Hi, I have studied the feasibility of making a Japanese Scrabble set for a couple of time. As we know Japanese consists of Kanji, Hiragana and Katakana, and there are too many units to include in a scrabble set (which is as the same situation to Chinese. I made a latinised Cantonese set of tiles, but it stuck as latinisation in Cantonese is not common, no lexicon could be used). Using romaji is the most feasible and playable approach for Japanese Scrabble. Instead of using a kana unit(eg. no, shi, ka), I decided to make a latin set (A,E,I..) since Japanese kanji vocabularies are deeply influenced by Chinese vocabularies. Most words are only with 2 to 4 moras (which is equivalent to 3 to 8 letters and hence fit for a romaji set scrabble). Otherwise it would become a stair game like the Korean Hangul Scrabble. For instance, to play the word "sashimi" in kana way would become sa-shi-mi (only 3 tiles played). For loanwords, Japanese are used to simplifying the words like Television becomes te-li-bi.
Kana frequency has been considered as reference for tile distribution. The proportion of vowels must be higher than one-half.
Here is my 102-tile Japanese Romaji Scrabble set: Letter Value Amount A 1 11 B 4 2 C 10 1 D 8 1 E 1 5 F 10 1 G 4 2 H 2 5 I 1 11 J 6 1 K 2 5 M 5 2 N 1 10 O 1 12 P 5 2 R 3 3 S 2 5 T 2 5 U 1 12 (V 10 0) W 8 1 Y 3 2 Z 6 1 [- 8 1] Blank 0 2
As L,Q,X do not exist in Japanese, and V is only used in an extremely infrequent amount of loanwords, I decided not to include them. A hyphen (Choonpu) could be included depends on the method of romanisation (oo/uu, or o-/u-) C can only be played in Ch form (like Q in english). I valued 10 points though chi is very common. F can only be played in Fu form. But amount of U is highest in the tile set, thus there is no barrier to play the F (which is proved by simulation). This situation is similar to French -ez, Z is a very strong letter when grouped with E.
This is a 1-game simulation of romaji Scrabble (by using a 3000-entry lite lexicon)
Great work, although it has a Western look due to the Latin characters. But if it's playable...
Zitat von kaytse im Beitrag #1 ... I'm glad to receive any suggestions and full lexicon to test the playability and optimize the tile face values and frequency.
There was a thread for Japanese/Hiragana, which contains attachments of or links to dictionaries. No idea if the conversion would be easy or even possible.
Zitat von kaytse im Beitrag #1...V is only used in an extremely infrequent amount of loanwords, I decided not to include them.
That's quite common and usually realized per count=0, meaning you would have to utilize a joker to place a word with this letter. But it would be possible, in contrast to not include the letter at all.
Zitat von kaytse im Beitrag #1A hyphen (Choonpu) could be included depends on the method of romanisation (oo/uu, or o-/u-)
Another option is to understand it as a diacritical letter. You replace a single character like 1 by any combination such as O-. The dictionary contains of CH1NPU, the letter set has 1=8,1, and during the game this 1 is replaced by O- on a single tile. The help has some information on this topic (Spanish and Scottish Gaelic dics make use of this feature). This is likely to be relevant also for C=CH and F=FU.