UFC Fight Night: Nurmagomedov vs. Song
Round 2 · 1:48
The result, the defining moments, and what the verdict means.
Song Yadong blew the roof off Oriental Sports Center to close out the show on Saturday, stopping Umar Nurmagomedov in the second round.
After choking out Deiveson Figueiredo in Macau, this performance could catapult Song into the No. 1 contender position in the bantamweight division.
Nurmagomedov is 20-1 and has never been finished in twenty-one professional fights; the one defeat is a decision to Merab Dvalishvili over five rounds. He wins by making a fight expensive rather than short — 4.03 takedowns per 15 minutes, 79% takedown defence, and 1.86 significant strikes absorbed per minute, under half what Song gives up.
In May, Song took Deiveson Figueiredo's neck in the second round; in January, Nurmagomedov needed all three to beat the same man on the cards. Song is the harder puncher by a distance — nine knockouts to two — and lands 4.34 significant strikes a minute, but he absorbs 3.81 getting there.
Career stats per UFC.com; recent form is each fighter's last five bouts. Record is as of fight night; career stats, recent form and win totals are the latest on file and may include this fight. Bars compare each stat on a fixed scale; the “Edge” marker flags the leading corner where the gap is meaningful. The snapshot plots five career rates — output, striking defense, durability (strikes absorbed, inverted), takedown defense and finish rate — each normalised on its own scale.
Outcome footage, post-fight coverage, and the pre-fight lead-up.
How the analysts' picks held up once the cage door closed.
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov by decision.
“He obviously wins this fight by decision.”
Track record: winner called in 132 of 185 resolved picks. How winner calls are counted
Winner Accuracy 71.4% (132 hit, 53 miss, across 185 resolved calls). Method Accuracy 61.9% (60 hit, 37 miss, across 97 method-scored calls on 185 resolved calls). Round Accuracy 40.7% (11 hit, 16 miss, across 27 round-scored calls on 185 resolved calls). Underdog Accuracy 48.7% (19 hit, 20 miss, across 39 underdog calls on 185 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov to control the bout with wrestling and win a decision.
“I'm ultimately going to go with the Umar to win this by a decision.”
Track record: winner called in 115 of 163 resolved picks. How winner calls are counted
Winner Accuracy 70.6% (115 hit, 48 miss, across 163 resolved calls). Method Accuracy 61.8% (47 hit, 29 miss, across 76 method-scored calls on 163 resolved calls). Round Accuracy 36.5% (19 hit, 33 miss, across 52 round-scored calls on 163 resolved calls). Underdog Accuracy 52.9% (9 hit, 8 miss, across 17 underdog calls on 163 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
FightIQ's public record lists Umar Nurmagomedov as its pick over Song Yadong.
Track record: winner called in 159 of 227 resolved picks. How winner calls are counted
Winner Accuracy 70.0% (159 hit, 68 miss, across 227 resolved calls). Method Accuracy 39.9% (59 hit, 89 miss, across 148 method-scored calls on 227 resolved calls). Round Accuracy: No round picks. Underdog Accuracy 55.0% (11 hit, 9 miss, across 20 underdog calls on 227 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov by decision.
“My official pick is Umar by decision.”
Track record: winner called in 175 of 264 resolved picks. How winner calls are counted
Winner Accuracy 66.3% (175 hit, 89 miss, across 264 resolved calls). Method Accuracy 58.7% (61 hit, 43 miss, across 104 method-scored calls on 264 resolved calls). Round Accuracy 77.8% (7 hit, 2 miss, across 9 round-scored calls on 264 resolved calls). Underdog Accuracy 36.4% (12 hit, 21 miss, across 33 underdog calls on 264 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov by points.
“Umar by points is the most likely outcome. I got Umar by points.”
Track record: winner called in 43 of 65 resolved picks. How winner calls are counted
Winner Accuracy 66.2% (43 hit, 22 miss, across 65 resolved calls). Method Accuracy 58.3% (7 hit, 5 miss, across 12 method-scored calls on 65 resolved calls). Round Accuracy 0.0% (0 hit, 2 miss, across 2 round-scored calls on 65 resolved calls). Underdog Accuracy 63.6% (7 hit, 4 miss, across 11 underdog calls on 65 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov by decision for his striking, wrestling, movement, and conditioning.
“49-46 or 50-45.”
Track record: winner called in 101 of 154 resolved picks. How winner calls are counted
Winner Accuracy 65.6% (101 hit, 53 miss, across 154 resolved calls). Method Accuracy 60.0% (39 hit, 26 miss, across 65 method-scored calls on 154 resolved calls). Round Accuracy 35.7% (5 hit, 9 miss, across 14 round-scored calls on 154 resolved calls). Underdog Accuracy 33.3% (4 hit, 8 miss, across 12 underdog calls on 154 resolved calls).
Picked: Umar Nurmagomedov
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov without committing to a finish or decision.
“I think Umar gets it done.”
