The CanUFight algorithm estimates a person’s realistic chances in an unarmed one-on-one fight by combining multiple measurable attributes into a single, population-weighted model.
Each input category was designed with two main objectives:
1. Quantify real-world influence on fighting ability.
For each attribute, I researched available data from scientific literature, combat-sports statistics, and physiological studies to estimate how strongly that factor affects physical fighting performance between otherwise equal individuals.
Using that evidence, I defined a numerical multiplier that represents the relative advantage or disadvantage the attribute provides.
2. Weigh the multipliers by global population distribution.
Once each attribute's independent effect was defined, I then modeled how that attribute is distributed across the world's population using demographic and health data from various sources such as the WHO, CDC, and global census reports.
Each multiplier is therefore population-weighted, reflecting not only how much an attribute matters, but also how common it is.
In short, the tool first estimates how much of an advantage each attribute provides by itself, then adjusts that advantage based on how often the attribute appears in the global population.
Sex selection applies a small multiplier to your base score:
| Sex | Multiplier |
|---|---|
| Male | x 1.05 |
| Female | x 0.95 |
Scientific data consistently shows measurable physiological differences between males and females that impact combat performance, primarily in strength, power, and endurance.
On average, males exhibit 50–60% greater upper and lower body strength due to differences in lean mass and fat distribution. Female athletes have about 5–10% more body fat and roughly 85% of the lean body mass of equally trained males.
Males also tend to possess a higher proportion of Type II (fast-twitch) muscle fibers and greater tendon stiffness, contributing to superior explosiveness and force output. Additionally, VO₂max and hemoglobin levels are typically 10–20% higher, improving oxygen transport and endurance.
These combined advantages translate to a modest but consistent edge in physical exchanges, even when other attributes are equal. To represent this gap, the multipliers that were chosen correspond to roughly a 55-45 performance advantage for males under otherwise identical conditions. This is large enough to reflect biological reality, but not so large that it overshadows training and athleticism.
Then, because each sex makes up about 50% of the world, those multipliers are halved (males would have a 10pt advantage over half of the world = 5pt change in percentile ranking).
This multiplier is gradually phased in based on biological development. For inputs under 13, the sex-based multiplier is disabled, since prepubescent children show minimal physiological differences in strength and performance. From 13 to 18, the effect scales progressively as puberty introduces measurable hormonal and muscular changes, reaching the full multiplier by adulthood.
https://pubmed.ncbi.nlm.nih.gov/39501696/
https://pmc.ncbi.nlm.nih.gov/articles/PMC7930971/
https://pubmed.ncbi.nlm.nih.gov/37424380/
https://pubmed.ncbi.nlm.nih.gov/40657230/
https://epublications.marquette.edu/cgi/viewcontent.cgi?article=1223
Human physical performance changes predictably over the lifespan due to growth, hormonal development, and age-related muscular decline.
The multiplier system reflects these well documented physiological trends, using distinct phases calibrated from exercise-science research.
It is then weighted according to the global population distribution, factoring in approximately how many people exist in each age group.
This allows for a realistic estimate of how any given individual would perform against a random opponent from the world population.
| Age Group | Population Share | Relative Ability |
|---|---|---|
| Childhood (0-12) | ~25% | 0.0× → 0.4× |
| Adolescence (13-17) | ~10% | 0.4× → 1.0× |
| Prime Adulthood (18-35) | ~25% | 1.0× |
| Post-Prime Decline (36-65) | ~30% | 1.0× → 0.7× |
| Elderly (65+) | ~10% | 0.7× → 0.4× |
Children have limited muscle mass and coordiation; near zero fighting capability rising gradually to ~40% of adult capacity by age 12. Puberty increases muscle size, hormones, power. Strength rises rapidly twoards full adult performance by 17-18 years. A gradual decline of 1% strength per year beyond 35 is applied due to sacopenia and reduced power.
https://www.health.harvard.edu/exercise-and-fitness/age-and-muscle-loss
https://pmc.ncbi.nlm.nih.gov/articles/PMC2804956/?
https://ourworldindata.org/grapher/population-by-five-year-age-group
https://www.visualcapitalist.com/mapped-the-median-age-of-every-continent/
Height indirectly affects fighting performance through reach advantage, which is strongly correlated with height (wingspan ≈ height in most adults).
