What-If 0075: "Phone Keypad"

[rmsgrey]

I can type neither "We reserved seats at a secret Starcraft fest." using only my left hand, nor "You look upon my mom in Honolulu!" with my right hand only. However, I can type "We reserved seats a a secret Starcraft fest!" (no quotes) using only my left hand, and "You look upon my mom in Honolulu." (with or without quotes) using only my right hand. * Punctuation matters!

Also, the top row definitely does not contain the comma or period.

* - the right-hand one is easier because I get to use the backspace!

UK keyboards have the quotation marks as shift-2 - a left hand character (shift-' is @)

[Patrik3]

I use one of those old phones where you type with numbers—for example, to type "Y", you press 9 three times. Some words have consecutive letters on the same number. When they do, you have to pause between letters, making those words annoying to type. What English word has the most consecutive letters on the same key?

When I read the first half of the question, I had thought up a couple different questions to the one this guy actually asked...:

- What English word requires the most key presses in total?

It's probably just the same as "What is the longest word in the English dictionary?" but if there's any word which happens to have most of its letters being "C", "F", "I", "L", "O", "S", "V", "Z" in it, then it might overtake the longest word.

That said, "pneumonoultramicroscopiscilicovolcanoconiosis" already has a surprisingly high density of these letters.
(I listed 8 letters of 26 in the alphabet, and this word contained 29 of 45 letters. I know that not all letters are used exactly as much as each other, so my expected density of 0.308 wasn't very accurate, but even so, 29/45 is 0.644 - over twice my expected value.)

A1 B2 C3 D1 E2 F3 G1 H2 I3 J1 K2 L3 M1 N2 O3 P1 Q2 R3 S4 T1 U2 V3 W1 X2 Y3 Z4

So for Pneumonoultramicroscopicsilicovolcanoconiosis,

1 2 2 2 1 3 2 3 2 3 1 3 1 1 3 3 3 3 4 3 3 1 3 3 4 3 3 3 3 3 3 3 3 3 1 2 3 3 3 2 3 3 4 3 4 ... I'm sure if I had the tiniest amount of programming know-how, I could have automated that process.

Also it's probably really easy to sum this sequence of random numbers, but apart from re-entering them all into a spreadsheet I wouldn't know how, and I can add them in my head pretty easily, but I apologize if I make a mistake.

... My answer is 118?

So average key press density per letter is 118/45 = 2.622

Expected key press density...

1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 4 1 2 3 1 2 3 4 SUM/26 = 6 6 6 6 6 10 6 10 SUM/26 = (36 + 20)/26 = 2.154

So it's not double the expected key press density, but it's still quite a way above it.

I'll just check how "Antidisestablishmentarianism" stacks up against it...

1 2 1 3 1 3 4 2 4 1 1 2 3 3 4 2 1 2 2 1 1 3 1 2 3 4 3; total = 60... did I do that right? I can only count 27 numbers there... crap.

Oh well. Assuming that I missed one out, (60 + 2.154)/28 = 2.220 - which is still higher than the expected density, but only just. Pneumonoultramicroscopicsilicovolcanoconiosis is far beyond the expected density, and is 17 letters longer than Antidisestablishmentarianism, so I doubt any word would ever beat it.

----------------

The other question I thought of was "What sentences are the most annoying to type?" - which ones have the highest keypress density. I have no idea how I'd figure that one out without using computers, though.

[davidstarlingm]

The greater annoyance is needing to wait between key presses. Like, if you have a "ghi" or "rs" or "ff" sequence in your word, where you have to pause for a second after selecting the first letter before it automatically moves into the next space.

[mathmannix]

I think when I used to have a phone like that, I would skip the waiting time by hitting the right-arrow? I suppose if I were a "professional" texter, I would just learn to think in words without numerical repeats, such as "maybe" instead of "might", or whatever.

[Chicagojon]

Interesting...I never realized I type 15-20% faster with my right hand than my left, but random online typing tests confirm it. B is a left hand letter for me and Y is right hand.

[rmsgrey]

"highlight" is pretty annoying to text.

