least common letter combination?

I’m curious as to the least common letter combination with the lowest number of letters. I’m preferably interested in words, acronyms or initials. I assume that all the two, three, and probably four letter combinations are taken. I can’t think of a better way to find this out than by a search for all combinations on an internet search engine, although I realize that this would probably include letter combinations that don’t have any real significance.

Does anyone have an idea to the lowest number letter combination not in use?

Huh?

Perhaps if you gave us an example, I might understand what you’re looking for.

“gry” ? :wink:

Well, I’d post it for you, but then it would be used.

:smiley:

I couldn’t find any four letter combinations that didn’t show up on a search, so I went to five letters.

“qgqvq” doesn’t turn up any hits on Google. Of course, it will now show up in the near future. :slight_smile:

You searched for all 25[sup]4[/sup] combinations? PLEASE tell me you made a script that did this! (And please tell me how you did it!)
Back to the OP. I think it’s time to call in Skip Newhall, who did such a tremendous job in Another note regarding ETAOIN SHRDLU (near the bottom of the page)

uh, you gotta be careful, cuz a lot of those letter combinations (at least through my cursory look at google) are not in actual words in any language. some of them look like they are character combinations that turned up from foreign lettersets (where the ascii letters represent symbols unrelated to the Roman letterset); some look like binary data in ascii. And so forth.

There was a plot of two- and three-letter frequency ratios in an article in American Scientist about 15-20 years ago. There was a guy – and I wish I could remember his name – who simulated that joke about putting a large number of monkeys at a large number of typewriters and seeing what they generated. Only he used computers instead of monkeys at typewriters, picking kletter at random. They generated Random noise, mostly. Then he had the frequency with which a letter appeared be the frequency his computer “monkey” would type it, and things started to look more rational. So then he generated probability maps of pairs of letters and used those probabilities to generate new output. (He included periods and spaces. He generated the probability maps by counting the frequencies of two-letter groups in actual writing.) These “second-order” monkeys produced even better looking stuff. By the time you get to Fourth and Fifth order monkeys you’re getting strings of words, and Fifth order monkey output in a foreign languasge you don’t know looks like the Real Thing. This guy’s article has probability maps of 2nd and 3rd order probabilities. “THE” is a BIG peak.

Mneomic (sp) as in that film by William Gibson