On June 28, 2004,in Singapore, Kimberly Yeo beat 124 othter contestants and broke the world record for speed-of-text entry on a cell phone. In 43.66 (some sources cite 43.24) seconds she entered the following into her phone:

    The razor-toothed piranhas of the genera Serrasalmus and Pygocentrus are the most ferocious freshwater fish in the world. In reality they seldom attack a human.

She was timed in person by officiates from SingTel. Her speed beat out the May 06, 2004 record of 67 seconds as set by Briton James Trussler.

This phrase is fixed. If you want to spend your your 15 minutes on this particular renown, you have to use this exact text. Being a sucker for linguistic constants of most any sort, I became curious about the particular text. To answer my curiosity I applied some reason, built a quick program, and emailed GWR. Here's what I found out.

Why English?

The company Guinness World Records, which is the keeper of such things, was originally an Irish company.

Why this particular phrase?

It's the right length.
When comparing speeds of things (say, for example, cars) the larger the task to be accomplished (length of the race track) the more pronounced differences in speeds become. Though there are often upper complexity boundaries (set the finish line too far away and you have to stop for gas, adding logistics of a pit crew to the variables, no longer comparing just speed) it's a good idea for ease-of-measurement to maximize the size of the task.

Since SMS is limited to 160 characters (140 in some systems for goofy reasons, but 160 in actuality) and this doens't cross any complexity boundary, the phrase should be 160 characters. This is the length of the phrase.

It's not easy.
Once you've maximized the length of the SMS, the next dimensions of control is difficulty, both cognitive and physical.

It's cognitively difficult without being impossible.
If contestants don't speak the language or the phrase was gibberish, I suspect it would be too difficult to memorize a string of 160 nonsensical characters in perfect order. Speaking the language makes it easier, but you still have to memorize it correctly. That's made more difficult by the stumbling blocks of "Serrasalmus" and "Pygocentrus," uncommon words that have ambiguous phonemes. If you heard it and weren't a taxonimist, you might guess them as "serrosalmis" or "pigocentris." These two have to be memorized.

It's phycially difficult without being frustrating.
This one takes a little explanation. Most e2 readers at the time of this writeup will be quite familiar with this system, but for progeny, for whom this will seem laughable, I'll be specific. SMS communications give access to 52+ characters using 12 keys. While there are a number of ways to map this problem, the first way devised (and the only way permitted by Guinness for this world record) is the familiar multi-press system. i.e., one key maps to several characters, and each time you press a key the character about-to-be-inserted changes to the next in the cycling list. Following is a standard layout for such systems.

    1: 1,".","!","?",",","$","/","#","@", "-"
    2: a, b, c, 2
    3: d, e, f, 3
    4: g, h, i, 4
    5: j, k, l, 5
    6: m, n, o, 6
    7: p, q, r, s, 7
    8: t, u, v, 8
    9: w, x, y, z, 9
    0: <space>, 0

How does the device know when you're done selecting a character? If the user presses another key, then it can presume that the last selected character was correct and move on. But if the user needs to use the same key for a new character, there must be some delimiter, some way to say "I am done with that one, now I want to enter another one." While you could assign a "select" key, cell phone designers wisely minimize the number of keys (the explanation for why this is wise is beyond this already bloated scope) and instead rely on time. If the user does nothing for a set period of time, say, 1 second, then the currently selected character is inserted and a new press of the same key will append a new character.

While this works OK in most SMS circumstances, in a contest of text entry speed, that one second must be wildly frustrating. There is little visual feedback to let you know when the second is up. If you guess too soon, you screw up your last letter, which requires that you fix it and another second delay. If you wait too late, you've wasted precious seconds. Also, the more forced pauses there are in the phrase, the shorter your effective task length is, which is bad for reasons mentioned above.

LeoDV notes that rather than wait the 1 second you can press any other key which you then immediately delete, avoiding the frustrating pause. This may be the method Kimberly used, but I can't find an email for her to confirm. In any case, even if GWR relies on this method, they probably want to minimize its occurance...

So you'd like a phrase that minimzes the number of these maddening pauses. Curious about how this phrase stacks up and not wanting to actually enter the damned thing myself, I wrote a little program to translate it from the English to the equivalent key presses. Note that the phrase is mixed-case, and the caps-mode-key must be taken into account. I presume a one-key mode shift, marked in the output as an asterisk. Given that, the actual keys pressed look like this:

(note that for formatting reasons, I've added some arbitrary line breaks)

In the 160 characters, there are only 3 pauses. That's some swanky keystroke engineering.

Guinness confirms (kinda)

To see if I was correct, I emailed Guinness and asked them (though certainly not with this much detail). Following is their reply.
    Thank you for your recent enquiry.

    We choose this as our text to send as our researchers believe that this shows a degree of texting difficulty. Most of the characters are used, there are button repetitions and this also fits with the majority of handsets. Previously, we had several different texts, but we now liken this to people preparing for a 100m record and they can practise as much as they like before the attempt.

    I hope this answers your query.

    Yours sincerely,

    Scott Christie
    Guinness World Records

So, the source confirms some of what I was guessing. Back to Kimberly, presuming she had everything memorized and made NO mistakes, Kimberly was able to make 332 key presses in her 43.66 seconds. Including 3 pauses at a conservative 1 second each, that means she was pressing 8.1653 key presses per second. That's 8 Hz.

For comparison, that's the speed at which a woodpecker pecks.

Anyway, all these requirements, which do not concern the content of the phrase at all, result in this charmingly alliterative mantra. The piranha must be proud.

Log in or register to write something here or to contact authors.