[kcameron]

Being the obsessive sort, I had to test the results myself. Here's my perl script:
#!/usr/bin/env perl

use strict;

my %words;
while (<>) {
chomp; s/\015//g; # get rid of line terminators (including dos formatted files)
$_ = lc; # force lower case
next if /[^a-z]/; # skip words with non-alpha characters (e.g. hyphens)
next if !$_; # skip blank lines
next if $words{$_}; # ignore dups

# map word into raw key info
my $pcnt = $_;
my $kseq = $_;
$kseq =~ tr/a-z/22233344455566677778889999/; # key pressed for each letter
$pcnt =~ tr/a-z/12312312312312312341231234/; # key press count for each letter
# abcdefghijklmnopqrstuvwxyz

# create wait counts
my @rep_raw;
@rep_raw = $kseq =~ m/(.)(\1+)/g; # order pairs of first key press followed by repeats
my $odd = 1;
my @rep = grep $odd = !$odd, @rep_raw; # list of repeats
my $totw = length (join "", @rep); # total waits for word
my @waits = sort (map length, @rep);
my $maxw = $waits[$#waits] || "0"; # maximum waits before moving to new key

# total key presses
my $kcnt=0;
map $kcnt+=$_, (split //, $pcnt);

# produce qwerty keyboard stats
my $is_homerow = m/^[asdfghjkl]+$/;
my $is_lefthand = m/^[qwertasdfgzxcvb]+$/;
my $is_righthand = m/^[yuiophjklnm]+$/;

# save info about the word
$words{$_} = {
'kseq' => $kseq,
'pcnt' => $pcnt,
'wlen' => length,
'kcnt' => $kcnt,
'maxw' => $maxw,
'totw' => $totw,
'file' => $ARGV,
'is_homerow' => $is_homerow,
'is_lefthand' => $is_lefthand,
'is_righthand'=> $is_righthand,
'all' => 1,
};

#printf "$_,keys:$kseq,cnts:$pcnt,len:%d,kcnt:$kcnt,maxw:$maxw,totw:$totw,rep:@rep,is_homerow:$is_homerow,is_lefthand:$is_lefthand,is_righthand:$is_righthand\n", length;
}

my $wkey;
sub by_wkey {
$words{$b}->{$wkey} <=> $words{$a}->{$wkey} || # note reverse numeric sort built-in
$a cmp $b # break tie with word alpha sort
;
}
sub proc_wkey {
my ($msg, $wcond);
($wkey, $wcond, $msg) = @_;

# extract the set of words matching the max for the key
my @ord_words = grep $words{$_}->{$wcond}, sort by_wkey keys %words;
my @max_words = (shift @ord_words);
my $max_val = $words{$max_words[0]}->{$wkey};
while (my $word = shift @ord_words) {
my $val = $words{$word}->{$wkey};
last if $val < $max_val;
push @max_words, $word;
}
push @max_words, @ord_words[0..4]; # add a few close words for fun

# Output info for all the words which equal the max value
printf "$msg:\n";
foreach my $word (@max_words) {
printf " %-50s: length:%2d, total key presses:%3d, key waits on one key: %3d, total key waits:%3d, file:%s\n", $word, $words{$word}->{wlen}, $words{$word}->{kcnt}, $words{$word}->{maxw}, $words{$word}->{totw}, $words{$word}->{file};
}
}

printf "%d unique words were examined\n", scalar (keys %words);
proc_wkey ("wlen", "all", "Words with most letters");
proc_wkey ("kcnt", "all", "Words with max total key presses");
proc_wkey ("maxw", "all", "Words with max one key wait");
proc_wkey ("totw", "all", "Words with largest total wait");
proc_wkey ("wlen", "is_homerow", "Longest words on QWERTY home row");
proc_wkey ("wlen", "is_lefthand", "Longest words on QWERTY left hand");
proc_wkey ("wlen", "is_righthand", "Longest words on QWERTY right hand");

0;

