Title says most of it. I've been at this for hours with almost no progress. There are two codes (37068 and 57063) that didn't work although they seem to fit the conditions given. Can someone please give a clue to put me on the right track? I feel like I'm going absolutely nowhere, but would still like to figure it out.

Update: The owners of the website have fixed the code. Apparently it was bugged and the visible number code that were hints weren't changing, although they were supposed to. This made the puzzle impossible since the answer was changing each time the page was refreshed or you attempted to crack the code.

Thanks y'all for helping before and after the fix!

  • 7
    $\begingroup$ Since the 4th and the 1st are the same code but return 2 different results, we can guess that assertions are inclusive only : they do say something about the attempt, but there may be more than stated. Ie, we should read "at least two numbers" instead of just "two numbers". My question is : is this a rule we have to guess or are there some written rules you omitted? $\endgroup$
    – Marvin
    Apr 15, 2019 at 8:33
  • 4
    $\begingroup$ @Marvin I think you're right because 16374 and 50928 contain all 10 digits, so 5 of them should be correct (although probably in wrong positions), since each digit is used once, but the conditions state that only 1+2 = 3 digits are correct $\endgroup$
    – trolley813
    Apr 15, 2019 at 9:29
  • $\begingroup$ Is this actually a screenshot of a phone, or is the data at the top (e.g. time and power level) potentially part of the puzzle? $\endgroup$
    – Pugmonkey
    Apr 15, 2019 at 15:05
  • $\begingroup$ This looks like an interesting puzzle, but I suggest you might want to give a hint because it looks as if there was activity on this puzzle yesterday, but not today. - I noticed in the edits that you had riddle tag, but they were changed to 'number puzzle' - do you agree with the description of this puzzle as a number puzzle or is it more like a riddle? $\endgroup$
    – tom
    Apr 16, 2019 at 23:05
  • $\begingroup$ This particular version of the game seems broken. There is a working version that I completed. That last series of numbers on the bottom is not supposed to be identical to the top. Reset for new numbers and maybe you'll get a working series. $\endgroup$
    – mkinson
    Apr 17, 2019 at 11:49

6 Answers 6


The puzzle is broken. You can find it at project-archive.net. Looking at the Javascript code, the solution changes every time the page is loaded, but the tips are the same.

  • 2
    $\begingroup$ Well done, I guess (?????????) Can I ask how you found that? $\endgroup$
    – user46002
    Apr 17, 2019 at 6:02
  • $\begingroup$ I solved the puzzle from the project-archive.net site you show.. it says "Congratulations, you have completed Phase 1. Phase 2 starts April 19th. Text "PHASE2" to [a phone number I wont post] to begin. $\endgroup$
    – mkinson
    Apr 17, 2019 at 11:46
  • $\begingroup$ It actually seems to be working. The text is the static part, it's the solution as well as list of entries that change every time you reload the page. The screenshot posted here seems to be from an early version that was still under development. The current version writes the solution to the console, probably as part of figuring out and fixing bugs. $\endgroup$
    – Moghwyn
    Apr 17, 2019 at 11:52
  • $\begingroup$ They must have fixed it. Yesterday it was not working for me, showing the exact same clues as in the OP even after reloading, but today it shows fresh clues every time. $\endgroup$ Apr 17, 2019 at 12:23

There's multiple contradictions in the puzzle as presented:

  • Between them, 16374 and 50928 represent all ten digits. The first has one correct number, but in the wrong position. The second has two correct, one in the wrong position. So out of all possible digits, only 3 are correct. Yet we're later told none are repeated. So as it stands this appears impossible.

There is a lateral thinking workaround for this:

The responses given for each guess refer to 'number', not 'digit'. So they could be referring to more than one digit, e.g. 16 is correct in 16374, but in the wrong position, would mean the sequence 16 appears elsewhere in the correct code.

This does seem needlessly convoluted though.

  • 97360 appears twice, with different results. The only explanation I can see for this is that the code changes with each attempt, possibly according to some rule, but that pretty much renders the puzzle unsolvable.
  • 2
    $\begingroup$ What's strange about that last point is that the problem statement claims this is your last attempt before the combination resets. So, perhaps the combination reset after the first attempt, but can be deduced from the second, third and fourth. Or perhaps "reset" means the combination rotates through a short list of combinations, and after this last attempt it'll go back to the beginning, which would mean you want to get it wrong enough times to "reset" it so you can keep trying with a consistent problem. $\endgroup$ Apr 15, 2019 at 17:06

For 37068, the first result would have been 1 in the correct position and 2 in the wrong positions. The second result would have been 3 correct but in the wrong positions.

For 57063, the first result would have been 1 in the correct position and 2 in the wrong positions. The second result would have been 3 correct but in the wrong position.

You have to use all the information available to solve these, and narrow down what digits are ruled in or ruled out by considering all four results.

