A decision tree for guessing under negative marking
Two options eliminated is not the same as three. A simple tree that keeps expected marks positive.
Guessing is maths, not courage
With +1 for a correct answer and −0.25 for a wrong one, a blind guess across 4 options has an expected value of exactly zero: one quarter of the time you gain a mark, three quarters you lose a quarter-mark. Blind guessing neither helps nor hurts — which means every elimination you can make turns guessing profitable.
Eliminate one option and the expected value goes positive. Eliminate two and guessing becomes one of the best decisions on the paper. The tree below makes this mechanical.
The tree
Can you eliminate all three wrong options? Answer with confidence. Can you eliminate two? Guess — the odds pay. Can you eliminate exactly one? Guess only if the paper rewards attempts or you are behind your target score; otherwise skip. Can you eliminate nothing? Skip, always.
The common leak is the third branch. Learners guess on gut feeling with one elimination and bleed a quarter-mark at a time. Over a 100-question paper, ten such guesses cost roughly a full mark after accounting for the hits.
Calibrate with exam mode
Knowing the tree is not the same as following it under pressure. Run full papers in exam simulation with negative marking switched on and review every guess afterwards: which branch were you really on? Most learners discover they were guessing on the one-elimination branch far more often than they believed.
What to do this week
Take one old mock and re-solve only the questions you got wrong. For each, write down which branch of the tree you were on. If more than a third were blind or one-elimination guesses, your next three practice sessions should be exam-mode only.
Practise it
Reading is not training
Every idea above is a drill inside UNLOCK — timed questions, instant solutions, and XP for doing it daily.
Start practising freeKeep reading
Related notes
How to use a question timer without panicking
Timers only help if they train a decision rule. Here is the three-step rule we teach for a 30-second clock.
Twelve speed-maths shortcuts worth memorising
Percentage change, successive discounts and ratio splitting — the shortcuts that save the most seconds.
A repeatable framework for seating-arrangement puzzles
Stop drawing ten circles and hoping. Order the clues by constraint strength and the grid fills itself.