Hey
Manas1212 -
The GMAT scoring algorithm is based off a measurement theory called "item response theory". Item response theory basically says that there are three things that should go into scoring a question - difficulty, the amount of information it gives about a student's ability level, and how hard it is to make an educated guess on the question. Because of that, it's really hard to equate the number of questions you got wrong on any given practice test to a score without knowing the relative importance and difficulty of each question. Some of the questions you got wrong on the 35 may have been experimental questions, or you may have only missed the harder questions on that attempt but missed much easier questions on another. (So for example, if you got a couple 700-level questions right but got many 500-level questions wrong, the fact that you missed those 500-level questions can be more devastating to your score than the 700-level questions are helpful.)
For a fuller picture of how IRT works in general,
here's an article I wrote about item response theory and
here is an article I wrote about how Computer Adaptive Testing works.
All that said, the best thing you can do is focus on doing as well as you can on every question that you see (but know when to guess and move on if you have no clue). Remember that because the GMAT is adaptive, it's going to feel hard no matter what -- just do your best on every question and let the scoring algorithm do its work.