Published 2026-06-15 · 9 min read
Significant strikes, control time, takedown defense, force estimates, prediction models — the language of modern MMA analysis can feel like a different sport. We pulled the questions fans ask most often across MMA forums, Reddit and stats sites, and answered each one in plain English with the numbers and definitions analysts and broadcasters actually use.
A significant strike is any distance strike that lands, plus power strikes (not pitter-patter) thrown from the clinch or on the ground. Jabs at range count; short positional strikes in a hold typically don't. UFC Stats and most analytics providers follow this convention.
Accuracy is takedowns landed divided by takedowns attempted by the fighter. Defense is takedowns stuffed divided by takedowns the opponent attempted against them. Strong wrestlers usually push one number above 60% and the other above 70%.
Control time is the cumulative seconds a fighter spends in a dominant position — top in guard, side control, mount, back, or pressing the opponent against the cage. The clock starts when control is established and pauses when the opponent escapes or scrambles to neutral.
SLpM is significant strikes landed per minute. SApM is significant strikes absorbed per minute. A volume striker like Max Holloway lives above 6.0 SLpM; a defensive technician keeps SApM under 2.5.
Providers like UFC Stats, FightMetric and Compustrike use slightly different rules for what counts as a strike, when control time starts, and how clinch strikes are categorized. Live human scorers also disagree on borderline calls. Differences of 5–15% per fight are normal.
Historically, all by humans — trained scorers watching the broadcast feed with a tagging tool. Computer vision systems like HITAI now produce the same metrics in real time from a single camera, removing scorer fatigue and the 24–48h delay before stats are published.
Only when it produces damage, submission threats, or visible round dominance. Judges have increasingly discounted 'lay-and-pray' control with no output. Control time correlates with winning rounds at about r=0.55 — strong, but striking output and damage matter more.
Three reasons. (1) Knockouts and submissions end fights before volume matters. (2) Judges score rounds 10-9 regardless of margin, so blowing out one round doesn't bank credit. (3) Damage per strike isn't captured in raw counts — a single clean head kick outweighs 40 leg kicks on the scorecards.
Strawweight and flyweight fights average 40% higher strike volume than heavyweight. Strength of schedule matters too: a 70% takedown defense against ranked wrestlers means more than 90% against journeymen. Always read stats next to opponent quality.
In order of predictive power: significant strike differential (SLpM minus SApM), takedown defense, and knockdown rate. Combining the three explains roughly 65% of variance in outcomes. Pure striking accuracy and submission attempts add smaller signal.
No. Stats give you a base rate — usually a 55–65% confidence pick — but style matchups, recent injuries, weight cuts and cage rust routinely flip favorites. Treat the model as a prior, not a verdict.
Small sample sizes (most fighters have under 20 pro fights), inconsistent scoring rules between promotions, no public damage metric, and survivorship bias in retired fighters' records. Aggregating across promotions also mixes different rule sets.
UFC Stats is the official source but updates after the fight. ESPN MMA and Tapology mirror it. For live, in-round numbers, broadcaster overlays powered by computer vision (such as HITAI) publish per-strike data to OBS, vMix and NDI feeds in under 200ms.
Vision systems triangulate the limb's velocity at impact, the striker's body mass, and the contact frame's deceleration. HITAI's force estimate is calibrated against sensor-mouthguard and accelerometer data within ±3% across thousands of labeled strikes.
Output drop-off across rounds — typically measured as strike volume in round 3+ as a percentage of round 1. Fighters under 75% are visibly gassing; elite cardio fighters stay above 90% deep into championship rounds.
Any time a fighter applies a recognizable submission hold (RNC, armbar, triangle, guillotine, kimura, etc.) with intent to finish. Position threats without a locked hold — like a body triangle or back control — don't count.
If you train, the most accessible workflow is filming sparring on a phone, running it through a vision-based analyzer for strike counts and target zones, then comparing rounds week-over-week. Gyms using HITAI get this automatically from a single camera mounted above the cage.
Official UFC Stats publishes 1–3 hours after a card ends. Real-time systems publish per-strike data while the round is happening — useful for broadcast overlays, in-corner coaching and live judging support.
Book a 20-minute walkthrough with our team. We'll process a clip from your event, show you the live overlays, and scope the integration.
%20--%3e%3csvg%20version='1.1'%20id='Layer_1'%20xmlns='http://www.w3.org/2000/svg'%20xmlns:xlink='http://www.w3.org/1999/xlink'%20x='0px'%20y='0px'%20viewBox='0%200%20198%2061.5'%20style='enable-background:new%200%200%20198%2061.5;'%20xml:space='preserve'%3e%3cstyle%20type='text/css'%3e%20.st0{fill:%23FFFFFF;}%20%3c/style%3e%3cg%3e%3cg%3e%3cg%3e%3cpath%20class='st0'%20d='M45.4,61.5H33.2l4.5-24.1H20.9l-4.5,24.1H0L10.9,4.1c0.4-2.4,2.5-4.1,5-4.1h12.2l-4.3,22.4h16.6L44.7,0h16.5%20L50.4,57.4C49.9,59.8,47.8,61.5,45.4,61.5z%20M38.1,57.4h7.3c0.5,0,0.9-0.3,1-0.8l10-52.6h-8.2l-4.3,22.4H18.9l4.3-22.4h-7.3%20c-0.5,0-0.9,0.3-1,0.8l-10,52.6H13l4.5-24.1h25L38.1,57.4z'/%3e%3c/g%3e%3cpath%20class='st0'%20d='M82.2,2L71.7,57c-0.3,1.4-1.5,2.5-3,2.5H59l10.4-55c0.3-1.4,1.5-2.5,3-2.5H82.2z'/%3e%3cpath%20class='st0'%20d='M101.9,12.8H86.7l1.6-8.3c0.3-1.4,1.5-2.4,3-2.4h40.2l-1.6,8.3c-0.3,1.4-1.5,2.5-3,2.5h-12.8l-8.9,46.7H93%20L101.9,12.8z'/%3e%3cpath%20class='st0'%20d='M198,2l-10.4,55c-0.3,1.4-1.5,2.5-3,2.5h-9.8l10.4-55c0.3-1.4,1.5-2.5,3-2.5H198z'/%3e%3c/g%3e%3cpath%20class='st0'%20d='M154.1,35.2l1.1,15.5l0.7,8.8h10.4c1.4,0,2.5-1.2,2.3-2.6L162,2h-14.3c-0.9,0-1.7,0.5-2.1,1.3l-27.3,56.1h13.1%20l6-13.4l4.5-10.9l1.6-3.9l3.9-8.5c1.5-3.3,3.2-7.8,4.6-11.2h0.2c0.3,3.5,0.6,8.1,0.8,11.2L154.1,35.2z'/%3e%3c/g%3e%3c/svg%3e)