Track record: winner called in 112 of 173 resolved picks. How winner calls are counted
Winner Accuracy 64.7% (112 hit, 61 miss, across 173 resolved calls). Method Accuracy 61.0% (61 hit, 39 miss, across 100 method-scored calls on 173 resolved calls). Round Accuracy 38.3% (18 hit, 29 miss, across 47 round-scored calls on 173 resolved calls). Underdog Accuracy 36.4% (8 hit, 14 miss, across 22 underdog calls on 173 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Angelo Bodetti of We Want Picks picks Umar Nurmagomedov by decision.
“I think Umar gets it done and it's likely a decision.”
Track record: winner called in 93 of 145 resolved picks. How winner calls are counted
Winner Accuracy 64.1% (93 hit, 52 miss, across 145 resolved calls). Method Accuracy 40.0% (2 hit, 3 miss, across 5 method-scored calls on 145 resolved calls). Round Accuracy: No qualifying picks. Underdog Accuracy 34.5% (10 hit, 19 miss, across 29 underdog calls on 145 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Livid Larry MMA picks Umar Nurmagomedov by decision.
Track record: winner called in 34 of 54 resolved picks. How winner calls are counted
Winner Accuracy 63.0% (34 hit, 20 miss, across 54 resolved calls). Method Accuracy 79.2% (19 hit, 5 miss, across 24 method-scored calls on 54 resolved calls). Round Accuracy 75.0% (3 hit, 1 miss, across 4 round-scored calls on 54 resolved calls). Underdog Accuracy 38.5% (5 hit, 8 miss, across 13 underdog calls on 54 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov by decision.
“Umar Nurmagomedov by decision.”
Track record: winner called in 92 of 148 resolved picks. How winner calls are counted
Winner Accuracy 62.2% (92 hit, 56 miss, across 148 resolved calls). Method Accuracy 60.8% (45 hit, 29 miss, across 74 method-scored calls on 148 resolved calls). Round Accuracy 48.8% (20 hit, 21 miss, across 41 round-scored calls on 148 resolved calls). Underdog Accuracy 26.9% (7 hit, 19 miss, across 26 underdog calls on 148 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Lucrative James picks Umar Nurmagomedov by decision.
Track record: winner called in 48 of 79 resolved picks. How winner calls are counted
Winner Accuracy 60.8% (48 hit, 31 miss, across 79 resolved calls). Method Accuracy 54.2% (26 hit, 22 miss, across 48 method-scored calls on 79 resolved calls). Round Accuracy 50.0% (3 hit, 3 miss, across 6 round-scored calls on 79 resolved calls). Underdog Accuracy 41.2% (7 hit, 10 miss, across 17 underdog calls on 79 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov to dominate and win a decision.
“I think that Umar has him beat in all levels and is going to win via decision.”
Track record: winner called in 41 of 69 resolved picks. How winner calls are counted
Winner Accuracy 59.4% (41 hit, 28 miss, across 69 resolved calls). Method Accuracy 57.7% (15 hit, 11 miss, across 26 method-scored calls on 69 resolved calls). Round Accuracy 12.5% (1 hit, 7 miss, across 8 round-scored calls on 69 resolved calls). Underdog Accuracy 30.8% (4 hit, 9 miss, across 13 underdog calls on 69 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
Picks Umar Nurmagomedov to win by decision.
“I'll take Umar Nagamemed off to win a decision here.”
Track record: winner called in 58 of 109 resolved picks. How winner calls are counted
Winner Accuracy 53.2% (58 hit, 51 miss, across 109 resolved calls). Method Accuracy 71.4% (10 hit, 4 miss, across 14 method-scored calls on 109 resolved calls). Round Accuracy 66.7% (2 hit, 1 miss, across 3 round-scored calls on 109 resolved calls). Underdog Accuracy 21.4% (6 hit, 22 miss, across 28 underdog calls on 109 resolved calls).
Picked: Umar Nurmagomedov · Dec
Fight result: Song Yadong · KO/TKO
MMA Mania's UFC Shanghai preview has Jesse Holland picking Umar Nurmagomedov by decision.
Track record: winner called in 27 of 55 resolved picks. How winner calls are counted
Winner Accuracy 49.1% (27 hit, 28 miss, across 55 resolved calls). Method Accuracy 55.6% (15 hit, 12 miss, across 27 method-scored calls on 55 resolved calls). Round Accuracy: No round picks. Underdog Accuracy 27.8% (5 hit, 13 miss, across 18 underdog calls on 55 resolved calls).
Picked: Song Yadong
Fight result: Song Yadong · KO/TKO
Picks Song Yadong as the underdog, pointing to his power, sprawl, conditioning, and opportunities during Umar Nurmagomedov's entries.
“I'm gonna go with Song. I think he chins Umar.”
Track record: winner called in 213 of 322 resolved picks. How winner calls are counted
Winner Accuracy 66.1% (213 hit, 109 miss, across 322 resolved calls). Method Accuracy 53.4% (62 hit, 54 miss, across 116 method-scored calls on 322 resolved calls). Round Accuracy 35.9% (14 hit, 25 miss, across 39 round-scored calls on 322 resolved calls). Underdog Accuracy 47.4% (18 hit, 20 miss, across 38 underdog calls on 322 resolved calls).
