UFC 221: Romero vs. Rockhold: Predictions & Analysis
UFC 221: Romero vs. Rockhold lands on Saturday, February 10, 2018 in Perth, Western Australia, Australia with 12 bouts on the card. Below is our fight-by-fight breakdown, combining Elo ratings, rolling statistical trends, style matchup data, and betting market context into a pick for every bout.
Quick Picks
| Matchup | Pick | Confidence | Prob |
|---|---|---|---|
| Yoel Romero vs Luke RockholdMiddleweight | Yoel Romero | Toss-up | 52% |
| Curtis Blaydes vs Mark HuntHeavyweight | Curtis Blaydes | Strong | 81% |
| Tai Tuivasa vs Cyril AskerHeavyweight | Tai Tuivasa | Strong | 82% |
| Jake Matthews vs Li JingliangWelterweight | Jake Matthews | Lean | 58% |
| Tyson Pedro vs Saparbeg SafarovLight Heavyweight | Tyson Pedro | Strong | 80% |
| Dong Hyun Ma vs Damien BrownLightweight | Dong Hyun Ma | Confident | 68% |
| Israel Adesanya vs Rob WilkinsonMiddleweight | Israel Adesanya | Confident | 70% |
| Alexander Volkanovski vs Jeremy KennedyFeatherweight | Jeremy Kennedy | Lean | 60% |
| Jussier Formiga vs Ben NguyenFlyweight | Jussier Formiga | Lean | 56% |
| Ross Pearson vs Mizuto HirotaLightweight | Ross Pearson | Lean | 61% |
| Jose Quinonez vs Teruto IshiharaBantamweight | Jose Quinonez | Lean | 63% |
| Luke Jumeau vs Daichi AbeWelterweight | Luke Jumeau | Lean | 57% |
Fight-by-Fight Breakdown
Yoel Romero vs Luke Rockhold
The Middleweight matchup features Yoel Romero (9-3) taking on Luke Rockhold (6-4). Rockhold is the bigger frame at 6'3" with a 4-inch reach advantage.
Romero is rated at 1613 — 311 points above Rockhold's 1302. Gaps this large usually mean one fighter has been consistently beating better opponents.
Stylistically this is Romero's striker game against Rockhold's wrestler approach. Romero brings a versatile approach, while Rockhold looks to close distance and put the fight on the mat. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Rockhold throws significantly more leather — a 2.9 sig. strike per minute gap. Romero is far more active with takedowns, averaging 0.4 more per 15 minutes. Romero has tighter striking defense, making opponents miss more often.
The Pick: Yoel Romero over Luke Rockhold. This is essentially a pick'em. The model nudges toward Romero at 52%, but there's almost nothing separating these two. Notably, the betting market has Romero at 48% implied while our model sees 52% — a 3-point disagreement that could signal value.
Curtis Blaydes vs Mark Hunt
The Heavyweight matchup features Curtis Blaydes (13-5) taking on Mark Hunt (8-7-1). Blaydes is the bigger frame at 6'4" with a 8-inch reach advantage.
Blaydes is rated at 1634 — 466 points above Hunt's 1169. Gaps this large usually mean one fighter has been consistently beating better opponents.
Both fighters land in our "Striker" archetype — fighters with a well-rounded skill set. When mirror matchups like this happen, the edge usually goes to whoever can impose their preferred pace and range.
A few statistical edges stand out. Blaydes throws significantly more leather — a 2.6 sig. strike per minute gap. Blaydes is far more active with takedowns, averaging 6.4 more per 15 minutes. Blaydes has tighter striking defense, making opponents miss more often.
The Pick: Curtis Blaydes over Mark Hunt. The model is firm on this one: Blaydes at 81%. Notably, the betting market has Blaydes at 62% implied while our model sees 81% — a 19-point disagreement that could signal value.
Tai Tuivasa vs Cyril Asker
The Heavyweight matchup features Tai Tuivasa (8-8) taking on Cyril Asker (2-2).
Tuivasa is rated at 1107 — 203 points above Asker's 904. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Tuivasa throws significantly more leather — a 1.1 sig. strike per minute gap. Asker is far more active with takedowns, averaging 5.1 more per 15 minutes. Tuivasa has tighter striking defense, making opponents miss more often.
