A fight statistic answers a particular question: how many counted strikes landed, what share of attempts connected, or how much output occurred per minute. It becomes misleading when that answer quietly changes into “who deserved to win” or “who will win next”. Here is a way to check the arithmetic and keep the conclusion within the evidence.
Start with the right counts
UFC's statistical definition includes strikes at distance, plus power strikes in clinch and ground exchanges. “Significant” is a category in the dataset; it does not mean every entry had the same effect. UFC athlete pages distinguish significant strikes landed from significant strikes attempted. Source: UFC's statistics definitions.
Use the Sig. str. totals for both fighters from the same bout. A row written as “27 of 106” means 27 landed from 106 attempts. Do not mix this column with total strikes, career averages or a single round. Record the source link and access date beside your notes so another reader can check the inputs.
The four calculations
- Accuracy: landed ÷ attempted × 100. The denominator includes both landed and missed attempts.
- Accuracy gap: A's accuracy minus B's accuracy, in percentage points. Keep the unrounded values until the final display.
- Net landed: A's landed count minus B's. Say which fighter comes first; a negative number means A landed fewer.
- Landed per minute: landed × 60 ÷ elapsed seconds. Use completed round time plus time elapsed in the final round, excluding breaks.
Percentage points and percent are different comparisons. In a hypothetical example, 60% accuracy versus 40% is a 20-percentage-point gap. The relative increase is 20 ÷ 40 × 100 = 50%. Neither number describes a probability of winning.
Example 1: more landed strikes, a different winner
Use Rob Font as A and Marlon Vera as B. The following figures come from the UFC Stats full-bout table; calculated values are rounded to two decimal places.
| Measure | Font (A) | Vera (B) |
|---|---|---|
| Landed | 271 | 159 |
| Attempted | 516 | 283 |
| Accuracy | 52.52% | 56.18% |
| Elapsed bout time | 25:00 | 25:00 |
| Landed per minute | 10.84 | 6.36 |
For Font, 271 ÷ 516 × 100 = 52.52%. The A-minus-B accuracy gap is −3.66 percentage points; net landed is 271 − 159 = +112. With 1,500 elapsed seconds, Font's rate is 271 × 60 ÷ 1,500 = 10.84. He landed more, while Vera connected on a larger share of his attempts.
The result on 30 April 2022 was a unanimous decision for Vera: 48–47, 49–46 and 49–46. Source: UFC's official result and scorecards. UFC's contemporary report describes Vera dropping Font in rounds two, three and four. Source: UFC's bout report. The counting comparison alone does not explain those moments or establish a scoring error.
Example 2: use the whole elapsed time
At UFC 217, Georges St-Pierre defeated Michael Bisping by submission at 4:23 of round three. The UFC Stats bout page records five-minute rounds and these significant-strike totals:
| Measure | Bisping (A) | St-Pierre (B) |
|---|---|---|
| Landed / attempted | 27 / 106 | 59 / 119 |
| Accuracy | 25.47% | 49.58% |
| Landed per minute | 1.88 | 4.10 |
Elapsed time is 5:00 + 5:00 + 4:23 = 14:23, or 863 seconds. St-Pierre's rate is 59 × 60 ÷ 863 = 4.10. Dividing by only 4:23 would incorrectly place all three rounds' strikes into the final round's time. The decimal-minute equivalent is 14 + 23 ÷ 60, not 14.23. Calculate accuracy from the raw counts rather than copying the source's whole-number percentage display.
What the worksheet leaves for you to judge
The fight-stat worksheet applies these calculations to the figures you enter. It rounds accuracy and gaps to one decimal; the tables here use two. A per-minute rate uses the entire elapsed bout time, including grappling exchanges. It is not a rate for time spent standing, attacking or in a particular position. Comparing two opponents in one bout also does not establish how either would perform against someone else.
Judging is a separate task. The ABC's July 2025 clarification evaluates rounds individually and prioritises the effect of legal offence, with damage the most valued component. That document is a scoring reference, not a retrospective re-scoring of these bouts. For practice, watch the action and record your own round assessments in the fan scorecard.
Before sharing a comparison, check three things:
- Missing is not zero. If attempts are unknown, leave accuracy uncalculated. With zero attempts, accuracy is undefined; with zero landed from a positive number of attempts, it is 0%.
- Keep the sample attached. Label one bout as one bout. A short appearance and a full decision offer different amounts of observation, even after conversion to a per-minute rate.
- Do not manufacture a conclusion. These calculations do not produce automatic round scores, a forecast, a measure of injury severity or a training prescription. A stronger claim needs evidence beyond the entries in this worksheet.
Correction and method note — 12 September 2026: This replacement removes the earlier article's unsourced “Fighter A” and “Fighter B” claims and promised analysis without identifiable bouts. The examples now use linked primary data and explicit arithmetic. Prepared with AI assistance using source checks; this is desk-based explanation, not ringside reporting or a claim of judging expertise. The original publication date is retained.