Picked: Song Yadong
Fight result: Song Yadong · KO/TKO
FIGHT//AI’s replay record preserves its original live winner selection: Song Yadong.
Track record: winner called in 49 of 153 resolved picks. How winner calls are counted
Winner Accuracy 32.0% (49 hit, 104 miss, across 153 resolved calls). Method Accuracy 50.0% (1 hit, 1 miss, across 2 method-scored calls on 153 resolved calls). Round Accuracy: No round picks. Underdog Accuracy 20.0% (20 hit, 80 miss, across 100 underdog calls on 153 resolved calls).
Analyst picks: Bedtime MMA Miss; TheWeasle Miss; FightIQ Miss; AJ Devito Miss; KennyOkoyeMMA Miss; The MMA Guru Hit; JakeGerberMMA Miss; DFS By The Numbers Miss; We Want Picks Miss; Livid Larry MMA Miss; HopperMMA Miss; James Blissett Miss; MMA Prediction Guru (MMAPG) Miss; Half The Battle Miss; Jesse Holland Miss; FIGHT//AI Hit.
Only 2 of 16 called the upset.
Chart: selected analyst picks, not a win probability. Pill badge: Winner hit rate (total scored picks before this event).
| Expand row |
|---|
The experts above are a lens, not credentials we conferred — analyst, fighter or betting desk.
Each agent’s latest pre-fight call vs result; forecast dates may differ. Market: sportsbook snapshot or historic market reconstruction.
Actual Song · KO/TKO · R2
Closest probability Gemini put 30% on Song.
Each agent’s latest pre-fight forecast. Dates are shown per agent.
63 of 108 winner picks correct · 58% hit rate
Pick Nurmagomedov
Most likely: Nurmagomedov by decision, about 61 in 100
Win chance: Nurmagomedov 74%, Song 26%.
Confidence 7 of 10 · High
61 of 101 winner picks correct · 60% hit rate
Pick Nurmagomedov
Most likely: Nurmagomedov by decision, about 50 in 100
Win chance: Nurmagomedov 78%, Song 22%.
Confidence 8 of 10 · High
68 of 108 winner picks correct · 63% hit rate
Pick Nurmagomedov
Most likely: Nurmagomedov by decision, about 50 in 100
Win chance: Nurmagomedov 70%, Song 30%.
Confidence 7 of 10 · High
3 of 3 models pick Nurmagomedov.
ChatGPT
Nurmagomedov: 74% (about 74 in 100 fights). Song: 26%.
Claude
Nurmagomedov: 78% (about 78 in 100 fights). Song: 22%.
Gemini
Nurmagomedov: 70% (about 70 in 100 fights). Song: 30%.
ChatGPT
Nurmagomedov by decision: 61% (about 61 in 100).
Claude
Nurmagomedov by decision: 50% (about 50 in 100).
Gemini
Nurmagomedov by decision: 50% (about 50 in 100).
ChatGPT
1. Nurmagomedov by decision (61%); 2. Song by KO/TKO (12%); 2. Song by decision (12%)
Claude
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (20%); 3. Song by KO/TKO (14%)
Gemini
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (15%); 2. Song by KO/TKO (15%)
ChatGPT
7 of 10, based on Nurmagomedov's 74% win chance.
Claude
8 of 10, based on Nurmagomedov's 78% win chance.
Gemini
7 of 10, based on Nurmagomedov's 70% win chance.
ChatGPT
Finish (KO/TKO or submission): 27%. Decision: 73%.
Claude
Finish (KO/TKO or submission): 45%. Decision: 55%.
Gemini
Finish (KO/TKO or submission): 38%. Decision: 62%.
ChatGPT
Nurmagomedov wins 74%: KO/TKO 4%, submission 9%, decision 61%.
Claude
Nurmagomedov wins 78%: KO/TKO 8%, submission 20%, decision 50%.
Gemini
Nurmagomedov wins 70%: KO/TKO 5%, submission 15%, decision 50%.
ChatGPT
Song wins 26%: KO/TKO 12%, submission 2%, decision 12%.
Claude
Song wins 22%: KO/TKO 14%, submission 3%, decision 5%.
Gemini
Song wins 30%: KO/TKO 15%, submission 3%, decision 12%.
All models
The models agree: all 3 pick Nurmagomedov.
ChatGPT
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (61%).
Claude
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's number 3 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
Gemini
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
ChatGPT
KO/TKO 16%, submission 11%, decision 73%.
Claude
KO/TKO 22%, submission 23%, decision 55%.
Gemini
KO/TKO 20%, submission 18%, decision 62%.
ChatGPT
Song: 14%, against 13% for Nurmagomedov.
Claude
Nurmagomedov: 28%, against 17% for Song.
Gemini
Nurmagomedov: 20%, against 18% for Song.
ChatGPT
Nurmagomedov: 13% (KO/TKO 4%, submission 9%). Song: 14% (KO/TKO 12%, submission 2%).