Here's the results:
localhost:~/try/keypad%./countkeys.pl wordsEn.txt corncob_lowercase.txt /usr/share/dict/words
418034 unique words were examined
Words with most letters:
pneumonoultramicroscopicsilicovolcanoconiosis : length:45, total key presses:118, key waits on one key: 3, total key waits: 6, file:/usr/share/dict/words
cyclotrimethylenetrinitramine : length:29, total key presses: 65, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
trinitrophenylmethylnitramine : length:29, total key presses: 62, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
antidisestablishmentarianism : length:28, total key presses: 61, key waits on one key: 1, total key waits: 1, file:wordsEn.txt
hydroxydehydrocorticosterone : length:28, total key presses: 68, key waits on one key: 1, total key waits: 3, file:/usr/share/dict/words
electroencephalographically : length:27, total key presses: 60, key waits on one key: 1, total key waits: 3, file:/usr/share/dict/words
Words with max total key presses:
pneumonoultramicroscopicsilicovolcanoconiosis : length:45, total key presses:118, key waits on one key: 3, total key waits: 6, file:/usr/share/dict/words
hydroxydehydrocorticosterone : length:28, total key presses: 68, key waits on one key: 1, total key waits: 3, file:/usr/share/dict/words
hydroxydesoxycorticosterone : length:27, total key presses: 68, key waits on one key: 1, total key waits: 4, file:/usr/share/dict/words
cyclotrimethylenetrinitramine : length:29, total key presses: 65, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
philosophicopsychological : length:25, total key presses: 64, key waits on one key: 1, total key waits: 5, file:/usr/share/dict/words
microspectrophotometrically : length:27, total key presses: 63, key waits on one key: 1, total key waits: 4, file:/usr/share/dict/words
Words with max one key wait:
deedeed : length: 7, total key presses: 11, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
fiddledeedee : length:12, total key presses: 23, key waits on one key: 6, total key waits: 7, file:/usr/share/dict/words
homonomous : length:10, total key presses: 24, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
mononomial : length:10, total key presses: 22, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
mononomian : length:10, total key presses: 21, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
nonmonogamous : length:13, total key presses: 28, key waits on one key: 6, total key waits: 7, file:/usr/share/dict/words
nonmonogamously : length:15, total key presses: 34, key waits on one key: 6, total key waits: 7, file:/usr/share/dict/words
nonmonopolistic : length:15, total key presses: 37, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
nonmonotheistic : length:15, total key presses: 35, key waits on one key: 6, total key waits: 6, file:/usr/share/dict/words
asarabacca : length:10, total key presses: 20, key waits on one key: 5, total key waits: 5, file:/usr/share/dict/words
bacaba : length: 6, total key presses: 10, key waits on one key: 5, total key waits: 5, file:/usr/share/dict/words
baccaceous : length:10, total key presses: 24, key waits on one key: 5, total key waits: 5, file:/usr/share/dict/words
caccabis : length: 8, total key presses: 20, key waits on one key: 5, total key waits: 5, file:/usr/share/dict/words
chrononomy : length:10, total key presses: 25, key waits on one key: 5, total key waits: 5, file:/usr/share/dict/words
Words with largest total wait:
gnomonologically : length:16, total key presses: 38, key waits on one key: 5, total key waits: 8, file:/usr/share/dict/words
nonaccommodable : length:15, total key presses: 31, key waits on one key: 3, total key waits: 8, file:/usr/share/dict/words
nonaccommodably : length:15, total key presses: 32, key waits on one key: 3, total key waits: 8, file:/usr/share/dict/words
nonaccommodatingness : length:20, total key presses: 43, key waits on one key: 3, total key waits: 8, file:/usr/share/dict/words
nonautonomousness : length:17, total key presses: 41, key waits on one key: 4, total key waits: 8, file:/usr/share/dict/words
nonmonarchically : length:16, total key presses: 38, key waits on one key: 5, total key waits: 8, file:/usr/share/dict/words
nontrigonometrically : length:20, total key presses: 46, key waits on one key: 3, total key waits: 8, file:/usr/share/dict/words
pseudomonocotyledonous : length:22, total key presses: 52, key waits on one key: 4, total key waits: 8, file:/usr/share/dict/words
ammonocarbonous : length:15, total key presses: 34, key waits on one key: 4, total key waits: 7, file:/usr/share/dict/words
automorphically : length:15, total key presses: 33, key waits on one key: 2, total key waits: 7, file:/usr/share/dict/words
cabbageheadedness : length:17, total key presses: 33, key waits on one key: 4, total key waits: 7, file:/usr/share/dict/words
cacodemonomania : length:15, total key presses: 30, key waits on one key: 4, total key waits: 7, file:/usr/share/dict/words
cannonballed : length:12, total key presses: 25, key waits on one key: 3, total key waits: 7, file:wordsEn.txt
Longest words on QWERTY home row:
haggadahs : length: 9, total key presses: 14, key waits on one key: 1, total key waits: 1, file:/usr/share/dict/words
falashas : length: 8, total key presses: 19, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
flagfall : length: 8, total key presses: 18, key waits on one key: 1, total key waits: 1, file:/usr/share/dict/words
galagala : length: 8, total key presses: 12, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
galahads : length: 8, total key presses: 14, key waits on one key: 0, total key waits: 0, file:wordsEn.txt
hadassah : length: 8, total key presses: 16, key waits on one key: 1, total key waits: 1, file:/usr/share/dict/words
Longest words on QWERTY left hand:
redrawerredrawers : length:17, total key presses: 36, key waits on one key: 1, total key waits: 4, file:/usr/share/dict/words
devertebrated : length:13, total key presses: 24, key waits on one key: 1, total key waits: 2, file:/usr/share/dict/words
tesseradecade : length:13, total key presses: 27, key waits on one key: 1, total key waits: 4, file:/usr/share/dict/words
aftereffects : length:12, total key presses: 28, key waits on one key: 3, total key waits: 3, file:wordsEn.txt
decerebrated : length:12, total key presses: 23, key waits on one key: 1, total key waits: 2, file:/usr/share/dict/words
desegregated : length:12, total key presses: 21, key waits on one key: 1, total key waits: 2, file:wordsEn.txt
Longest words on QWERTY right hand:
phyllophyllin : length:13, total key presses: 32, key waits on one key: 1, total key waits: 2, file:/usr/share/dict/words
miminypiminy : length:12, total key presses: 26, key waits on one key: 0, total key waits: 0, file:/usr/share/dict/words
hypolimnion : length:11, total key presses: 26, key waits on one key: 1, total key waits: 2, file:/usr/share/dict/words
hypophyllum : length:11, total key presses: 24, key waits on one key: 1, total key waits: 1, file:/usr/share/dict/words
kinnikinnik : length:11, total key presses: 26, key waits on one key: 1, total key waits: 2, file:/usr/share/dict/words
polyphonium : length:11, total key presses: 24, key waits on one key: 1, total key waits: 1, file:/usr/share/dict/words

