Skip to content

Klausler

~`
!1
@2
#3
$4
%5
^6
&7
*8
<9
>0
{[
}]
K
(,
U
Y
P
W
L
M
F
C
?/
+=
|\
O
A
E
I
D
R
N
T
H
S
_-
Q
).
"'
:;
Z
X
V
G
B
J
Finger 123478910
Usage 7.25% (P88)8.21% (P51)12.76% (P80)12.49% (P44)15.76% (P84)11.55% (P22)6.93% (P7)7.82% (P58)
Same Finger Full Scissor Half Scissor Lat. Stretch
Bigram 1.62% (P83) 0.11% (P52) 2.77% (P91) 1.75% (P91)
Skipgram 5.55% (P79) 0.34% (P63) 3.16% (P88) 2.16% (P94)
No Thumbs Left Space Right Space
Weak-ish Redirs. 0.25% (P13) 0.28% (P5) 0.68% (P35)
Weak Redirects 0.37% (P53) 0.18% (P55) 0.18% (P55)
Other Same Finger 22.86% (P84) 14.89% (P84) 14.89% (P84)
Rolls : Alts 1.19 (P9) 1.39 (P19) 1.66 (P70)
2-Roll In : Out 0.66 (P3) 0.91 (P32) 0.71 (P0)
3-Roll In : Out 0.50 (P14) 0.90 (P30) 0.30 (P6)
Author
Peter Klausler
Year
2003
Finger Map
Traditional

Klausler, or Klausler’s Evolved Layout, is notable for being generated using an evolutionary algorithm other than simulated annealing, the standard for later layout optimization programs. It uses a tournament format in which randomly generated layouts compete with each other, and the top 1/4 layouts undergo random mutations and move on to the next round; the tournament ends when the top-performing layout has not changed for 50 rounds. The fitness function is also unique: it is derived from real-world keystroke speed data from an included keylogger, rather than using hand-selected heuristics as in most later optimizer programs.

The Evolved Layout is very similar to Dvorak, with the vowels and punctuation in a block on the left and nths on the right hand home keys, like Dvorak’s htns. Like Dvorak, it prefers alternation over rolls. However, it does not improve on any of Dvorak’s stats, besides a slight decrease in SFB%.