The asymmetric scoring of the KNX Basic exam (+1 for a correct answer, -0.5 for a wrong one, 0 for an abstention) makes it sensible to abstain on questions whose answer is completely unknown rather than to guess at random.

True. With a +1/-0.5/0 scheme, a pure random guess on a True/False question has an expected value of 0.5 x 1 + 0.5 x (-0.5) = +0.25, which looks positive but is highly volatile and easily wiped out by a streak of wrong guesses on borderline items. Abstaining yields a guaranteed 0, which is strictly better than the negative outcome that follows a wrong guess and protects the cumulative score on questions where you have no signal at all. The rule of thumb taught in KNX training centres is: guess when you can eliminate at least one option or when intuition is clearly biased, abstain when the topic is genuinely unknown.

Track your own hit rate on doubtful questions during practice — if you score worse than 60% when you guess, switch to systematic abstention on those items.