Not much, you would say? So I thought, but for some strange reason this is the second time this particular word comes up as a 70% match when my source segment consists of a tag and a word "following", and one time it came up when I had a segment consisting of a tag and word "press".
The source word is in lower case, LIECHTENSTEIN in upper case. The source word had one tag or none, LIECHTENSTEIN has two. Does anybody have a sane explanation to this?
I should add that this happens when I use TE. And Liechtenstein in my TM has a corresponding translation, so I have not accidentally translated "following" or "press" as Liechntenstein in my TM.
The Trados matching algorthms are aberrant for short and/or heavily tagged segments.
The same occurs for short segments containing numbers.
It seems the tag/number weight is least two times greater than the word weight which gives at least strange results for a human.
It was discussed here.
For some kind of text (e.g. partlists), the Trados wodcount may be completely unreliable, so beware.
BTW, today I had some 70% matches like:
(bold)ESM (/bold): for (bold)Rappel :
Because in the TM I had (bold)ESM(/bold): for (bold)ESM (/bold):
You're not the only one