I believe Randall got it wrong. It seems that the question was asking which English word is most annoying to type on a cell keypad based on the total number of times one must pause prior to starting a new letter. From my wordlists, the answer is "gnomonologically" which requires 8 pauses. "nonmonogamous" has only 7 pauses. Note that the longest word "pneumonoultramicroscopicsilicovolcanoconiosis" requires 118 key presses but only 6 pauses.

[ps.02]

Almost on topic:

maybe the spelling "masterbate" is so common because it is easier to type with one hand

That is all.

[willpellmn]

This was far and away the best XKCD of any variety for weeks, perhaps months. I laughed so hard I thought my eyeballs would explode.

[limentani]

Stewardesses is a cool word, but I believe there is a longer left-hand word: sweaterdresses.

Also, my favourite word to type is strewn, because it has four letters in a row on a qwerty keyboard. Apart from qwerty itself, which is now often recognised as a word, I think this is probably the most in any English word. It's very fun to type, makes me feel like a pianist!

[najodleglejszy]

Those old phones still exist?

yes, they do. at the end of 2012, 46% of all mobile phones in Poland were "dumbphones".
also: Stewart, why don't you use T9?

[PinkShinyRose]

Those old phones still exist?

yes, they do. at the end of 2012, 46% of all mobile phones in Poland were "dumbphones".
also: Stewart, why don't you use T9?

It's one of those old phones, it didn't have T-9 yet ?

I think when I used to have a phone like that, I would skip the waiting time by hitting the right-arrow?

