Evaluation scale (including precision) is increased with the introduction of negative points. The range of theoretically possible points in the previous example is 100 + 25 = 125, and without the negative points it is 100 points. In the same example 50 points equal number 75 on the scale of 0 - 125, or 60%. Without negative points, 60% would equal 60 points. The range of passing grades without negative points is 40, while with negative points it is 50. Therefore it is bigger!
  It is clear that when negative points are appointed, it is in the best interest of the students to answers all the questions, contrary to conventional wisdom. On a concrete example, without answering the questions a student is not sure about, the number of points would be only 48, which is lower than the required score. Students are not completely ignorant about the right answer, even if they are not sure about it. Informed guessing is the reason they achieve greater number of points than just monkey score.

  If you don't use negative points, take into consideration randomly guessed answers when determining the passing scale.

  This can be achieved by correcting (calculating) the percentage of points (K) based on the percentage of points won in the test (P). If it is a test consisting of just one type of MC questions, where each question has "n" suggested answers, correction formula is relatively simple:

K = (P x n - 100)/(n-1)

For example, if you decide that 50% of questions should be correctly answered in order to pass the test, the minimum required score level will not be 50%. It will depend on the number of offered answers "n". The following table shows how:

Number of offered answers	Minimum score (%) for 50% deserved points
2	75
3	67
4	63
5	60

   Minimum required score to pass a test usually ranges from 50 -60% of correct answers (most often 55%).