How Is UFC Takedown Defense Calculated? 5,598 Bouts Studied

Quick answer

UFC takedown defense is the share of an opponent's recorded takedown attempts that fail. AgentMMA analyzed 5,598 UFC bouts. In the 3,815 with unequal totals, the takedown leader won 69.9% across all outcomes and 70.2% of decisions. Fighters who stopped every attempt, when opponents tried at least three, won 73.5% of their appearances. Since 2016, fighters completed 37.8% of takedown attempts league-wide, making average defense 62.2%. Those figures show a strong association, not proof that takedowns alone caused the wins.

Data snapshot: AgentMMA analysis of 5,598 UFC bouts with a winner and round-by-round statistics, from 1994-03-11 through 2026-10-10; computed 2026-10-17.

What AgentMMA's data shows

Win rate by takedown defense in the bout (opponent tried 3+ takedowns)

The gradient is steep. Fighters who stopped every shot won nearly three quarters of these appearances, while those defending 33% or less won 22.0%. The intervals are also separated enough to make the pattern hard to dismiss as sampling noise. It still doesn't prove that defense caused the result: stronger fighters may defend better, create more offense, and choose where the bout happens.

Takedowns by division, 2016-2026

Flyweights attempted takedowns most often, at 4.84 per fighter per 15 minutes. Women's bantamweights had the highest accuracy, 43.2%, and therefore the lowest defense, 56.8%. These are division-level descriptions, not standards for individual fighters. Style, opponent mix, bout length, and small division samples all shape the figures.

How is UFC takedown defense calculated?

The formula uses the attempts and completions recorded for a fighter's opponents:

TD Def = 1 - (opponent takedowns landed / opponent takedown attempts)

Multiply the result by 100 to show a percentage. If opponents land 5 takedowns from 25 attempts, the fighter stopped 20. Their takedown defense is 80%.

This is the same arithmetic behind the TD Def meaning shown on UFC fighter pages. Takedown accuracy looks from the attacker's side: landed takedowns divided by attempts. Defense looks from the opponent's side: failed attempts divided by all attempts faced.

No attempts means no useful rate. A fighter who has never faced a recorded shot hasn't shown 100% defense; the denominator is zero. A database should return a missing value, not reward an empty sample.

Career figures require totals, not an average of bout percentages. Suppose a fighter stops 7 of 10 shots in one bout and 1 of 2 in another. The bout rates are 70% and 50%, but the combined rate is 8 stops from 12 attempts, or 66.7%. A simple average would wrongly give 60%.

What counts as a takedown attempt?

UFC Stats publishes landed and attempted takedowns, but live grappling doesn't arrive in neat units. FightMetric's public glossary described a landed takedown as deliberate grappling that brings an opponent down and establishes an advantageous position for an appreciable time. That final part matters. Touching a knee or hand to the canvas doesn't always create a credited completion.

An attempt is also a scorer's call. A clear double-leg that gets sprawled on can be one attempt. A long fence sequence may contain several shots, mat returns, or changes of grip. The scorer has to decide where one attempt ends and another begins.

This is why takedown counts aren't a direct measure of control time. A fighter can score several mat returns inside one long exchange. Another can land one takedown and hold top position for most of a round. The first fighter has more credited takedowns; the second may have produced more sustained control or offense.

The counts are still useful. Human judgment becomes a larger problem when you inspect one disputed sequence than when you study thousands of bouts.

Why does the denominator matter so much?

A displayed 100% can mean a fighter stopped three attempts or thirty. Those records don't deserve the same confidence. One conceded takedown would reduce the first record to 75%, but the second only to 96.8%.

The UFC Record Book handles part of this problem by requiring at least five UFC fights and 20 opponent attempts for its takedown-defense leaderboard. That removes the smallest samples. Twenty attempts can still come from a narrow set of matchups, so the raw count remains useful.

AgentMMA used a separate rule for its bout-level defense buckets. An appearance qualified only when the opponent tried at least three takedowns. That prevents a single stuffed shot from entering the 100% group, but it doesn't make every appearance equally informative.

Opponent quality matters too. Defending 75% against persistent chain wrestlers may tell you more than defending 90% against reluctant shooters. The public percentage makes no adjustment for opponent quality or fatigue. It also misses whether shots started in open space or against the fence.

Do takedowns and takedown defense score with judges?

Defense doesn't earn points by itself under the Unified Rules. A stuffed shot can preserve a useful position and lead to offense, but the defensive act isn't a scoring event. Judges first assess effective striking and effective grappling.

A takedown also isn't a fixed number of points. The ABC's 2025 clarification defines successful grappling through damage produced by takedowns, reversals, submission attempts, and advantageous positions. A takedown that starts damaging strikes or a submission threat carries more weight than a brief change of position followed by an instant stand-up.

That distinction helps explain why statistics and scorecards can disagree. TD Def describes what happened to recorded attempts. It doesn't grade the damage, submission danger, control, or offense that followed.