I did that too, although, before I found this out I used to type a space and then immediately delete it (or was that to type compound words, and other words that could be formed using short parts of other words because of the very limited T-9 dictionary?).

[The Synologist]

Hey, quick question, how do you guys write "my"? I use my right index finger for the "m", and since it's so far out of position I just use my left index finger for the "y", even though that's not correct, since it would be much slower to bring my right index finger back up to strike both.

Experimenting a little further it looks like I use my left index finger for "y" about half the time.

[Pfhorrest]

Hey, quick question, how do you guys write "my"? I use my right index finger for the "m", and since it's so far out of position I just use my left index finger for the "y", even though that's not correct, since it would be much slower to bring my right index finger back up to strike both.

I do that as well. I never properly learned to touch type, I just gradually evolved from hunt-and-peck to a multi-finger hunt-and-peck with predictive hand positioning, so that as I'm moving to type one character, another finger (whichever could hit it fastest) is already approaching the next character. I'm easily the fastest typist I've met in person (nearly everyone who sees me type comments on my speed); it's been a long time since I've taken a typing test, but I seem to recall a 140wpm peak number (I don't recall my sustained rate but the number 90 is coming to mind). I believe in the end I end up very closely approximating standard touch typing, but there are differences, such as "my" here, where it's so much faster to let my left index finger be on its way to the Y while I'm still typing the M than to wait until I've finished hitting the M before I can even begin going for the Y. I would suspect that whenever there are two letters that "should" be hit with the same finger in succession, I probably diverge from the standard touch typing. I know my resting position is slightly different; my index fingers do rest on F and J, but my middle and ring fingers rest on E/I and W/O respectively, and my pinkies hover nonspecifically around the A/capslock/Q/tab and colon/quote/P/left-bracket areas.

Here, let me describe how I type "The quick brown fox jumps over a lazy dog.", and someone with proper touch typing training tell me where I diverge:
T - left index (with left pinky on left shift)
h - right index
e - left middle
space - right thumb
q - left ring
u - right middle
i - right ring
c - left index
k - right index
space - right thumb
b - right index
r - left index
o - right middle
w - left middle
n - right index
space - right thumb
f - left index
o - right middle
x - left middle
space - left thumb
j - right index
u - right middle
m - right index
p - right ring
s - left middle
space - left thumb
o - right middle
v - left index
e - left middle
r - left index
space - left thumb
a - left ring
space - left thumb
l - right ring
a - left ring
z - left ring -- interesting, this one has two letters in succession with the same finger
y - left index -- I'm not sure why I do this big reach with the same hand I was just using here instead of just using my right index, except maybe...
space - right thumb -- ...to enable me to do this, which in turn...
d - left index -- ...enables me to do this more easily
o - right middle
g - left index
period - right ring

That Y with my left index finger is driving me nuts now. I don't know why I do it. I can actually feel myself having to stretch to reach it. I just tried it with a bunch of other sentences, mostly ones that involve a Z-Y-space-D in them somewhere, and while it doesn't feel especially unnatural if I slow down and think "when I get to that Y I want to use my right hand", when I just go with the flow I end up using my left hand for the Y. I even use left shift with my left pinky while doing it. The only exceptions I can find are sentences with a lot of left-side action after it: something like "lazy days" gets me to use my right for the Y. I think maybe my brain is hesitant to bring my right hand over to the Y when it's expecting to need to use it to hit keys far to the right soon thereafter.

[Klear]

I try to learn touch-typing roughly once in a couple of years. Usually I get to the point where I master two full rows, but then I get some sort of transcribing job and before it's done, I revert to my weird standard style for speed.

[airdrik]

* - the right-hand one is easier because I get to use the backspace!

AFfter all, the backspace key sis nearly as essential to typing as the spacebar. (wWell, bmaybe closier to some of the medium-frequencely letters, though it depends on your accuaracy as to how much you use it in comparison to other keys).

[Copper Bezel]

I have Caps set to a second Backspace. Either hand will do, if I could only train myself to remember it's there.

[Sir Hotzenplotz]

Your supercomputer dials (9!)! What are the first 10 digits, and how long does it take given that every digit takes 1/3 second to dial? (paraphrased)

Hint: you can compute the number of digits and the phone number with most spreadsheet programs in under five minutes, if you do it right.

