Mary scored 81 on the chemistry test. Three scores averaging 82 add up to a total of 246 points (82 × 3), and Karen's 95 plus Jill's 70 account for 165 of those points. That leaves 81 for Mary. The arithmetic is settled in one line: 246 − 165 = 81. Anyone who splits the average three ways, or subtracts each known score from 82 separately, gets a different and wrong answer — which is exactly why this small problem trips up so many people.
That confusion is worth unpacking. Backward average problems like this one show up constantly in algebra workbooks, standardized-test prep (SAT, ACT, GRE quantitative sections), and classroom quizzes, and most explanations skip the step where errors actually happen. Below you'll find the full solution walked through, a verification pass, the mistakes that produce wrong answers such as 82 or 71, a few variants of the same problem, and a practice set with worked answers. The math is simple. The reasoning discipline is what transfers.
The total-first method solves the problem before any division
The reliable way to reverse an average is to sum first. Three test scores with a mean of 82 must total 82 × 3 = 246 points. Once you fix that total, the problem becomes ordinary subtraction: 246 − 95 − 70 = 81.
Work it in the other order and you're subtracting averages from scores, which is meaningless. Karen is 13 points above the mean; Jill is 12 points below it. Those deviations nearly cancel, leaving Mary one point below the mean, at 81. That's a useful sanity check, not a solving method.
Write the equation form once and it generalizes to any count of scores: sum of all scores = mean × number of scores. For this problem, 95 + 70 + Mary = 82 × 3. From there, Mary = 246 − 165 = 81.
If the total-first method feels like a trick, test it on a case where you already know the answer. Scores of 80, 80, and 80 average 80, and the total is 240. Scores of 70, 80, and 90 also total 240 and also average 80. The total carries all the information the average encodes. That's why the total is the right thing to reconstruct.
The full worked calculation, step by step
Here's the entire solution laid out in the order you'd write it on paper. Nothing here requires algebra beyond multiplication and subtraction, which is why the problem appears in middle-school curricula and on test-prep worksheets aimed at adults returning to quantitative material.
1: Multiply the average by the count of scores: 82 × 3 = 246. 2: Add the two known scores: 95 + 70 = 165. 3: Subtract the known total from the full total: 246 − 165 = 81. 4: State the answer with units: Mary scored 81 points.
Step 2 has a shortcut worth noticing. Because Karen and Jill's scores sum to 165, you can subtract them in either order — 246 − 95 = 151, then 151 − 70 = 81 — and land in the same place. Addition is commutative, so the order of subtraction of the known scores never changes the result.
The units matter more than they look. The 82 is points per test, and the 246 is total points across three tests. Writing "246 points" instead of a bare 246 keeps the arithmetic honest and prevents the classic error of treating the total as another average.
If any of the three scores were missing — say Jill's instead of Mary's — you'd run the identical three steps with different numbers. The method doesn't care which score is the unknown.
Checking the answer by rebuilding the average
Verification on backward-average problems means going forward again. Add all three scores: 95 + 70 + 81 = 246. Divide by three: 246 ÷ 3 = 82. The average matches the stated one, so 81 is correct.
This check catches sign errors and off-by-one slips that survive the first pass. If you'd answered 71, the check would give (95 + 70 + 71) ÷ 3 = 236 ÷ 3 ≈ 78.67, which fails immediately. If you'd answered 82, the check gives (95 + 70 + 82) ÷ 3 = 247 ÷ 3 ≈ 82.33, which also fails. The rebuild takes ten seconds and is the single highest-value habit in this whole exercise.
Do the check with exact division, not estimation. Averages of three scores land on repeating decimals whenever the total isn't divisible by three, and 82.33 registers as "close enough" to a tired eye. Write the fraction or compare totals instead: 247 ≠ 246 settles it without any rounding.
Checking also confirms you answered the right question. The problem asks for Mary's score, not the total or the deviation. A correct computation of 246 that gets reported as the final answer is still wrong.
Three wrong answers this problem reliably produces — and why
The most common wrong answer is 82. People see the average and assume it's also one of the scores, or they "split the difference" between 95 and 70 and stop at the midpoint, 82.5, then round. The midpoint of Karen and Jill's scores is genuinely 82.5, and it looks suspiciously close to the stated average, which makes the trap feel right. It isn't: Mary's actual score is 81, and the similarity is a coincidence of these particular numbers.
The second trap is subtracting each known score from the average: 82 − 95 = −13, then −13 − 70 = −83. Negative scores don't exist on most chemistry tests, so the absurdity is a signal to stop and reconsider the method. When a computation about test scores produces a negative value, the setup is wrong, not the test.
