puzzle can be solved in 6 attempts with an ideal strategy in standard mode, but not in 5 attempts. Playing in hard mode means that even with an ideal strategy some solutions can take 14 attempts.

  42. I’m old, but not THAT old…am I the only one who wishes the letters on the ‘keyboard’ were a few pt. sizes larger???


    Oh my GOD it was so frustrating trying to chase down this simple fact. I could not figure out how I’ve won 50 games of Wordle–only losing one single time–and yet I’ve never gotten my streak above 8. My current streak is 2, what?? It’s because missing a day ruins your streak.

    PLEASE CONSIDER UPDATING YOUR ARTICLE TO MAKE THIS CLEAR. So others after me don’t spend an hour trying to google this simple fact. UG.

  44. I have played wordle since mid January. After playing sloppily and missing the clue word- swill- I have a streak of 122. Here’s my raw numbers: 0, 6, 50, 70, 28, 6, and the one miss. I switched strategies, no longer try for two guesses, and always start with two words that use all the vowels and the letter y. Unless I get lucky and hit three or more letters on the first guess.

  45. Eric, I agree that the lead-in to this page ambiguously references “win streak” when streak is merely the number of CONSECUTIVE wins. Your plea to update this article leads to a series of questions I have:
    Is there anyone who actually represents Wordl who can clean up this page?
    Back in April Jon referenced 2309 potential solutions and 12,972 acceptable words to seek those solutions. Where can I find that (and other related) information?
    I’ve concluded that the Wordl solution will not be a 4-letter word made plural by adding an “s”. Is that an official rule/parameter or is that just the consensus of players never having encountered such a 5-letter solution?

  46. The wordle source code has appeared several times on the net and can be googled. Both the original word lists and the amended NYT times versions are available (after they took out Fetus for example.) I too concluded early on that Josh had made a deliberate decision to exclude plurals – ie four letter words with an ‘s’ added, and I also worked out by simple observation that it seemed not to repeat words and that there were no proper nouns – though I thought ‘Rupee’ was cutting it a bit close. One could check that these observations are correct by looking through the solution list but I’m still managing to resist the temptation to do that even though I have a saved copy of the source code. I am now finding, after 190 days, that knowing many words that have already ‘gone’ combined with the fact that I know it is not going to repeat, is sometimes giving me an edge – as today when I knew that one the only two possible words I could come up with as a solution had actually been the answer quire recently.

    1. Jeremy, I’m in the same boat (except don’t have the source code); eliminated one of my options because it was a solution recently. I’m also finding it’s harder to solve in three attempts; was fighting to stay under a 3.7 average, now at 3.72 and feeling like I’ll never see a 3.69 again.

      1. My long run average is at 3.90 but I know my regular starting word is not the best possible choice – I’m just waiting for it to come up as the word of the day. I was quite chuffed when today’s solution was my name!

        1. I have a half-dozen pairs of words, each pair composed of the top 6 consonants and 4 vowels. It works well, but the day I chose not to start off with “train”, one of my favorites, was the day that train was the solution.

    2. This is off-topic to your source code comment. But, I found out today as I reached 199 correct out of 200 plays, that my regrettable 99% reached 99.5% (199÷200=.995) that Wordle rounds UP!!! I’m back to 100% (even though I’m not technically). But it says so on the screen, so…!!!

      1. Say 12 Hail Mary’s and we’ll let it slide.

  47. In response to genius Jon, who I suppose solves Wordle daily in 1 or 2 tries, it’s not a matter of simply solving in 6 guesses, it’s what percentage of solves in 1,2,3,4,5,6 guesses. Moreover, who needs a brain twister at 7:00 am with your coffee. And yes, it is fun to play with your kids.

  48. A clarification regarding “hard mode:” it is harder to win but easier to play, less strategy/thought. For example, suppose after 3 guesses you have sta_e, and your 3 guesses have eliminated stare and state as solutions. You’re then left with stage, stake, stale, and stave as potential solutions. In hard mode, you’re simply guessing, no real strategy required, and have a 1-in-4 chance of losing. [I might have previously argued stave is the least common among the 4, so choose that last. I’ve since abandoned that theory.] But in regular mode you are searching for a word that incorporates at least three of potential remaining letters (g,k,l,v) and that in itself requires more strategy than the hard mode’s simply guessing. In this case, if you choose “glove” as your fourth word, you’re guaranteed to win in 5. So if your goal is winning and/or exercising your mind, why would you ever play hard mode?

Comments are closed.