Find what percentile a target value occupies in a dataset. Understand where a specific value ranks relative to all data points.
Last updated: March 2026
Percentile rank answers the question: "What percent of the data falls below (or at) my target value?" It converts a raw value into a percentile position ranging from 0 to 100. This is the inverse operation of finding a percentile—instead of asking "what value is at the 75th percentile?", you ask "what percentile is the value 85?"
The formula accounts for ties (equal values) by placing them at the midpoint of their rank range: Rank = [(Below + 0.5 × Equal) / N] × 100. For example, if 6 values are below your target, 2 equal it, and 2 above it in a dataset of 10, the rank is [(6 + 0.5×2)/10]×100 = 70th percentile.
Percentile rank is ubiquitous in real-world applications: standardized test scores ("your score is in the 92nd percentile"), growth charts for children ("baby's weight at 60th percentile"), employee performance ("top 10% of sales = 90th+ percentile rank"), and income distribution analysis. It provides intuitive, robust ranking that isn't affected by extreme outliers.
Sales Performance Ranking
Opposite of percentile. Percentile: 'What value is at P75?' vs Rank: 'What percentile is value 85?' Percentile rank tells you what percentage of data falls below your target value. Result is a percentage from 0 to 100.
Rank = [(#below + 0.5 × #equal) / N] × 100. The 0.5 × equal term handles ties by averaging their rank positions. Gives continuous percentile (0-100) rather than discrete ranks.
Still works! Calculator counts values strictly below your target. Example: value=65 in [10,20,...,100], 6 values below → rank = (6/10)×100 = 60th percentile. Value doesn't need to exist in data.
Percentile: 'What's the 75th percentile value?' (answer: specific number). Rank: 'What's the percentile of value 85?' (answer: percentage). They're inverse operations—one gives values, the other gives positions.
For ties, average their rank positions. If 3 values tied occupy ranks 8, 9, 10, their average rank is 9. Formula: (below + 0.5×equal)/N positions them fairly at midpoint of their tied range.
No. Rank always ranges [0, 100]. Value above all data → rank ≈ 100. Value below all → rank ≈ 0. Boundaries: minimum value ≈ (0 + 0.5)/N %, maximum ≈ (N - 0.5)/N %.
Test scores: 'Your score is 92nd percentile' (better than 92%). Growth charts: 'Baby weight at 60th percentile.' Employee performance: 'Top 10% of sales' (90th+ rank). Income distribution, benchmarking, rankings.
All get rank = 50%. Each value occupies the same average position. Logically: 0% strictly below, 100% equal, 0% above → (0 + 0.5×100)/100 = 50th percentile for all values.
Related Tools