Claude
Nurmagomedov: 28% (KO/TKO 8%, submission 20%). Song: 17% (KO/TKO 14%, submission 3%).
Gemini
Nurmagomedov: 20% (KO/TKO 5%, submission 15%). Song: 18% (KO/TKO 15%, submission 3%).
ChatGPT
Nurmagomedov is knocked out in 12% and submitted in 2%. Song is knocked out in 4% and submitted in 9%.
Claude
Nurmagomedov is knocked out in 14% and submitted in 3%. Song is knocked out in 8% and submitted in 20%.
Gemini
Nurmagomedov is knocked out in 15% and submitted in 3%. Song is knocked out in 5% and submitted in 15%.
ChatGPT
Lopsided: 74% vs 26%.
Claude
Lopsided: 78% vs 22%.
Gemini
Lopsided: 70% vs 30%.
ChatGPT
Song wins 26% (about 1 in 4 fights): a big upset.
Claude
Song wins 22% (about 1 in 5 fights): a big upset.
Gemini
Song wins 30% (about 1 in 3 fights): a big upset.
ChatGPT
73%.
Claude
55%.
Gemini
62%.
ChatGPT
If it goes to the judges: Nurmagomedov 84%, Song 16%.
Claude
If it goes to the judges: Nurmagomedov 91%, Song 9%.
Gemini
If it goes to the judges: Nurmagomedov 81%, Song 19%.
ChatGPT
Biggest difference: Nurmagomedov by decision, 61% here against 50% on average for the others.
Gemini
Biggest difference: Nurmagomedov by decision, 50% here against 56% on average for the others.
ChatGPT
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (74%).
Claude
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (78%).
Gemini
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (70%).
All models
On Nurmagomedov: Claude (78%). On Song: Gemini (30%).
All models
Half The Battle (Nurmagomedov by decision), TheWeasle (Nurmagomedov by decision), AJ Devito (Nurmagomedov by decision), MMA Prediction Guru (MMAPG) (Nurmagomedov by decision), Bedtime MMA (Nurmagomedov by decision), HopperMMA (Nurmagomedov by decision), KennyOkoyeMMA (Nurmagomedov by decision), We Want Picks (Nurmagomedov by decision), JakeGerberMMA (Nurmagomedov by decision), Jesse Holland (Nurmagomedov by decision), Livid Larry MMA (Nurmagomedov by decision), James Blissett (Nurmagomedov by decision) and FightIQ (Nurmagomedov by decision): the same winner and method as this model, tied; this model favours Nurmagomedov, most likely by decision.
ChatGPT
On 23 Aug 2026, by gpt-5.6-sol.
Claude
On 23 Aug 2026, by claude-opus-5.
Gemini
On 24 Aug 2026, by gemini-3.1-pro-high.
FightCard
Models research public information before the fight, such as records, styles, recent results and coverage. They cannot see inside training camps, injuries nobody has reported, or what happens on the night. A small chance is not no chance: outcomes the model rates at a few percent still happen, and big favourites lose sometimes.
FightCard
Confidence restates the favourite's win chance on a 10-step scale: the win chance in tens, rounded, so 55% is 6 of 10 and 95% or more is 10 of 10. 1 to 3 reads Low, 4 to 6 Medium and 7 to 10 High. A favourite usually has at least half the chance, so most picks show 5 or more. When the model has the fighters exactly even, there is no pick and no confidence level. It is not a separate judgement by the model.
FightCard
KO/TKO includes referee, doctor and corner stoppages, and disqualifications. A decision includes a fight stopped early and scored on the judges' cards. A stoppage between rounds counts as a finish in the round that just ended. A draw and a no contest share one outcome.
All models
Yes.
Pick
Confidence 7 of 10 · High
Most likely: Nurmagomedov by decision, about 61 in 100
Win chance: Nurmagomedov 74%, Song 26%.
AI model's own reasoning — not checked by FightCard
Umar Nurmagomedov is favored to win a five-round decision by repeatedly mixing distance striking with takedowns and control. Song Yadong’s clearest route is damaging Umar during extended boxing exchanges, but the matchup projects to give him too few sustained opportunities.
Whether Song can consistently break clinches and return to open-space boxing before Nurmagomedov banks prolonged control time.
ChatGPT
Nurmagomedov: 74% (about 74 in 100 fights). Song: 26%.
Claude
Nurmagomedov: 78% (about 78 in 100 fights). Song: 22%.
Gemini
Nurmagomedov: 70% (about 70 in 100 fights). Song: 30%.
ChatGPT
Nurmagomedov by decision: 61% (about 61 in 100).
Claude
Nurmagomedov by decision: 50% (about 50 in 100).
Gemini
Nurmagomedov by decision: 50% (about 50 in 100).
ChatGPT
1. Nurmagomedov by decision (61%); 2. Song by KO/TKO (12%); 2. Song by decision (12%)
Claude
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (20%); 3. Song by KO/TKO (14%)
Gemini
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (15%); 2. Song by KO/TKO (15%)
ChatGPT
7 of 10, based on Nurmagomedov's 74% win chance.