The Pick: Tai Tuivasa over Cyril Asker. The model is firm on this one: Tuivasa at 82%. Notably, the betting market has Tuivasa at 73% implied while our model sees 82% — a 9-point disagreement that could signal value.
Jake Matthews vs Li Jingliang
The Welterweight matchup features Jake Matthews (15-7) taking on Li Jingliang (11-6).
The Elo gap here is razor-thin — Matthews at 1295, Jingliang at 1270. On paper, this is as close to a coin flip as it gets. Matthews rides a 3-fight win streak into this one.
Stylistically this is Matthews's wrestler game against Jingliang's striker approach. Matthews looks to close distance and put the fight on the mat, while Jingliang brings a versatile approach. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Jingliang throws significantly more leather — a 3.2 sig. strike per minute gap. Matthews is far more active with takedowns, averaging 0.6 more per 15 minutes. Jingliang has tighter striking defense, making opponents miss more often.
The Pick: Jake Matthews over Li Jingliang. The model gives Matthews a slight nod at 58% — this could easily go either way. Notably, the betting market has Matthews at 40% implied while our model sees 58% — a 17-point disagreement that could signal value.
Tyson Pedro vs Saparbeg Safarov
The Light Heavyweight matchup features Tyson Pedro (6-4) taking on Saparbeg Safarov (1-2).
There's a real Elo separation here: Pedro at 1015 versus Safarov at 916. That 99-point gap typically reflects a meaningful difference in recent quality of competition and results.
A few statistical edges stand out. Safarov throws significantly more leather — a 2.2 sig. strike per minute gap. Pedro is far more active with takedowns, averaging 1.3 more per 15 minutes. Pedro has tighter striking defense, making opponents miss more often.
The Pick: Tyson Pedro over Saparbeg Safarov. The model is firm on this one: Pedro at 80%.
Dong Hyun Ma vs Damien Brown
The Lightweight matchup features Dong Hyun Ma (3-4) taking on Damien Brown (2-3).
The Elo gap here is razor-thin — Ma at 838, Brown at 854. On paper, this is as close to a coin flip as it gets.
Stylistically this is Ma's striker game against Brown's all-rounder approach. Ma brings a versatile approach, while Brown is comfortable adjusting on the fly, mixing strikes and grappling as openings appear. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Brown throws significantly more leather — a 0.3 sig. strike per minute gap. Ma is far more active with takedowns, averaging 1.4 more per 15 minutes. Ma has tighter striking defense, making opponents miss more often.
The Pick: Dong Hyun Ma over Damien Brown. We're leaning Ma here at 68%, a solid but not overwhelming edge. Notably, the betting market has Ma at 56% implied while our model sees 68% — a 12-point disagreement that could signal value.
Israel Adesanya vs Rob Wilkinson
The Middleweight matchup features Israel Adesanya (13-4) taking on Rob Wilkinson (0-1).
Adesanya is rated at 1559 — 736 points above Wilkinson's 823. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Wilkinson throws significantly more leather — a 2.9 sig. strike per minute gap. Wilkinson is far more active with takedowns, averaging 3.7 more per 15 minutes. Adesanya has tighter striking defense, making opponents miss more often.
The Pick: Israel Adesanya over Rob Wilkinson. We're leaning Adesanya here at 70%, a solid but not overwhelming edge. The market implies 74% for Adesanya, but our model sees only 70%. That 4-point gap favoring Wilkinson is worth watching.
Alexander Volkanovski vs Jeremy Kennedy
The Featherweight matchup features Alexander Volkanovski (14-3) taking on Jeremy Kennedy (3-0). There's a 5-inch height gap favoring Kennedy.
Volkanovski is rated at 1824 — 733 points above Kennedy's 1090. Gaps this large usually mean one fighter has been consistently beating better opponents. Kennedy has won 3 straight.
A few statistical edges stand out. Volkanovski throws significantly more leather — a 3.0 sig. strike per minute gap. Kennedy is far more active with takedowns, averaging 2.1 more per 15 minutes. Volkanovski has tighter striking defense, making opponents miss more often.
