UFC 39: The Warriors Return: Predictions & Analysis
UFC 39: The Warriors Return lands on Friday, September 27, 2002 in Uncasville, Connecticut, USA with 8 bouts on the card. The card is headlined by a championship fight. 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 |
|---|---|---|---|
| Ricco Rodriguez vs Randy CoutureHeavyweight | Ricco Rodriguez | Confident | 74% |
| Tim Sylvia vs Wesley CorreiraHeavyweight | Wesley Correira | Lean | 56% |
| BJ Penn vs Matt SerraLightweight | BJ Penn | Confident | 72% |
| Caol Uno vs Din ThomasLightweight | Caol Uno | Lean | 62% |
| Gan McGee vs Pedro RizzoHeavyweight | Pedro Rizzo | Toss-up | 54% |
| Phil Baroni vs Dave MenneMiddleweight | Phil Baroni | Toss-up | 54% |
| Matt Lindland vs Ivan SalaverryMiddleweight | Matt Lindland | Lean | 62% |
| Sean Sherk vs Benji RadachWelterweight | Sean Sherk | Confident | 70% |
Fight-by-Fight Breakdown
Ricco Rodriguez vs Randy Couture
The Heavyweight championship matchup features Ricco Rodriguez (5-1) taking on Randy Couture (16-7).
There's a real Elo separation here: Couture at 1248 versus Rodriguez at 1157. That 90-point gap typically reflects a meaningful difference in recent quality of competition and results. Couture has won 3 straight.
Both fighters land in our "Wrestler" archetype — fighters who win by dictating where the fight takes place, grinding out control time and wearing opponents down. 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. Rodriguez throws significantly more leather — a 0.5 sig. strike per minute gap. Rodriguez is far more active with takedowns, averaging 2.5 more per 15 minutes. Rodriguez has tighter striking defense, making opponents miss more often.
**The Pick: Ricco Rodriguez over Randy Couture.** We're leaning Rodriguez here at 74%, a solid but not overwhelming edge.
Tim Sylvia vs Wesley Correira
The Heavyweight matchup features Tim Sylvia (9-3) taking on Wesley Correira (2-1).
Sylvia is rated at 1296 — 251 points above Correira's 1045. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Correira throws significantly more leather — a 0.0 sig. strike per minute gap. Correira is far more active with takedowns, averaging 0.0 more per 15 minutes. Correira has tighter striking defense, making opponents miss more often.
**The Pick: Wesley Correira over Tim Sylvia.** The model gives Correira a slight nod at 56% — this could easily go either way.
BJ Penn vs Matt Serra
The Lightweight matchup features BJ Penn (12-12-2) taking on Matt Serra (7-6). Penn is the bigger frame at 5'9" with a 2-inch reach advantage.
Serra is rated at 1201 — 263 points above Penn's 938. Gaps this large usually mean one fighter has been consistently beating better opponents.
The style clash matters here: Penn is most dangerous on the ground, constantly threatening chokes and joint locks, while Serra is patient on the feet, timing counters and loading up when he sees openings. In our database, knockout artists own a 54% win rate against submission artists, giving Serra the stylistic edge.
A few statistical edges stand out. Penn throws significantly more leather — a 2.3 sig. strike per minute gap. Penn is far more active with takedowns, averaging 0.9 more per 15 minutes. Penn has tighter striking defense, making opponents miss more often.
**The Pick: BJ Penn over Matt Serra.** We're leaning Penn here at 72%, a solid but not overwhelming edge.
Caol Uno vs Din Thomas
The Lightweight matchup features Caol Uno (3-4-2) taking on Din Thomas (5-3).
Thomas carries a modest Elo edge (1012 to 944), the kind of gap that reflects a slightly better run of form rather than a talent chasm.
Both fighters land in our "Wrestler" archetype — fighters who win by dictating where the fight takes place, grinding out control time and wearing opponents down. 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. Thomas throws significantly more leather — a 0.6 sig. strike per minute gap. Uno is far more active with takedowns, averaging 1.2 more per 15 minutes. Uno has tighter striking defense, making opponents miss more often.
**The Pick: Caol Uno over Din Thomas.** The model gives Uno a slight nod at 62% — this could easily go either way.
Gan McGee vs Pedro Rizzo
The Heavyweight matchup features Gan McGee (2-1) taking on Pedro Rizzo (8-5).
Rizzo is rated at 1268 — 168 points above McGee's 1100. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Rizzo throws significantly more leather — a 0.8 sig. strike per minute gap. McGee is far more active with takedowns, averaging 2.9 more per 15 minutes. Rizzo has tighter striking defense, making opponents miss more often.
**The Pick: Pedro Rizzo over Gan McGee.** This is essentially a pick'em. The model nudges toward Rizzo at 54%, but there's almost nothing separating these two.
Phil Baroni vs Dave Menne
The Middleweight matchup features Phil Baroni (3-6) taking on Dave Menne (2-3).
Menne carries a modest Elo edge (814 to 762), the kind of gap that reflects a slightly better run of form rather than a talent chasm.
Stylistically this is Baroni's striker game against Menne's submission artist approach. Baroni brings a versatile approach, while Menne is most dangerous on the ground, constantly threatening chokes and joint locks. Historically these archetypes are dead-even when they collide.
A few statistical edges stand out. Baroni throws significantly more leather — a 0.3 sig. strike per minute gap. Baroni is far more active with takedowns, averaging 0.7 more per 15 minutes. Menne has tighter striking defense, making opponents miss more often.
**The Pick: Phil Baroni over Dave Menne.** This is essentially a pick'em. The model nudges toward Baroni at 54%, but there's almost nothing separating these two.
Matt Lindland vs Ivan Salaverry
The Middleweight matchup features Matt Lindland (8-3) taking on Ivan Salaverry (3-3).
Lindland is rated at 1239 — 365 points above Salaverry's 874. Gaps this large usually mean one fighter has been consistently beating better opponents.
The style clash matters here: Lindland looks to close distance and put the fight on the mat, while Salaverry is comfortable adjusting on the fly, mixing strikes and grappling as openings appear. In our database, wrestlers own a 59% win rate against all-rounders, giving Lindland the stylistic edge.
A few statistical edges stand out. Lindland throws significantly more leather — a 0.1 sig. strike per minute gap. Lindland is far more active with takedowns, averaging 0.8 more per 15 minutes. Salaverry has tighter striking defense, making opponents miss more often.
**The Pick: Matt Lindland over Ivan Salaverry.** The model gives Lindland a slight nod at 62% — this could easily go either way.
Sean Sherk vs Benji Radach
The Welterweight matchup features Sean Sherk (7-4) taking on Benji Radach (1-0).
Sherk is rated at 1371 — 377 points above Radach's 994. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Radach throws significantly more leather — a 1.0 sig. strike per minute gap. Sherk is far more active with takedowns, averaging 6.7 more per 15 minutes. Radach has tighter striking defense, making opponents miss more often.
**The Pick: Sean Sherk over Benji Radach.** We're leaning Sherk here at 70%, a solid but not overwhelming edge.
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.