When you find a word using this number system 91% of Welsh words are stripped out automatically. Then the image is compared to the Voynich word and it has a chance of 9% of the remaining Welsh dictionary. But the percentages are less than that, because the word is targeting an image and both the welsh word and the voynich word have to be equal. So in essence its narrowed down to 3% for a chance of a fit. Now in these posts you will find a Zodiac which all but one word is in Welsh and I don't know what the percentages of finding that fit is! So here it is again. However I will try .03x.12=.36% so I had a .36% chance of fitting a Zodiac in.
The people who are arguing against my method are not taking this into consideration. If all languages don't work based on rules to decoded the voynich than the argument for a number system is high for my method.
![[Image: 67r2welsh-recovered3.jpg?w=840]](https://voynichnumerology.files.wordpress.com/2016/09/67r2welsh-recovered3.jpg?w=840)
The people who are arguing against my method are not taking this into consideration. If all languages don't work based on rules to decoded the voynich than the argument for a number system is high for my method.