While height alone does not determine outcomes, a longer reach allows a fighter to strike or control distance more effictively.
A large-scale analysis by Bruins Sports Analytics examined over 6,000 UFC bouts (1993 to 2021) and found a link between reach advantage and win probability:
| Reach Advantage | Win Rate | Gain from Previous |
|---|---|---|
| +1 inch | 52.28% | +2.28% |
| +2 inches | 53.79% | +1.51% |
| +3 inches | 54.88% | +1.09% |
| +4 inches | 55.27% | +0.39% |
| +5 inches | 56.13% | +0.86% |
| +6 inches | 58.82% | +2.69% |
| +7 inches | 62.39% | +3.57% |
Across the analyzed ranges, the average gain per inch of reach advantage is about +1.68%.
To simulate real-world matchups, the calculator models height as a mixture of three age-specific distributions rather than a single adult-only curve.
Each age group contributes to the global height profile in proportion to its share of the world population:
| Age Group | Population Share | Mean Height | Standard Deviation |
|---|---|---|---|
| Children (0-12) | ~25% | 45 in (3'9") | ≈ 4 in |
| Adolescents (13-17) | ~10% | 60 in (5'0") | ≈ 3 in |
| Adults (18+) | ~65% | 66 in (5'6") | ≈ 3.2 in |
When combined, this mixture produces an overall world height distribution with a mean near
62 inches (5′2″) and a realistic spread that naturally accounts for children, teenagers,
and adults without over-representing either extreme.
The resulting behavior of the model is intuitive:
- Shorter individuals (below 5'2") experience a modest disadvantage.
- Average-height individuals (~5'2") are neutral.
- Taller individuals gain a measurable but saturating advantage - extreme heights (6'+) plateau since they already exceed nearly everyone globally.
Overall, the final height multiplier yields about a ±15% range in effect, matching the real-world influence of reach observed in professional fight data while remaining fair across the world's height diversity.
https://www.kaggle.com/datasets/rajeevw/ufcdata
https://www.bruinsportsanalytics.com/post/mma_reach
https://ourworldindata.org/human-height
https://distributionofthings.com/human-height/#fn-4
https://www.gigacalculator.com/calculators/height-percentile-calculator.php
https://www.worlddata.info/average-bodyheight.php#by-population
Weight influences fighting performance through strength, impact force, and physical durability.
Greater mass generally means more momentum and striking power, as well as improved ability to absorb blows or control clinches.
However, excessive or insufficient body mass can reduce agility, endurance, and overall effectiveness.
Just like the height calculation, the calculator models weight with a mixture of the same three age-specific distributions:
| Age Group | Population Share | Mean Weight (lbs) | Standard Deviation |
|---|---|---|---|
| Children (0-12) | ~25% | 47 | 25 |
| Adolescents (13-17) | ~10% | 117 | 25 |
| Adults (18+) | ~65% | 137 | 35 |
Around the global average (120 lbs), each 10 lbs of mass will alter the expected fighting performance multiplier by 0.8% (equating to a 2-3% change in final percentile).