A third error appears when the count is misread. Three students means three scores, so the total is 82 × 3. Someone who divides 82 by three, or who assumes the known scores also averaged 82, gets a contradiction and often guesses. Count the scores, not the names.
The deeper pattern behind all three mistakes is operating on averages directly. Averages don't add or subtract meaningfully; totals do. Once you convert the average into a total, the problem has one obvious path.
A comparison table for every rearrangement of this problem
Rearranged versions of this problem show up constantly in practice sets, and the same total-first logic handles all of them. The table below works each variant with a different starting unknown so you can see the pattern hold.
| Known information | Question asked | Calculation | Answer |
|---|---|---|---|
| Average 82; Karen 95, Jill 70 | Mary's score | 82 × 3 − 95 − 70 | 81 |
| Average 82; Karen 95, Mary 81 | Jill's score | 82 × 3 − 95 − 81 | 70 |
| Average 82; Jill 70, Mary 81 | Karen's score | 82 × 3 − 70 − 81 | 95 |
| Karen 95, Jill 70, Mary 81 | The average | (95 + 70 + 81) ÷ 3 | 82 |
| Average 82; four students incl. 95, 70, 88 | Fourth student's score | 82 × 4 − 95 − 70 − 88 | 75 |
| Four scores totaling 328; 95, 70, 82 | Fourth student's score | 328 − 95 − 70 − 82 | 81 |
Each row is the same problem wearing a different coat. The total-first move — multiply the average by the count, then work with the sum — handles every entry, including the last row, where the total is handed to you directly and the average never appears.
If you're preparing for a timed test, the practical takeaway from the table is that you don't need six methods. You need one, applied to six shapes of the question. Recognizing the shape is half the speed.
Mary's 81 is correct, but here's what an average like that can hide
The answer 81 is trustworthy arithmetic. The average of 82 is where judgment should slow down. A mean summarizes three scores with one number, and it conceals the spread completely: this group includes a 95 and a 70, a 25-point gap between two students in the same class on the same test.
That spread carries real information a teacher would want. A class whose scores cluster near 82 looks different from a class with a 95 and a 70 and whatever sits between them, even when the means match. Averages are sensitive to how scores are distributed, and the same mean can come from very different result patterns.
The mean is also not the only summary available. If the question had asked for the median instead of the average, the answers would diverge in interesting ways. These particular scores are all distinct, so no mode exists; with three values the median is the middle one, and the problem didn't specify it. When a teacher or grader says "average" without qualification, they almost always mean the arithmetic mean, but a test-prep problem that says "average" and means "median" is a classic trick.
For a rebuilt grade, though, the mean is exactly the right tool: a total of 246 points across three tests is the honest record, and dividing by three is the fair reconstruction for the missing student.
Why a 25-point gap makes Mary's result easy to second-guess
Karen's 95 sits 13 points above the average. Jill's 70 sits 12 below. Those two deviations nearly cancel, and Mary's 81 sits one point under the mean. That balance is why the "midpoint of 95 and 70" shortcut nearly worked earlier — and why people who use it land at 82.5 instead of 81.
Reconstruct Mary's position from the deviations directly and it's cleaner than subtracting totals. The full deviation set must sum to zero: +13 (Karen) − 12 (Jill) + Mary's deviation = 0, so Mary's deviation is −1, giving 82 − 1 = 81. That's a second, independent route to the same answer, and it's useful because it shows the 81 isn't a fluke of one subtraction — it's what the balance demands.
Deviation thinking also makes future problems faster. If a fourth student joined with an average that held at 82, their deviation would need to be zero — they'd have to score exactly 82. If the average needed to be 85, the required total would be 255, and you could calculate the new student's score in one step.
The 25-point gap itself is worth keeping in mind when the question says "on average." It's mathematically fine to summarize these scores at 82, but the summary hides exactly the difference a teacher notices.
The same idea at any scale, from three tests to a semester grade
Change the number of items and the method doesn't blink. A season of seven games, a quarter of four project scores, ten quiz grades: total = average × count, and the missing value is the total minus everything known. The arithmetic is identical; only the multiplier changes.
Take a worked case with a real scale. A student's quizzes average 86 across eight quizzes, so the total is 688. If the first seven quiz scores sum to 601, the missing eighth score is 688 − 601 = 87. Same three steps, different numbers. Test it on your own gradebook if you keep one, and the check works the same way.
