Jumble: That Scrambled Word Puzzle has been in my newspaper since forever, and I assume that most of you have seen it or something like it. That link takes to you to today’s game.
The game always consists of two five-letter words and two six-letter words, Monday through Saturday. Sundays have an extended game of six six-letter words.The twist is that there must be a unique solution for each set of letters. No anagrams allowed.
For some reason I started thinking about making these. They need 313 x 2 or 626 five-letter words and 313 x 2 + 52 x 6 or 938 six-letter words. And they all have to be common words that average people would recognize.
How many five and six-letter common words fit the requirement of one unique spelling? Are five-letter words or six-letter words more likely to have one unique solution? How quickly would the makers have to reuse words? I assume they try hard to avoid repeating words too closely together because daily users would start to recognize them. They do have the advantage that the scrambled letters are presented randomly so that PLEEO and OPELE and LEOPE are not instantly recognizable as the same word, whereas an alphabetical EELOP might trigger memories.
I’m putting this in GQ because the question isn’t so much about the game as it is about the English language, but do what you see fit.
I have a list of 235,000 English words on my computer. I wrote a quick program to look for words of a given length which are not anagrams of another word. Results:
It occurs to me that the words don’t have to be unique - ANTU can create AUNT and TUNA - but if the circled letter is the third, then both answers come up with “N” which is allows the rest of the puzzle to go to completion. You need sets of letters that only produce words that have at least one letter in the same position.
So many word lists to choose from! (And the famous Scrabble™ list is well-protected.) One list I use has 109,573 total words — it has inflected forms, but no hyphenated or apostrophized words; it has only a few proper nouns. Here are some stats for that list:
2-letter words: 97 out of 141 are unique under anagramming
3-letter words: 495 out of 852 are unique under anagramming
4-letter words: 1654 out of 3133 are unique under anagramming
5-letter words: 4110 out of 6915 are unique under anagramming
6-letter words: 7906 out of 11492 are unique under anagramming
7-letter words: 12939 out of 16879 are unique under anagramming
8-letter words: 16534 out of 19459 are unique under anagramming
9-letter words: 15257 out of 16694 are unique under anagramming
(Perhaps I should e-mail, say the 11492 six-letter words, to markn+, or vice versa, to get a ‘diff’ and a take on which list is better for these sorts of queries.)
Thanks, guys. I was hoping somebody could program this.
The numbers are certainly lower because some of the words would never be used, but having thousands to choose from makes the construction a lot easier.
At one time I used to do the word unscramble part of Jumble. I rarely found the final answer, but that’s because I usually didn’t even look at that half of the puzzle. Anyway, I would occasionally find an anagram of the intended answer. The answer I found was always a somewhat obscure word, but which I knew because my interest in certain types of words or certain subjects. One example I remember had the letters LLSYY (not the actual clue). My immediate solution was LYSYL (a palindrome), but they wanted SLYLY. I thought my answer was better.
BTW, I once (back in the 90s) had a telephone conversation with the man who invented Jumble. I don’t remember his name and can’t even remember how I happened to have this call. It had something to do with some article I wrote for Word Ways, but that’s as far as I can say. Anyway, he asked if I was interested in word puzzles and it resulted in him sending me a book of these Jumbles that he’d published. The ones in the book were harder in some way (which I can’t remember) than the ones you see in the daily newspaper. Whoever bought the rights to the puzzle dumbed down the concept to make it more palatable to the general public. I still have the book, but it’s packed away, so I can’t tell you the guy’s name.
Good grief. The only reason I would even try to do one is if I got to roll in the mountains of money the Jumble generates. Since that won’t happen…
I don’t know whether it’s a sign of age or if 720 possibilities really is a mental leap from 120 but I can solve the five-letter words in my head while the six-letter words are too hard. That’s the only reason I ask, just cogitating randomly from that base. (And I also can often guess the pun from the clue, which allows me to cheat and work backwards if I care. Which I don’t. I bet I haven’t completed a half-dozen jumbles.)
None of those names (from Wikipedia) ring a bell, but the phone conversation was a long time ago (about 30 years). I’d have to dig out the book to give you the guy’s name.
A few times I’ve come across Jumbles with more than one possible solution. Usually the alternate (“incorrect”) solution is something really obscure. One example that pops up from time to time is DEILY: the correct answer is assumed to be YIELD, but YLIDEis also a real word. A few months ago there was a somewhat more glaring example: CEHILS. The answer was supposed to be CHISEL, but IMO, CHILES (as in chile peppers) is equally valid.
I have a book from England with words from Chambers dictionary sorted by length and then by letters in the word a, b, c etc. Thus jumble would appear as
bejlmu
Very useful for anagrams in cryptic puzzle. Alas, it is in a box somewhere and a quick search on Amazon did not turn it up. I got it 40 years ago - I’d guess a version would be on line somewhere.
There are also Jumble books, btw. I’ve given a few to my wife at Christmas.