What is concerning here is that the usual expectation is that you'll be told how many (if any) digits are correct and in the correct position, and how many (if any) are correct but in the wrong position. Since we see the same guess generate two different outputs, either the puzzle is flawed or there is more going on here than meets the eye. If it's giving incomplete information for each guess, this doesn't seem like a fair puzzle.


First of all, we can state:

Of the combination 97360, 2 digits are in the right position, and 1 digit is in the incorrect position.

Also, 50928 has 1 digit in the right position, and 1 digit in the incorrect position.

When we look at the overlap in used digits between 97360 and 50928, we find that only 9 and 0 occur in both. Because the positions of these digits aren't the same for both, (1,5 and 2,3 respectively), and 50928 doesnt have "2 correct digits in the incorrect position", we can conclude that "9 AND 0" aren't both in the final code.

If either 9 or 0 is in the correct position in 97360, the condition for 50928 "one digit is in the incorrect position" is satisfied.

I feel like it's fair to say that 16374 doesn't have any numbers in the correct position. If it did have numbers in the correct position, the puzzle should've mentioned that.

Because the overlap in numbers between 97360 and 16374 are the numbers 3, 6, and 7, where the number 3 is on the third digit in both of the codes, we could rule out 3 as being "one of the numbers in the correct position" in 97360. I'm basing this on the assumption that the puzzle didnt provide information about any digit being in the correct position in 16374.

Because '3' is neither in a correct nor incorrect position in the final code, we can conclude that 3 is not in the final code.

Therefore, 37068 and 57063 are not possible final codes.

We can also conclude that of the digits 0,6,7 and 9, two are in the correct position and one is in an incorrect position, so one of those digits is not in the final code.

Looking at my earlier conclusion where either 9 OR 0 is in the final code, we can conclude that 6 and 7 are in the final code.

This would contradict the statement about 16374 though, meaning we can't assume that 16374 doesn't have any digits in the correct position.

I'm stuck here, this was just my line of thought regarding this puzzle..


Your version seems broken. Posting this method of solution in case you get it working.

So I went to http://project-archive.net/ as DASF indicated that the puzzle existed there and failed a couple times. The numbers changed each time but the clues remained the same, and that's because they were accurate each time as the puzzle reset with a new code and new numbers to correspond to it.

My numbers were:

4 2 5 3 7
7 4 5 1 0
6 2 3 0 5
5 8 7 2 6

My solution:

The first thing I did was get rid of all numbers that existed in at least three rows to remove clutter, so 7, 5 and 2 were out.

Then, focusing on the 4th set I looked for numbers that existed in that series but not the first two series, so 8 and 6. One of these was in the correct position, the other was not.

Then the 3rd set of numbers had only 6 # 3 0 # left, and since 6 existed for that and the 4th set I assumed 6 was in the correct position for the 4th set and was the correct number but wrong position for the 3rd set.

This meant I had a number that was in the correct position still for the 3rd set, and since I have a correct number in the 1st set with only 4 and 3 left as options and no 3 in the 2nd set, I figured 3 must be the correct number.

So now I had # 8 3 # 6 for my answer. The 2nd set had to have a number, so all that was left was 1, since 0 existed in the 3rd set, so that must be my first position.. then all that was left was 9, which didn't appear anywhere.

Correct solution: 1 8 3 9 6


If you go to the webpage in question, right-click for 'View Source', and then go to Console, you can see your answer (top right):

enter image description here

and the clues fit! Refreshing the page changes the answer in the console and the combinations on the main page, but not the clues.

It works by:

creating a vertical permutation of the rows of:


and then assigning a unique digit to A-J.

Then the code displays the first four columns as previously tried combinations, for example, FJHDI.

This means the code is cracked. For example, A never appears (so A=0 in my example), and H appears four times, and so H=7. As H appears 3rd on the bottom guess, we know G=8, B=6, D=2, and 0 is in position 3 in the answer. From row 4 on the original array, D is in position 5, etc...

  • $\begingroup$ As I said as a comment on another answer - Yesterday that website's script was broken, and it always showed the numbers shown in the picture posted in the question. They must have fixed something, cause today using the same browser as yesterday it suddenly works the way it is supposed to. $\endgroup$ Apr 17, 2019 at 14:26
  • $\begingroup$ @JaapScherphuis; I'm using Edge on Windows10 and it works for me. $\endgroup$
    – JMP
    Apr 17, 2019 at 14:30
  • $\begingroup$ As I said, it is working for me now too, but yesterday it wasn't. $\endgroup$ Apr 17, 2019 at 14:39
  • $\begingroup$ @JaapScherphuis; starting at line 64 in Debugger, there is a hardcoded effort which gives OP's question - they must not have overwritten it or something. $\endgroup$
    – JMP
    Apr 17, 2019 at 14:46

Not the answer you're looking for? Browse other questions tagged or ask your own question.