Claude
8 of 10, based on Nurmagomedov's 78% win chance.
Gemini
7 of 10, based on Nurmagomedov's 70% win chance.
ChatGPT
Finish (KO/TKO or submission): 27%. Decision: 73%.
Claude
Finish (KO/TKO or submission): 45%. Decision: 55%.
Gemini
Finish (KO/TKO or submission): 38%. Decision: 62%.
ChatGPT
Nurmagomedov wins 74%: KO/TKO 4%, submission 9%, decision 61%.
Claude
Nurmagomedov wins 78%: KO/TKO 8%, submission 20%, decision 50%.
Gemini
Nurmagomedov wins 70%: KO/TKO 5%, submission 15%, decision 50%.
ChatGPT
Song wins 26%: KO/TKO 12%, submission 2%, decision 12%.
Claude
Song wins 22%: KO/TKO 14%, submission 3%, decision 5%.
Gemini
Song wins 30%: KO/TKO 15%, submission 3%, decision 12%.
ChatGPT
The models agree: all 3 pick Nurmagomedov.
ChatGPT
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (61%).
Claude
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's number 3 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
Gemini
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
ChatGPT
KO/TKO 16%, submission 11%, decision 73%.
Claude
KO/TKO 22%, submission 23%, decision 55%.
Gemini
KO/TKO 20%, submission 18%, decision 62%.
ChatGPT
Song: 14%, against 13% for Nurmagomedov.
Claude
Nurmagomedov: 28%, against 17% for Song.
Gemini
Nurmagomedov: 20%, against 18% for Song.
ChatGPT
Nurmagomedov: 13% (KO/TKO 4%, submission 9%). Song: 14% (KO/TKO 12%, submission 2%).
Claude
Nurmagomedov: 28% (KO/TKO 8%, submission 20%). Song: 17% (KO/TKO 14%, submission 3%).
Gemini
Nurmagomedov: 20% (KO/TKO 5%, submission 15%). Song: 18% (KO/TKO 15%, submission 3%).
ChatGPT
Nurmagomedov is knocked out in 12% and submitted in 2%. Song is knocked out in 4% and submitted in 9%.
Claude
Nurmagomedov is knocked out in 14% and submitted in 3%. Song is knocked out in 8% and submitted in 20%.
Gemini
Nurmagomedov is knocked out in 15% and submitted in 3%. Song is knocked out in 5% and submitted in 15%.
ChatGPT
Lopsided: 74% vs 26%.
Claude
Lopsided: 78% vs 22%.
Gemini
Lopsided: 70% vs 30%.
ChatGPT
Song wins 26% (about 1 in 4 fights): a big upset.
Claude
Song wins 22% (about 1 in 5 fights): a big upset.
Gemini
Song wins 30% (about 1 in 3 fights): a big upset.
ChatGPT
73%.
Claude
55%.
Gemini
62%.
ChatGPT
If it goes to the judges: Nurmagomedov 84%, Song 16%.
Claude
If it goes to the judges: Nurmagomedov 91%, Song 9%.
Gemini
If it goes to the judges: Nurmagomedov 81%, Song 19%.
ChatGPT
Biggest difference: Nurmagomedov by decision, 61% here against 50% on average for the others.
Gemini
Biggest difference: Nurmagomedov by decision, 50% here against 56% on average for the others.
ChatGPT
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (74%).
Claude
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (78%).
Gemini
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (70%).
ChatGPT
On Nurmagomedov: Claude (78%). On Song: Gemini (30%).
ChatGPT
Half The Battle (Nurmagomedov by decision), TheWeasle (Nurmagomedov by decision), AJ Devito (Nurmagomedov by decision), MMA Prediction Guru (MMAPG) (Nurmagomedov by decision), Bedtime MMA (Nurmagomedov by decision), HopperMMA (Nurmagomedov by decision), KennyOkoyeMMA (Nurmagomedov by decision), We Want Picks (Nurmagomedov by decision), JakeGerberMMA (Nurmagomedov by decision), Jesse Holland (Nurmagomedov by decision), Livid Larry MMA (Nurmagomedov by decision), James Blissett (Nurmagomedov by decision) and FightIQ (Nurmagomedov by decision): the same winner and method as this model, tied; this model favours Nurmagomedov, most likely by decision.
ChatGPT
On 23 Aug 2026, by gpt-5.6-sol.
Claude
On 23 Aug 2026, by claude-opus-5.
Gemini
On 24 Aug 2026, by gemini-3.1-pro-high.
FightCard
Models research public information before the fight, such as records, styles, recent results and coverage. They cannot see inside training camps, injuries nobody has reported, or what happens on the night. A small chance is not no chance: outcomes the model rates at a few percent still happen, and big favourites lose sometimes.
FightCard
Confidence restates the favourite's win chance on a 10-step scale: the win chance in tens, rounded, so 55% is 6 of 10 and 95% or more is 10 of 10. 1 to 3 reads Low, 4 to 6 Medium and 7 to 10 High. A favourite usually has at least half the chance, so most picks show 5 or more. When the model has the fighters exactly even, there is no pick and no confidence level. It is not a separate judgement by the model.
