• Practice
  • About 14 min
  • 5 practice questions

Break-even accuracy: the minimum hit rate that makes an attempt worthwhile

Derive the break-even accuracy for any illustrative marking scheme, compare it with your measured accuracy on doubtful questions, and decide when those attempts pay.

If something looks wrong, please tell us.

Expected value tells you what one kind of attempt is worth. A quicker test for everyday use is the break-even accuracy: the lowest accuracy at which a group of attempts stops losing marks. If your accuracy on a kind of question is above it, attempting pays on average; below it, attempting loses. One number per scheme, worked out once, guides every decision.

Deriving the break-even point

Let the mark for a right answer be R, the penalty be N and your accuracy on a group of questions be p. Each attempt earns p x R - (1 - p) x N on average. Set that equal to zero: p x R = (1 - p) x N, so p x (R + N) = N, which gives the rule below.

Break-even accuracy = penalty / (mark + penalty). Attempt a group of questions on average only if your accuracy on that group is above it.

Break-even accuracy for illustrative schemes
Scheme (illustrative)WorkingBreak-even accuracyWrong answers that cancel one right
+1 / -1/3(1/3) / (4/3)1/4 = 25%3
+1 / -1/4(1/4) / (5/4)1/5 = 20%4
+2 / -0.50.5 / 2.51/5 = 20%4
+1 / -1/2(1/2) / (3/2)1/3 = about 33.3%2
+1 / -11 / 21/2 = 50%1

Two patterns stand out. First, +1 / -1/4 and +2 / -0.5 have the same break-even point, because only the ratio of penalty to mark matters. Second, under +1 / -1/3 the break-even point is 25%, exactly the chance of a blind four-option guess, which is why blind guessing averaged zero in the last two lessons.

Worked example 1: Yash and his doubtful questions

Yash marks a small "D" next to every doubtful answer in his mocks. Over several mocks he has 40 doubtful attempts, of which 14 were right: an accuracy of 14 / 40 = 35%. What are they worth under three illustrative schemes?

  • +1 / -1/3 (break-even 25%): 35% is above it. Total = 14 - 26 x 1/3 = 14 - 8.67 = about +5.33. Keep attempting doubtful questions.
  • +1 / -1/2 (break-even about 33.3%): 35% is just above. Total = 14 - 26 x 1/2 = 14 - 13 = +1. Barely worth it; attempt them only if time is free.
  • +1 / -1 (break-even 50%): 35% is well below. Total = 14 - 26 = -12. Skip doubtful questions under this scheme.

The same student with the same knowledge should behave differently under different schemes. That is why you must work out the break-even point for your own exam's scheme, taken from its instructions.

Worked example 2: Nandini and a scheme stated as a fraction

Suppose Nandini's practice paper has 2-mark questions with one-third of the marks deducted for a wrong answer. Penalty: 2 x 1/3 = 2/3. Break-even: (2/3) / (2 + 2/3) = (2/3) / (8/3) = 2/8 = 1/4 = 25%. That is the same as +1 / -1/3, because the penalty is still one-third of the mark. Nandini realises she can remember a single rule for her paper: "attempt a group when my accuracy on it is above 25%", whatever the marks per question.

Using break-even accuracy well

  • Work it out from your exam's instructions before you start taking that exam's mocks.
  • Apply it to groups of similar questions (doubtful ones, one subject, one question type), not to your overall accuracy.
  • Treat results close to the break-even point as "no real gain": small samples and bad days can tip them either way.
  • Recalculate only if the scheme changes.

Key takeaways

  • Break-even accuracy = penalty / (mark + penalty).
  • Only the ratio of penalty to mark matters, so a fraction-based penalty gives one break-even point for all question values.
  • Compare it with your measured accuracy on a group of similar questions.
  • Close to the break-even point, an attempt gains almost nothing; prefer time for checking.

Check your understanding

  1. 1. Suppose +1 for right and -1/2 for wrong. What is the break-even accuracy?

    1. 25%
    2. About 33.3%
    3. 50%
    4. 20%
    Show answer

    Answer: B. (1/2) / (1 + 1/2) = (1/2) / (3/2) = 1/3, about 33.3%.

  2. 2. Suppose +4 for right and -1 for wrong. What is the break-even accuracy?

    1. 20%
    2. 25%
    3. 33.3%
    4. 40%
    Show answer

    Answer: A. 1 / (4 + 1) = 1/5 = 20%.

  3. 3. Yash is right on 35% of his doubtful questions. Under which illustrative scheme should he skip them?

    1. +1 / -1/3
    2. +1 / -1
    3. +2 / -0.5
    4. +1 / -1/4
    Show answer

    Answer: B. Break-even under +1 / -1 is 50%, and 35% is below it. Under the others the break-even point is 25% or 20%, which 35% clears.

  4. 4. Why do +1 / -1/4 and +2 / -0.5 have the same break-even accuracy?

    1. Both have a maximum of 100 marks
    2. Both have four options
    3. Both deduct less than one mark
    4. The penalty is one-fourth of the mark in both
    Show answer

    Answer: D. Break-even depends only on the ratio of penalty to mark: 0.25/1 and 0.5/2 are both 1/4, giving 1/5 = 20%.

  5. 5. Suppose +1 for right and -1/3 for wrong. Over many mocks you are right on 30 of 100 questions in one subject. What do those 100 attempts earn?

    1. +30
    2. -3.33
    3. About +6.67
    4. 0
    Show answer

    Answer: C. 30 right, 70 wrong: 30 - 70 x 1/3 = 30 - 23.33 = about +6.67. Accuracy 30% is just above the 25% break-even point.

Privacy · Terms · Refunds · Delivery