The Pick: Jeremy Kennedy over Alexander Volkanovski. The model gives Kennedy a slight nod at 60% — this could easily go either way.
Jussier Formiga vs Ben Nguyen
The Flyweight matchup features Jussier Formiga (9-6) taking on Ben Nguyen (4-2).
There's a real Elo separation here: Formiga at 1149 versus Nguyen at 1036. That 113-point gap typically reflects a meaningful difference in recent quality of competition and results.
Stylistically this is Formiga's wrestler game against Nguyen's knockout artist approach. Formiga looks to close distance and put the fight on the mat, while Nguyen is patient on the feet, timing counters and loading up when he sees openings. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Nguyen throws significantly more leather — a 3.1 sig. strike per minute gap. Formiga is far more active with takedowns, averaging 0.9 more per 15 minutes. Formiga has tighter striking defense, making opponents miss more often.
The Pick: Jussier Formiga over Ben Nguyen. The model gives Formiga a slight nod at 56% — this could easily go either way. Notably, the betting market has Formiga at 50% implied while our model sees 56% — a 6-point disagreement that could signal value.
Ross Pearson vs Mizuto Hirota
The Lightweight matchup features Ross Pearson (12-12) taking on Mizuto Hirota (1-4-1).
Pearson carries a modest Elo edge (849 to 777), the kind of gap that reflects a slightly better run of form rather than a talent chasm.
Both fighters land in our "Striker" archetype — fighters with a well-rounded skill set. When mirror matchups like this happen, the edge usually goes to whoever can impose their preferred pace and range.
A few statistical edges stand out. Hirota throws significantly more leather — a 0.2 sig. strike per minute gap. Hirota is far more active with takedowns, averaging 1.4 more per 15 minutes. Pearson has tighter striking defense, making opponents miss more often.
The Pick: Ross Pearson over Mizuto Hirota. The model gives Pearson a slight nod at 61% — this could easily go either way. The market and our model are aligned — the line looks fair.
Jose Quinonez vs Teruto Ishihara
The Bantamweight matchup features Jose Quinonez (5-3) taking on Teruto Ishihara (3-4-1).
Quinonez carries a modest Elo edge (882 to 817), the kind of gap that reflects a slightly better run of form rather than a talent chasm.
Stylistically this is Quinonez's all-rounder game against Ishihara's striker approach. Quinonez is comfortable adjusting on the fly, mixing strikes and grappling as openings appear, while Ishihara brings a versatile approach. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Quinonez throws significantly more leather — a 0.5 sig. strike per minute gap. Quinonez is far more active with takedowns, averaging 2.6 more per 15 minutes. Quinonez has tighter striking defense, making opponents miss more often.
The Pick: Jose Quinonez over Teruto Ishihara. The model gives Quinonez a slight nod at 63% — this could easily go either way. The market and our model are aligned — the line looks fair.
Luke Jumeau vs Daichi Abe
The Welterweight matchup features Luke Jumeau (2-1) taking on Daichi Abe (1-1).
There's a real Elo separation here: Jumeau at 1005 versus Abe at 923. That 82-point gap typically reflects a meaningful difference in recent quality of competition and results.
A few statistical edges stand out. Abe throws significantly more leather — a 1.2 sig. strike per minute gap. Abe is far more active with takedowns, averaging 1.0 more per 15 minutes. Abe has tighter striking defense, making opponents miss more often.
The Pick: Luke Jumeau over Daichi Abe. The model gives Jumeau a slight nod at 57% — this could easily go either way. Notably, the betting market has Jumeau at 41% implied while our model sees 57% — a 16-point disagreement that could signal value.
Methodology
Predictions are generated by our ensemble model combining LightGBM (65%) and CatBoost (35%), trained on every UFC fight since 1994. The model uses 23 features including Elo ratings, rolling 5-fight statistical averages, style matchup history, physical attributes, and market odds when available.
On our held-out test set (402 fights from January-September 2023), the model achieves 63.4% accuracy with a log-loss of 0.626. High-confidence picks (>75% probability) hit at 82.7%. For full model transparency, visit our Model page.