The same warning applies to the 70.2% decision figure. Takedown leaders often win decisions, but judges aren't simply counting completions. Fighters who land more takedowns may also be winning the striking, creating threats, or controlling the terms of the round.

Do takedowns win UFC fights?

They are strongly associated with winning. In AgentMMA's sample, one fighter landed more takedowns in 3,815 bouts. Equal counts were excluded. The takedown leader won 69.9% of those bouts and 70.2% of the decisions.

That leaves almost three fights in ten won by the fighter with fewer takedowns. A clean knockdown, sustained striking damage, submission attacks, or more effective work after being taken down can outweigh the raw count. Some completed takedowns create almost no offense.

The defense buckets tell a related story. When an opponent tried at least three times, fighters who stuffed every attempt won 73.5%. Keeping the fight in a preferred phase is valuable. Yet the statistic doesn't show whether the defender then used that position well.

Association isn't causation. Skill differences affect both takedown success and the final result. Score effects may matter because a trailing fighter can shoot more often while the leader focuses on balance and position. The database records the sequence, not each fighter's intent.

What can TD Def miss?

It misses the cost of defending. A fighter may stop six shots but spend four minutes pinned to the fence and enter the next round exhausted. The official rate can look excellent while the tactical outcome is poor.

It also misses recovery. Some fighters concede takedowns but stand within seconds. Others defend the first shot, lose the next exchange, and stay underneath. Get-ups, reversals, submission exposure, control time, and damage complete the picture.

Career averages hide change. A fighter can improve after switching camps or decline after injuries and age. Recent round-level data and film can reveal that movement earlier than one cumulative percentage.

Attempt selection isn't random. Opponents shoot more when they expect success. They may abandon wrestling against a strong defender or avoid taking down a dangerous submission fighter. Each fighter helps shape the sample used to judge them.

How to use this when you read a fight

Start with attempts faced. Treat 80% across 100 attempts as firmer evidence than 80% across ten. Then check when those attempts occurred and who made them.

Match the wrestling style. Open-space doubles, reactive singles, body locks, trips, and fence chains test different skills. A single career TD Def number compresses all of them.

Next, look past the first completion. Ask how quickly the fighter stands, whether they give up the back, and whether repeated defense cuts their pace. Check what the attacker creates after landing: damage, passes, submissions, or only a brief position change.

Finally, compare career data with recent rounds. If the percentage and current film disagree, don't force a clean story. Injury, opponent level, tactical choice, or ordinary sample noise may explain the gap.

For prediction models, use raw attempts and completions rather than the rounded display alone. Opponent-adjusted rates, control time, get-ups, submission exposure, age, and recent pace can add context. TD Def is one feature, not a solved grappling matchup.

How solid is this data?

AgentMMA analyzed 5,598 UFC bouts with a winner and round-by-round statistics from 1994-03-11 through 2026-10-10. The takedown-leader result covers 3,815 bouts where landed totals differed. Equal totals were excluded. Defense buckets required at least three attempts by the opponent, and the division table uses bouts since 2016.

The main pattern is large and ordered: win rate falls across every lower defense bucket, from 73.5% to 22.0%. The 95% confidence intervals are reported in the table. They describe sampling uncertainty, not bias in the underlying records.

Three limits matter. First, career takedown defense on fighter pages is cumulative. One bout with one stopped attempt can display 100%, so always check attempts. Second, takedown counts are UFC Stats' calls. Scrambles, trips, and mat returns require judgment. Third, round-by-round statistics cover about 58-64% of bouts in AgentMMA's database, so the analyzed population isn't every UFC bout.

Missingness may not be random. Older events are less likely to have complete round statistics, which can shift the population toward modern rules, training, and divisions. The division table limits that problem by starting in 2016, but doesn't remove differences in opponent mix or style.

The conclusion would weaken if corrected historical records erased the ordered defense gradient, or if opponent-adjusted analysis moved the takedown-leader rate close to 50%. Neither test is part of this descriptive study. The safe reading is narrower: better bout-level takedown outcomes are strongly associated with winning in the available UFC data.

FAQ

What does TD Def mean in UFC?

TD Def means takedown defense. It is the percentage of an opponent's recorded attempts that didn't become landed takedowns.

How is UFC takedown defense calculated?

Subtract opponent takedowns landed from opponent attempts, then divide by opponent attempts and multiply by 100. If an opponent lands 2 of 10, the defense rate is 80%.

What is a good takedown defense percentage in the UFC?

Context matters, but AgentMMA found a clear gradient. Fighters with 67-99% bout-level defense won 57.5% of qualifying appearances; fighters who stopped every attempt won 73.5%. Both groups faced at least three attempts.

How accurate are UFC takedown statistics?

They are coded observations, not sensor readings. Clear shots are usually straightforward, but scrambles, trips, and mat returns can depend on where the scorer separates attempts.

Do takedowns win UFC fights?

Often, but not automatically. The fighter with more landed takedowns won 69.9% of 3,815 unequal-total bouts and 70.2% of decisions in AgentMMA's analysis. The figures show association, not a judging formula or proof of causation.