Finally, a BMI adjustment then refines the result based on body composition:
| Category | Condition | Effect on Multiplier |
|---|---|---|
| Underweight | BMI < 18 | Penalty for low muscle mass and fragility |
| Obese | BMI ≥ 35 | Penalty for reduced agility and endurance |
| Morbidly Obese | BMI ≥ 50 | Severe penalty for near-immobility |
Exception: If the user's activity level is high (one of the top three "Athlete" options), the obesity penalties don't apply until morbidly obese, recognizing that a higher BMI in this case likely reflects muscle mass rather than fat.
https://pubmed.ncbi.nlm.nih.gov/34854804/
https://thesportjournal.org/article/an-analysis-of-weight-and-fighting-styles-as-predictors-of-winning-outcomes-of-elite-mixed-martial-arts-athletes/
https://thesportjournal.org/article/the-correlation-between-weight-divisions-and-methods-used-by-winning-mixed-martial-arts-athletes/
https://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-12-439
https://fpnotebook.com/Endo/Exam/WghtMsrmntInChldrn.htm
https://pmc.ncbi.nlm.nih.gov/articles/PMC3408371/
Activity in this tool represents a person's general physical conditioning - their strength, speed, endurance, coordination, and mobility, not their fighting skill or martial arts technique.
The multiplier was modeled primarily using data collected by the CDC that measured the number of leisure-time periods per week of vigorous physical activity lasting 10 minutes or more.
| Tier | Population Share | Multiplier |
|---|---|---|
| Severely Disabled / Immobile | 2% | 0.28× |
| Significantly Disabled / Impaired | 14% | 0.42× |
| No Activity / Sedentary | 45% | 0.8× |
| <1x per week / Rarely moves | 4% | 0.9× |
| 1–2x per week / Low Activity | 12% | 1.0× |
| 3–4x per week / Regular Activity | 13% | 1.1× |
| 5+ per week / Highly Active | 8.5% | 1.2× |
The CDC data does not account for the upper tiers of athletic performance - individuals who train or compete at levels beyond general fitness.
To represent these groups, an additional 3 tiers were included:
| Tier | Population Share | Multiplier |
|---|---|---|
| Recreational Athlete | 1% | 1.3× |
| College / Elite Athlete | 0.09% | 1.5× |
| Olympic / Pro Athlete | 0.01% | 1.6× |
Each category is given a multiplier that factors in how activity and fitness might affect a person's chances in an unarmed fight. These multipliers were chosen arbitrarily but logically - informed by experience, general athletic research, and practical realism.
https://www.genspark.ai/spark/global-estimate-of-professional-athletes/1ee3cadd-ef2e-4668-8d01-ba0a1dba19a8
https://www.thelancet.com/journals/langlo/article/PIIS2214-109X(24)00150-5/fulltext
https://www.cdc.gov/nchs/data/series/sr_10/sr10_228.pdf
Martial arts training was modeled using the best available global estimates of participation in real, contact-based combat sports.
Surveys and practitioner statistics suggest that roughly 3% of the world's population has ever trained in a legitimate sparring art (ex. Wrestling, Boxing, BJJ, Judo, Muay Thai, etc.).
To represent skill progression, this trained population was distributed across experience tiers measured in approximate years of training.
Each tier was then assigne a relative "combat ability" value. The calculator then simulates matchups against this entire global distribution (97% untrained and 3% trained) to estimate how a user's experience compares to the world average.
| Tier | Trained Pop. Share | Global Pop. Share | Ability |
|---|---|---|---|
| ≤ 6 months | 20% | 0.6% | 1.5× |
| 1 year | 20% | 0.6% | 1.8× |
| 2 years | 20% | 0.6% | 2.05× |
| 3 years | 15% | 0.45% | 2.25× |
| 4 years | 15% | 0.45% | 2.4× |
| 5+ years | 8.5% | 0.26% | 2.6× |
| Pro level | 0.5% | 0.015% | 3.0× |
https://goldbjj.com/blogs/roll/statistics?srsltid=AfmBOoqvVzJpeqrDHIZr317VxOlFUlna2fiIY8nPV7OG-K-uI6FWrd79
https://zipdo.co/martial-arts-statistics/
https://jiujitsuhaus.com/the-rarity-of-the-brazilian-jiu-jitsu-black-belt-numbers-dropouts-and-the-journey-of-a-lifetime/