FightCard
KO/TKO includes referee, doctor and corner stoppages, and disqualifications. A decision includes a fight stopped early and scored on the judges' cards. A stoppage between rounds counts as a finish in the round that just ended. A draw and a no contest share one outcome.
ChatGPT
Yes.
gpt-5.6-sol · 23 Aug63 of 108 winner picks correct · 58% hit rate
Pick
Confidence 8 of 10 · High
Most likely: Nurmagomedov by decision, about 50 in 100
Win chance: Nurmagomedov 78%, Song 22%.
AI model's own reasoning — not checked by FightCard
Umar Nurmagomedov by five-round decision is the modal outcome (~78% overall win probability, in line with the -455/-550 market). His grappling volume (~4 TD/15, 79% TD defense) and low-risk, high-accuracy striking (3.86 landed/min, 1.86 absorbed) map directly onto how Song Yadong has lost judged fights to Petr Yan and Sean O'Malley, with an added wrestling layer Song has never had to answer. Song's live path is early: 0.48 knockdowns/15 and nine career KOs against a chin that hasn't been tested at 135. Expect Umar to bank rounds and hunt a back-take as Song's output dips late.
Umar's 7-month layoff vs. Song's sharp 3-month turnaround off a finish; a slow first round in a hostile Shanghai main event could hand Song the early window where his power matters most.
Claude
Nurmagomedov: 78% (about 78 in 100 fights). Song: 22%.
ChatGPT
Nurmagomedov: 74% (about 74 in 100 fights). Song: 26%.
Gemini
Nurmagomedov: 70% (about 70 in 100 fights). Song: 30%.
Claude
Nurmagomedov by decision: 50% (about 50 in 100).
ChatGPT
Nurmagomedov by decision: 61% (about 61 in 100).
Claude
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (20%); 3. Song by KO/TKO (14%)
ChatGPT
1. Nurmagomedov by decision (61%); 2. Song by KO/TKO (12%); 2. Song by decision (12%)
Gemini
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (15%); 2. Song by KO/TKO (15%)
Claude
8 of 10, based on Nurmagomedov's 78% win chance.
ChatGPT
7 of 10, based on Nurmagomedov's 74% win chance.
Gemini
7 of 10, based on Nurmagomedov's 70% win chance.
Claude
Finish (KO/TKO or submission): 45%. Decision: 55%.
ChatGPT
Finish (KO/TKO or submission): 27%. Decision: 73%.
Gemini
Finish (KO/TKO or submission): 38%. Decision: 62%.
Claude
Nurmagomedov wins 78%: KO/TKO 8%, submission 20%, decision 50%.
ChatGPT
Nurmagomedov wins 74%: KO/TKO 4%, submission 9%, decision 61%.
Gemini
Nurmagomedov wins 70%: KO/TKO 5%, submission 15%, decision 50%.
Claude
Song wins 22%: KO/TKO 14%, submission 3%, decision 5%.
ChatGPT
Song wins 26%: KO/TKO 12%, submission 2%, decision 12%.
Gemini
Song wins 30%: KO/TKO 15%, submission 3%, decision 12%.
Claude
The models agree: all 3 pick Nurmagomedov.
Claude
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's number 3 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
ChatGPT
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (61%).
Gemini
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
Claude
KO/TKO 22%, submission 23%, decision 55%.
ChatGPT
KO/TKO 16%, submission 11%, decision 73%.
Gemini
KO/TKO 20%, submission 18%, decision 62%.
Claude
Nurmagomedov: 28%, against 17% for Song.
ChatGPT
Song: 14%, against 13% for Nurmagomedov.
Gemini
Nurmagomedov: 20%, against 18% for Song.
Claude
Nurmagomedov: 28% (KO/TKO 8%, submission 20%). Song: 17% (KO/TKO 14%, submission 3%).
ChatGPT
Nurmagomedov: 13% (KO/TKO 4%, submission 9%). Song: 14% (KO/TKO 12%, submission 2%).
Gemini
Nurmagomedov: 20% (KO/TKO 5%, submission 15%). Song: 18% (KO/TKO 15%, submission 3%).
Claude
Nurmagomedov is knocked out in 14% and submitted in 3%. Song is knocked out in 8% and submitted in 20%.
ChatGPT
Nurmagomedov is knocked out in 12% and submitted in 2%. Song is knocked out in 4% and submitted in 9%.
Gemini
Nurmagomedov is knocked out in 15% and submitted in 3%. Song is knocked out in 5% and submitted in 15%.
Claude
Lopsided: 78% vs 22%.
ChatGPT
Lopsided: 74% vs 26%.
Gemini
Lopsided: 70% vs 30%.
Claude
Song wins 22% (about 1 in 5 fights): a big upset.
ChatGPT
Song wins 26% (about 1 in 4 fights): a big upset.
Gemini
Song wins 30% (about 1 in 3 fights): a big upset.
Claude
55%.
