Least effort generator

[oshfdk]

Some points:
- One can generate non VM words if tokens are chosen totally randomly. Some external rules are required. 
- Real VM words do not use much s2 and s3. Omitting them drops coverage from 75% to 66%.
- Impossible to draw line how much is template and how much is just random on the fly variation of scribes
- There are some words that have high occurrence but complicate the set, they need a token that is used mostly on that word. 

Hi!

This is certainly a possibility, the idea itself is not particularly new. Did you look into a way to generate statistically probable distributions or sequences of these tokens?

Also, there can be easy counterexamples. If you take Chinese written phonetically (pinyin), then overwhelmingly all words will be created using a combination of 1-3 simple tokens. It's likely the same is true for Vietnamese, which actually uses a script based on Latin characters. This doesn't mean that all Vietnamese books are just a few scribes randomly combining tokens.