🌐 中文 · 🌐 English · 🌐 Español · 🌐 Français · 🌐 العربية · 🌐 Русский · 🌐 Deutsch · 🌐 日本語 · 🌐 Tiếng Việt · 🌐 한국어

Coding Statistics & Analysis

The current code table contains 308,453 distinct characters and words: 8,214 single characters, 189,571 two-character words, 74,312 three-character words, and 36,356 words of four or more characters. They correspond to 390,046 coded candidate entries and 184,055 code groups in use.

Dictionary Size and Coding Space

The 26 letters form the space of one- to four-key codes, with 26ⁿ letter combinations for each code length. Candidate positions allow several characters or words to share a code, but do not increase the number of distinct letter codes. By default, one-, two-, and three-key codes each retain five candidates, and four-key codes retain ten; full codes for single characters are always retained. The table below lists the total candidate positions at these default capacities.

Letter-Code Space and Default Candidate Capacity
Code lengthLetter codesDefault candidates per codeTotal candidate positions
One-key265130
Two-key67653,380
Three-key17,576587,880
Four-key456,976104,569,760
Total475,254—4,661,150

Candidate Counts and Utilization by Code Length

The following table counts coded entries by code length, candidate position, and character or word length. The first candidate is the preferred entry in each code group in use. Among first candidates for four-key codes, only entries with no other candidates under the same code meet the condition for direct commit with four keystrokes. Positions two through five and six through ten are listed separately to help assess candidate selection and paging needs.

Coded Candidate Entry Counts
Code lengthCandidate positionSingle charactersTwo-character wordsThree-character wordsWords of ≥4 charactersTotal
One-keyFirst candidate2600026
One-keySecond through fifth104000104
Two-keyFirst candidate60175117676
Two-keySecond through fifth1,2288783572412,704
Three-keyFirst candidate2,7957,9453,7922,74017,272
Three-keySecond through fifth2,64531,78117,47711,03462,937
Four-keyFirst candidate5,34596,54137,88826,307166,081
Four-keySecond through fifth1,28881,06331,8949,436123,681
Four-keySixth through tenth9311,7184,21454016,565
Total14,125229,93395,67350,315390,046
Candidate-Position Utilization by Range
Code lengthFirst candidateSecond through fifth candidatesSixth through tenth candidatesAll positions
One-key100.00%100.00%—100.00%
Two-key100.00%100.00%—100.00%
Three-key98.27%89.52%—91.27%
Four-key36.34%6.77%0.72%6.70%

Code Collisions and Conditions for Four-Key Direct Commit

A single-candidate group in the table below is a code group containing only one candidate. Its percentage uses the number of code groups in use at that code length as the denominator. The first candidate in a collision group still needs to be distinguished from other candidates; these first candidates cannot all be counted as direct commits.

Code Collisions by Code Length
Code lengthCode groups in useSingle-candidate groupsSingle-candidate shareCollision groupsMaximum candidatesEntries
One-key2600.00%265130
Two-key67600.00%67653,380
Three-key17,2725032.91%16,769580,209
Four-key166,081104,82663.12%61,25510306,327

Frequency-Weighted Minimum Keystroke Estimates

Frequency-Weighted Keystrokes Estimated from the Shortest Path for Each Character or Word
CategorySingle charactersTwo-character wordsThree-character wordsWords of ≥4 charactersOverall
Keystrokes per character or word2.663.713.973.992.95
Keystrokes per output character2.661.851.320.992.23

Key Distribution for Basic Single Characters

The table below summarizes the codes of all single characters. The counts across keys and code positions are fairly evenly distributed. In the Total column, the keys in the 米/人 main area, “EDC—SDFG—HJKL—I”, have the highest counts, as intended by the design.

Form-Code Positions and Primary-Pronunciation Supplements for Basic Single Characters
KeyFirst radicalSecond radicalThird radicalFourth radicalPhonetic supplementTotal
A2071011073436485
B130173195134217849
C1794443921142521,381
D2384434672131941,555
E476368342205351,426
F4826154501241661,837
G3354553153212161,642
H600339150462001,335
I39645134520601,398
J3853512741513811,542
K4013935892821411,806
L3114004212182991,649
M3842132471741861,204
N2732922441171061,032
O26926415811210813
P2092652332801531,140
Q3081641971792341,082
R37237820966681,093
S296359397952841,431
T322269533175822
U238328108140688
V54216820822601,144
W2843451961231551,103
X20815618567294910
Y19825712121398995
Z17173123113376856
Total8,2148,0646,7263,6384,57631,218

Overall Coding Assessment

The analysis shows the following advantages of the Fengshui Form-Phonetic Input Method:

  • Large dictionary: 8,214 single characters and 300,239 words, totaling 308,453 distinct characters and words, cover the vast majority of everyday input needs.
  • High short-code utilization: one- and two-key candidate positions are 100% utilized. For three-key codes, first-candidate utilization is 98.27%, and utilization of positions two through five is 89.52%, making good use of the short-code space.
  • Frequency priority: high-frequency characters and words take priority.
  • Keystroke efficiency: the frequency-weighted minimum is 2.66 keystrokes for single characters and 2.95 for all characters and words. High-frequency characters and words are assigned short codes.
  • Even key distribution: the 26 letter keys are fairly evenly distributed across code positions, with the main area (EDC—SDFG—HJKL—I) used most, in line with the ergonomic design.