ChatGPT
73%.
Gemini
62%.
Claude
If it goes to the judges: Nurmagomedov 91%, Song 9%.
ChatGPT
If it goes to the judges: Nurmagomedov 84%, Song 16%.
Gemini
If it goes to the judges: Nurmagomedov 81%, Song 19%.
Claude
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (78%).
ChatGPT
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (74%).
Gemini
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (70%).
Claude
On Nurmagomedov: Claude (78%). On Song: Gemini (30%).
Claude
Half The Battle (Nurmagomedov by decision), TheWeasle (Nurmagomedov by decision), AJ Devito (Nurmagomedov by decision), MMA Prediction Guru (MMAPG) (Nurmagomedov by decision), Bedtime MMA (Nurmagomedov by decision), HopperMMA (Nurmagomedov by decision), KennyOkoyeMMA (Nurmagomedov by decision), We Want Picks (Nurmagomedov by decision), JakeGerberMMA (Nurmagomedov by decision), Jesse Holland (Nurmagomedov by decision), Livid Larry MMA (Nurmagomedov by decision), James Blissett (Nurmagomedov by decision) and FightIQ (Nurmagomedov by decision): the same winner and method as this model, tied; this model favours Nurmagomedov, most likely by decision.
Claude
On 23 Aug 2026, by claude-opus-5.
ChatGPT
On 23 Aug 2026, by gpt-5.6-sol.
Gemini
On 24 Aug 2026, by gemini-3.1-pro-high.
FightCard
Models research public information before the fight, such as records, styles, recent results and coverage. They cannot see inside training camps, injuries nobody has reported, or what happens on the night. A small chance is not no chance: outcomes the model rates at a few percent still happen, and big favourites lose sometimes.
FightCard
Confidence restates the favourite's win chance on a 10-step scale: the win chance in tens, rounded, so 55% is 6 of 10 and 95% or more is 10 of 10. 1 to 3 reads Low, 4 to 6 Medium and 7 to 10 High. A favourite usually has at least half the chance, so most picks show 5 or more. When the model has the fighters exactly even, there is no pick and no confidence level. It is not a separate judgement by the model.
FightCard
KO/TKO includes referee, doctor and corner stoppages, and disqualifications. A decision includes a fight stopped early and scored on the judges' cards. A stoppage between rounds counts as a finish in the round that just ended. A draw and a no contest share one outcome.
Claude
Yes.
claude-opus-5 · 23 Aug61 of 101 winner picks correct · 60% hit rate
Pick
Confidence 7 of 10 · High
Most likely: Nurmagomedov by decision, about 50 in 100
Win chance: Nurmagomedov 70%, Song 30%.
AI model's own reasoning — not checked by FightCard
Umar Nurmagomedov faces Song Yadong in a five-round bantamweight main event. The fight represents a classic clash of styles: Nurmagomedov's elite combat sambo against Yadong's Sanda-based striking. While Yadong has demonstrated solid takedown defense and finishing ability, Nurmagomedov's relentless grappling pressure is expected to dictate where the fight takes place. His ability to chain takedowns and maintain top control will likely neutralize Yadong's striking threats, leading to a dominant decision victory.
Yadong's ability to maintain his knockout power in later rounds while fending off relentless takedown attempts.
Gemini
Nurmagomedov: 70% (about 70 in 100 fights). Song: 30%.
ChatGPT
Nurmagomedov: 74% (about 74 in 100 fights). Song: 26%.
Claude
Nurmagomedov: 78% (about 78 in 100 fights). Song: 22%.
Gemini
Nurmagomedov by decision: 50% (about 50 in 100).
ChatGPT
Nurmagomedov by decision: 61% (about 61 in 100).
Gemini
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (15%); 2. Song by KO/TKO (15%)
ChatGPT
1. Nurmagomedov by decision (61%); 2. Song by KO/TKO (12%); 2. Song by decision (12%)
Claude
1. Nurmagomedov by decision (50%); 2. Nurmagomedov by submission (20%); 3. Song by KO/TKO (14%)
Gemini
7 of 10, based on Nurmagomedov's 70% win chance.
ChatGPT
7 of 10, based on Nurmagomedov's 74% win chance.
Claude
8 of 10, based on Nurmagomedov's 78% win chance.
Gemini
Finish (KO/TKO or submission): 38%. Decision: 62%.
ChatGPT
Finish (KO/TKO or submission): 27%. Decision: 73%.
Claude
Finish (KO/TKO or submission): 45%. Decision: 55%.
Gemini
Nurmagomedov wins 70%: KO/TKO 5%, submission 15%, decision 50%.
ChatGPT
Nurmagomedov wins 74%: KO/TKO 4%, submission 9%, decision 61%.
Claude
Nurmagomedov wins 78%: KO/TKO 8%, submission 20%, decision 50%.
Gemini
Song wins 30%: KO/TKO 15%, submission 3%, decision 12%.
ChatGPT
Song wins 26%: KO/TKO 12%, submission 2%, decision 12%.
Claude
Song wins 22%: KO/TKO 14%, submission 3%, decision 5%.