The scaled version is where the method earns its keep. On a single three-score problem, guess-and-check nearly works. At forty scores, a backward-average problem built from partial records becomes unmanageable without the total-first structure. That's why the technique is taught early: it's the general case, and the three-student problem is a small demonstration.
For students, the practical takeaway is to record totals, not just averages. A term total survives 82 or 82.33 as a running average but not an average computed with an incomplete set. One accurate total prevents a lot of hidden trouble downstream.
Practice problems with answers you can check
Five variations below use the same logic and different numbers. Work each one forward, then verify by rebuilding the average.
1: Three tests average 76. First two scores are 81 and 68. Third score? Total 228; 228 − 149 = 79. 2: Four tests average 90. Scores are 95, 84, 92, and one missing. Missing score? Total 360; 360 − 271 = 89. 3: Three tests average 82. Karen has 95 and Mary has 81. Jill's score? Total 246; 246 − 176 = 70 — the same set as the main problem, read from a different angle. 4: Five quizzes average 84. Four known scores sum to 341. Fifth score? Total 420; 420 − 341 = 79. 5: Two tests average 82, but one is 70. The other score is 164 − 70 = 94. With only two scores, the missing score is just twice the average minus the known one.
Notice problem 5's shortcut: with exactly two scores, the unknown is either the average plus the gap — 70 + (82 − 70) × 2 = 94 — or, identically, 2 × 82 − 70. For three or more scores, no shortcut survives, and the full total is required. If you're checking your own practice work, verify every answer by re-dividing the full set, not just the subtraction.
What actually explains the wrong answers in these problems
Most errors on reverse-average questions aren't arithmetic errors at all. People add and subtract perfectly and still miss, because they compute the right total and then answer the wrong question. The arithmetic was never the hard part.
The reliable fix is a habit: write the meaning of the total before computing it. Say "246 points across three tests" out loud or in the margin, then ask what the question is asking for. The translation step is where the marks are lost, and the practice is where they're recovered.
Second, match the total to the count before anything else. The single most common count error is using the number of known scores instead of the number of students — with two knowns and three students, the multiplier is still 3, because the average describes all three. Write the count first and the average second, in that order.
Third, keep the verification pass from earlier. It costs seconds and catches both sign errors and wrong-question errors, which together account for most misses. The rebuilt average is a complete audit of the answer against the problem's own conditions.
This article is general educational information about arithmetic, not a diagnosis, treatment plan, legal opinion, or investment recommendation, and it promises no outcome. Anyone who needs academic advice should consult a qualified teacher or tutor, and anyone facing a crisis or emergency should contact local emergency or crisis services.
FAQ
Can Mary's score be negative in some versions of this problem?
Yes, when the known scores exceed the total. If the average were 60 for three students and the two known scores were 95 and 70, the required total is 180 and the two knowns sum to 165, leaving 15 — still positive. Push Karen to 98 and Jill to 78 and the total of 180 minus 176 leaves 4. Push further and the result goes negative, which tells you the stated average and the stated scores can't both be true — an inconsistency you should flag rather than compute through.
What if the problem says the average is 82 but asks for the total?
That's a direct multiplication, not a reverse-average problem: 82 × 3 = 246. The confusion between the total and the average is common enough that some problems deliberately test whether you can tell the two apart. The clue is the word "total" in the question versus "average" or "mean."
Is this the same as finding a missing number in a mean?
Yes, it's the classic missing-value-in-a-mean problem, and the procedure is identical at any scale. The missing score equals the mean times the count, minus the sum of the known scores. The only change as the numbers grow is more subtraction; the structure never changes, which is why one worked example, properly understood, covers the whole category.
Doesn't the median give a different answer?
Only if the problem asks for the median, which it doesn't. The median and the mean coincide here — the problem essentially asks for a score that makes 82 the mean, and median questions would need a full set of numbers and a different calculation. Averages are means by default; medians are specified separately.
What does the 25-point spread between 95 and 70 tell a teacher?
That the average hides variation. Two students who both scored 82 could be a steady class or a split class, and the 95 and 70 here show real spread. Teachers looking at grade data often want both the mean and the range; the mean alone, as this example shows, can describe a group loosely.
How do I avoid errors under time pressure?
Write the total first, say what it represents, then subtract. The most common error under time pressure is answering 82, which is simply the given average, and a written total makes that answer visibly absurd. Verification by re-dividing takes ten seconds and catches it, too.
Does the count ever include the missing student?
Always. The average of 82 describes all three students, including Mary, so the multiplier is three whether her score is known or not. Using a count of two — the known scores — produces the wrong answer 130 or similar nonsense.