Do you mean one can solve the thing in 5 minutes, or did computers 10 years ago need about 5 minutes to run an appropriate script? (I really don't know..)

1-609-714-4004 is in Medford, NJ.
...
619978 s, a little over a week.
...
I can compute the same in Mathematica in about 5 seconds ☺

Wait, this is only the runtime, not including the time to write the script, right? Otherwise: what kind of huge values can Mathematica handle?

Besides that: Nice, I get the same values, that is a good sign : )
(9!)! = 1.609714400409960 * 10^1859933, computed in 0.66 seconds! (On a netbook!) I then 'streamlined' the code such that it used only 1/3 of the lines and became much more elegant, but then it took 5 times as long to run... Anyway:

Spoiler:

What I did was start multiplying the integers that make up the factorial, while checking the product doesn't exceed the double floating point limit. When (before) it does divide by 10^n, with n = floor(log10(product)). Sum all values for n to get the exponent (with base 10)

Is this the 'intended' solution method? Or are there other viable methods for computing this?

[Shidoshi]

I use a typing method similar to Pfhorrest's. It just evolved from hunt-and-peck after I learned some resting positions for my hands because of key-intensive MMOs.
I can type much better with my left hand, though. Mainly because Ragnarok Online would let you type while you went around with your character using the Mouse, so I'd chat while crossing empty deserts. That's probably part of the reason that my left hand keeps invading the space that should be the right hand's, not to mention that the right hand has to go for the End, Home, Arrow Keys and the Num Pad sometimes.

[davidstarlingm]

I did Mavis Beacon Teaches Typing in at least a half-dozen of its incarnations when I was in elementary school and junior high. Those games were the easiest way to get on the computer, haha.

The closest I've come to developing my own typing is the one-handed business. Haven't been able to come up with a consistent pattern.

[Klear]

I did Mavis Beacon Teaches Typing in at least a half-dozen of its incarnations when I was in elementary school and junior high. Those games were the easiest way to get on the computer, haha.

The closest I've come to developing my own typing is the one-handed business. Haven't been able to come up with a consistent pattern.

Best game to learn typing is this:

http://www.youtube.com/watch?v=sNfQ_B6_xy8

[rmsgrey]

I use a ~7 finger modified hunt-and-peck that's now, through long practice, both fast and accurate, and, if I'm not paying attention, even works without glancing at the keyboard more often than every few lines...

I was playing PC games before WSAD mouse-and-keyboard layouts became standard, so got very good at finding the right key without looking across a wide range of control schemes...

[Klear]

I was playing PC games before WSAD mouse-and-keyboard layouts became standard, so got very good at finding the right key without looking across a wide range of control schemes...

Ahhh... the good old days when I didn't even own a mouse and had to use alt to strafe in Doom and Wolfenstein. Vanilla Duke Nukem 3D is pretty much impossible to play conveniently with full mouse look. PgUp and PgDn was all you needed to aim up and down.

Of course, back then keyboards didn't have that dangerous windows key between ctrl and alt, so these keys were usually used for main actions like jumping and shooting. Though space is sometimes making a comeback as a use button.

I remember playing X-Wing and Tie Fighter with a joystick for steering, mouse for shooting (the stick's buttons were busted) and keyboard for everything else...

[dalcde]

I use a ~7 finger modified hunt-and-peck that's now, through long practice, both fast and accurate, and, if I'm not paying attention, even works without glancing at the keyboard more often than every few lines...

I was playing PC games before WSAD mouse-and-keyboard layouts became standard, so got very good at finding the right key without looking across a wide range of control schemes...

I can type with two fingers faster than most people do with all ten (albiet less accurately, since I don't do that a lot)

[Pfhorrest]

I should probably add that the "hunt" part of the "modified hunt and peck" I now do has atrophied entirely. I can type with my eyes closed almost as well as with them open, although somehow typing with them open still helps with accuracy, peripheral vision I guess. I typed that preceding sentence with my eyes closed at full speed and didn't correct anything afterward (though I just went back and added "with accuracy").

