I would need your view on the Project Statistics feature.
I am positive a lot of terms of the texts I am working on are contained in my memory and glossary (I put it in .tmx format so that I can perform statistics), and I am very surprised as I am getting back a disappointing poor percentage of matches...
Has the same occurred to you? Any comments on why this is happening?
Thanks in advance. Looking forward to hearing your comments,
I've never experienced any problems with the match statistics. They seem to be as good as those of other CAT tools.
One of my clients sends me Trados match analyses regularly, and I've never experienced any big differences (only small ones, probably due to differences in segmentation). And what's more, the CT match percentages have never been lower than the Trados percentages (only slightly higher in some cases, which is good).
But I've only used the Statstics feature for segment matches. You seem to have used it for term matches. I've never done that, and I'm not sure whether the feature is intended for term match statistics at all.