Gemini
The models agree: all 3 pick Nurmagomedov.
Gemini
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
ChatGPT
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's joint number 2 outcome. Its most likely outcome was Nurmagomedov by decision (61%).
Claude
Result: Song by KO/TKO in round 2. Song by KO/TKO was the model's number 3 outcome. Its most likely outcome was Nurmagomedov by decision (50%).
Gemini
KO/TKO 20%, submission 18%, decision 62%.
ChatGPT
KO/TKO 16%, submission 11%, decision 73%.
Claude
KO/TKO 22%, submission 23%, decision 55%.
Gemini
Nurmagomedov: 20%, against 18% for Song.
ChatGPT
Song: 14%, against 13% for Nurmagomedov.
Claude
Nurmagomedov: 28%, against 17% for Song.
Gemini
Nurmagomedov: 20% (KO/TKO 5%, submission 15%). Song: 18% (KO/TKO 15%, submission 3%).
ChatGPT
Nurmagomedov: 13% (KO/TKO 4%, submission 9%). Song: 14% (KO/TKO 12%, submission 2%).
Claude
Nurmagomedov: 28% (KO/TKO 8%, submission 20%). Song: 17% (KO/TKO 14%, submission 3%).
Gemini
Nurmagomedov is knocked out in 15% and submitted in 3%. Song is knocked out in 5% and submitted in 15%.
ChatGPT
Nurmagomedov is knocked out in 12% and submitted in 2%. Song is knocked out in 4% and submitted in 9%.
Claude
Nurmagomedov is knocked out in 14% and submitted in 3%. Song is knocked out in 8% and submitted in 20%.
Gemini
Lopsided: 70% vs 30%.
ChatGPT
Lopsided: 74% vs 26%.
Claude
Lopsided: 78% vs 22%.
Gemini
Song wins 30% (about 1 in 3 fights): a big upset.
ChatGPT
Song wins 26% (about 1 in 4 fights): a big upset.
Claude
Song wins 22% (about 1 in 5 fights): a big upset.
Gemini
62%.
ChatGPT
73%.
Claude
55%.
Gemini
If it goes to the judges: Nurmagomedov 81%, Song 19%.
ChatGPT
If it goes to the judges: Nurmagomedov 84%, Song 16%.
Claude
If it goes to the judges: Nurmagomedov 91%, Song 9%.
Gemini
Biggest difference: Nurmagomedov by decision, 50% here against 56% on average for the others.
ChatGPT
Biggest difference: Nurmagomedov by decision, 61% here against 50% on average for the others.
Gemini
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (70%).
ChatGPT
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (74%).
Claude
Of 16 analyst picks, 14 back Nurmagomedov and 2 back Song. This model picks Nurmagomedov (78%).
Gemini
On Nurmagomedov: Claude (78%). On Song: Gemini (30%).
Gemini
Half The Battle (Nurmagomedov by decision), TheWeasle (Nurmagomedov by decision), AJ Devito (Nurmagomedov by decision), MMA Prediction Guru (MMAPG) (Nurmagomedov by decision), Bedtime MMA (Nurmagomedov by decision), HopperMMA (Nurmagomedov by decision), KennyOkoyeMMA (Nurmagomedov by decision), We Want Picks (Nurmagomedov by decision), JakeGerberMMA (Nurmagomedov by decision), Jesse Holland (Nurmagomedov by decision), Livid Larry MMA (Nurmagomedov by decision), James Blissett (Nurmagomedov by decision) and FightIQ (Nurmagomedov by decision): the same winner and method as this model, tied; this model favours Nurmagomedov, most likely by decision.
Gemini
On 24 Aug 2026, by gemini-3.1-pro-high.
ChatGPT
On 23 Aug 2026, by gpt-5.6-sol.
Claude
On 23 Aug 2026, by claude-opus-5.
FightCard
Models research public information before the fight, such as records, styles, recent results and coverage. They cannot see inside training camps, injuries nobody has reported, or what happens on the night. A small chance is not no chance: outcomes the model rates at a few percent still happen, and big favourites lose sometimes.
FightCard
Confidence restates the favourite's win chance on a 10-step scale: the win chance in tens, rounded, so 55% is 6 of 10 and 95% or more is 10 of 10. 1 to 3 reads Low, 4 to 6 Medium and 7 to 10 High. A favourite usually has at least half the chance, so most picks show 5 or more. When the model has the fighters exactly even, there is no pick and no confidence level. It is not a separate judgement by the model.
FightCard
KO/TKO includes referee, doctor and corner stoppages, and disqualifications. A decision includes a fight stopped early and scored on the judges' cards. A stoppage between rounds counts as a finish in the round that just ended. A draw and a no contest share one outcome.
Gemini
Yes.
gemini-3.1-pro-high · 24 Aug68 of 108 winner picks correct · 63% hit rate
AI-generated · Not betting advice · 21+. Gambling problem? Call 1-800-MY-RESET
No current consensus line is saved on this page.
Comments are currently closed while we finish moderation and privacy controls.