Total Points vs Weighted Grades: Which Denominator Applies?

Total points use possible points as influence

Add earned points and divide by possible points. A 100-point assessment has ten times the denominator influence of a ten-point assessment. The point model does not treat every row equally unless the possible-point values are equal.

Explicit weights use the syllabus column

Convert each assessment to a percentage, multiply by its stated weight and normalize over the supplied weights. Maximum points are used to obtain the score percentage; they do not replace the separate course weights.

Compare one record both ways

For 9/10 and 35/50, total points give 44/60 × 100 = 73.333…%. Scores are 90% and 70%. If explicit weights are 20 and 80, weighted average is 90 × 0.20 + 70 × 0.80 = 74%. Equal-assessment weighting gives 80%. None of these methods should be switched midway to produce a preferred grade.

Know when they agree

The two methods agree when explicit weights are proportional to possible points. For the example, weights 10 and 50 normalize to one sixth and five sixths, matching the total-points denominator. This provides a useful cross-check rather than a rule that weights should always be entered this way.

Resolve categories before merging

A course can use points within each category and percentages across categories. Calculate each category’s earned/possible percentage first, then weight the category scores. Do not flatten the full gradebook into one points sum unless the grading policy says to do so.

Adding a row can change influence in different ways

Start with 8/10 and 30/50. In points mode the result is 38/60 = 63.3333…%. Adding 10/10 gives 48/70 = 68.5714…%, because the new assignment supplies only one seventh of the updated possible points. An equal-row model starts at the average of 80 and 60, or 70%, then becomes (80 + 60 + 100)/3 = 80%.

The improvement is different because the new row has different influence under the two models. Neither calculation changes the earned marks. If a tool returns the same result for every weighting pattern, that may indicate it is ignoring the selected weights.

Work inside categories before combining the course

Assume homework records 18/20 and 42/60. Its category percentage is 60/80 = 75%, not the equal-row mean of 90 and 70, which would be 80%. If that homework category has 30% course weight and an exam category scores 84% at 70% weight, final grade is 75 × .30 + 84 × .70 = 81.3%.

Flattening all possible points together would require knowing the exam’s possible marks and would apply a different implicit weighting. A course can intentionally make a small exam category influence the course more than its raw mark count suggests. The category label is therefore part of the grading record, not decorative text.

Missing, excluded and zero are not the same record

A missing 0/20 assignment counted by policy adds twenty possible points and no earned points. An excluded assignment adds neither. An ungraded future assignment may be absent from the current average but still belong in a final projection. State which case is intended before interpreting the displayed percentage.

Under explicit weights, dropping a category can also change normalization. Removing a 20% category and normalizing over the remaining 80% describes the surviving subset; it is not automatically the final course rule. The instructor might transfer the weight to another category instead. A generic average does not infer that redistribution.

Use denominator tests to resolve disagreements

Check whether doubling an assignment’s possible and earned marks changes the course result. In a total-points model, the row’s influence doubles while its percentage stays fixed. Under a fixed explicit weight, its influence stays unchanged. This thought experiment distinguishes the methods without relying on a tool label. Preserve the original gradebook rows, category rules, exclusions and full precision when reconciling a discrepancy; do not edit the marks to match a preferred answer.