Consider a standard game of Scrabble® using the csw19 dictionary.

Many board-states exist where ZERO possible next moves exist. That is, no combination of 1-7 remaining letters can make a legal move.

My question is, how fast can you legally get the board into this state? I have a solution with an answer of FOUR moves, and will award the check to the first to provide one such solution.

Note: Better solutions may exist.


  • The 'players' obviously will not be playing intelligently or even logically, just legally.
  • The moves may take certain tiles out of play that would have been the only way to make a valid word later
  • 1
    $\begingroup$ The tag excerpt for [open-ended] starts with "OPEN-ENDED PUZZLES ARE NO LONGER IN SCOPE." - i.e. off-topic. See here for why. If a provable optimum exists, then [optimization] would work. $\endgroup$
    – bobble
    Sep 16, 2021 at 15:55
  • 3
    $\begingroup$ I am asking for a particular solution. The puzzle itself is NOT open ended. I added the tag because optimizing the solution may be of interest to some people as an additional exercise. I'll remove the tag, but you're being pedantic. $\endgroup$ Sep 16, 2021 at 16:18
  • 6
    $\begingroup$ @TwoBitOperation I wouldn't say it's pedantic. Part of moderation is ensuring that users are following site guidelines. If your puzzle isn't open ended, then don't use the open ended tag. As you and bobble both pointed out, the optimization tag is a better fit. $\endgroup$ Sep 16, 2021 at 16:24
  • 5
    $\begingroup$ @TwoBitOperation correctly tagging questions is an important part of curating the site. This isn't open ended, and bobble helpfully provided an alternative. $\endgroup$ Sep 16, 2021 at 16:25
  • 4
    $\begingroup$ I guess my gripe stemmed from the fact that so often a puzzle posted here and the discussion immediately goes meta: Was it posted the right way? Can I find a duplicate? Etc. Anyway, I suppose that's the internet, and it is a very fine moderation line to walk. And, all things considered I came across as way more a jerk than I intended to, so sorry about that. Back to puzzling. $\endgroup$ Sep 23, 2021 at 16:47

2 Answers 2


How about this succession of plays:

h6 QAT; 7g qAT; 8h (T)Ix/(T)I; 9i XU/(TI)X/(x)U

producing a grid


  • $\begingroup$ This looks good; give me a little time to try to find a play, but I can't see one at the moment! Great answer! $\endgroup$ Sep 28, 2021 at 11:53
  • $\begingroup$ If there's a play I can't find it. Nice work! I'll post my answer for comparison. $\endgroup$ Sep 28, 2021 at 20:11

Please UPVOTE @Rosie F's Answer

The 4 move solution I found when designing the question:

1) GIBUS across
2) (Q)IS down
3) SU(Q) down
4) GIBUS down

4 Move Solution

The only acceptable move is 'gibuses' which cannot be played since no blank or S remains.

  • $\begingroup$ Forgive my possible shortsightedness here, but where are the two blanks used? I see one for the second 'Q', but where is the second blank played in this grid? Thanks. $\endgroup$
    – Stiv
    Sep 28, 2021 at 20:26
  • $\begingroup$ I wasn't really clear. The main point is that both blanks are both used somewhere other than the S's. I had them on both Q's, and have updated the picture to show. $\endgroup$ Sep 28, 2021 at 20:35
  • $\begingroup$ Ah, okay gotcha - so the blanks have both been used in place of other tiles which are still available in the bag but which were not drawn by the players in the course of this game. Got it :) $\endgroup$
    – Stiv
    Sep 28, 2021 at 21:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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