[davidstarlingm]

I should probably add that the "hunt" part of the "modified hunt and peck" I now do has atrophied entirely. I can type with my eyes closed almost as well as with them open, although somehow typing with them open still helps with accuracy, peripheral vision I guess. I typed that preceding sentence with my eyes closed at full speed and didn't correct anything afterward (though I just went back and added "with accuracy").

I had a huge hurdle at first trying to learn to type without looking at the keys. I was training all the right muscle memory but it was impossibly slow. Then I tried looking down after the muscle memory was all trained, just to get my bearings, and it became a thousand times easier as long as I glanced down now and then. Soon I didn't need to.

[najodleglejszy]

Those old phones still exist?

yes, they do. at the end of 2012, 46% of all mobile phones in Poland were "dumbphones".
also: Stewart, why don't you use T9?

It's one of those old phones, it didn't have T-9 yet ?

my first phone (Siemens C55) already had T9, I don't think he was talking about phones even that old.

[Kaiman]

Sometimes I'd rather have the old multiple key press version than T9, since at least you can just hit the keys accurately and know exactly what you typed.

We reserved seats at a secret Starcraft fest.

I'd say that uses the right hand twice - once to hit SHIFT for the W and once to hit SHIFT for the S. :p

[12obin]

This is the last thread I had posted to before yesterday.

[Logological]

I was pleased to see "What If?" tackle some good old-fashioned logology in this entry. I'm not sure if Randall knew, but people have been collecting and writing about unusual telephone and keyboard words for decades. It certainly seems to have sparked a lot of discussion on the forum here, so in case anyone's interested I thought I'd provide some pointers to older printed works on these subjects:

• In 1965 Dmitri Borgmann published Language on Vacation: An Olio of Orthographical Oddities, which was probably the world's first treatise on wordplay. The book includes a discussion of the longest words and sentences which can be typed on certain rows of a QWERTY keyboard, or with one hand. Back then Markov chaining wasn't well known (and without home computers was impractical to implement on the scale Randall uses) though Borgmann came up with some pretty respectable results. For example, for right-hand QWERTY he constructed, "In July, oh my killjoy Johnny, I'll look in upon my jumpy polo pony up in hilly Honolulu."
  
• Borgmann's keyboard recreations were expanded and improved upon by a succession of contributors to Word Ways, the recreational linguistics magazine that he edited for its inaugural year. These include Eugene Ulrich, who in 1983 constructed or reported some single- and alternating-hand paragraphs ("Typewriter Recreations"); Mike Keith (of "Near A Raven" fame) who in 1999 was probably the first to tackle the problem computationally ("Typewriter Words"); and Rex Gooch and Susan Thorpe who continued this work in 2000 ("Typewriter Words: All Fingers Used", "New Categories of Typewriter Words"). These authors cover such oddities as words which can be typed using a single finger, words where each finger is used at most once, words where the same finger is never used twice in a row, words whose letters progress in one vertical or horizontal direction across the keyboard, words where successive letters are on adjacent keys, and words with the lowest mean Euclidean distance between successive keys.
  
• I see there are some Dvorak enthusiasts in this thread, so I might point out my own contribution a few years back, which was to find the same sorts of oddities on the Dvorak keyboard: "Dvorak Typewriter Words"
  
• With respect to words typed on a (traditional) telephone keypad, this subject is also covered in Borgmann's Language on Vacation—he discusses records such as the longest words that can be dialled using only even or odd letters (NONCOMMUNICATING, WELL-EXPRESED).
  
• Telephone words have also cropped up in various math journals and puzzle magazines, including the Journal of Recreational Mathematics (e.g., "Telephone Synonyms", 1989).
  
• And again, Borgmann's investigation was expanded by various Word Ways authors, including A. Ross Eckler, Jr. ("Sending Messages by Telephone", 1995), Rex Gooch ("Telephone Words", 1996), and Mike Keith and Dave Morice ("Let Your Fingers Do the Wordplay", 1999).

Language on Vacation and the Journal of Recreational Mathematics are out of print and not available online, but the Word Ways archives are open-access for articles older than two years, so most or all of the ones linked to above should work.