UFC 46: Supernatural: Predictions & Analysis

Saturday, January 31, 2004·Las Vegas, Nevada, USA
Published February 27, 2026
Predictions are for entertainment purposes only and do not constitute financial advice. Please gamble responsibly.

UFC 46: Supernatural lands on Saturday, January 31, 2004 in Las Vegas, Nevada, USA with 8 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

MatchupPickConfidenceProb
Vitor Belfort vs Randy CoutureLight HeavyweightRandy CoutureLean62%
Renato Verissimo vs Carlos NewtonWelterweightCarlos NewtonConfident74%
BJ Penn vs Matt HughesWelterweightMatt HughesToss-up53%
Frank Mir vs Wes SimsHeavyweightFrank MirConfident67%
Lee Murray vs Jorge RiveraMiddleweightJorge RiveraConfident65%
Georges St-Pierre vs Karo ParisyanWelterweightKaro ParisyanLean59%
Josh Thomson vs Hermes FrancaLightweightJosh ThomsonConfident67%
Matt Serra vs Jeff CurranLightweightJeff CurranToss-up54%

Fight-by-Fight Breakdown

Vitor Belfort vs Randy Couture

Light Heavyweight
62%
Randy Couture
Belfort
15-9
Elo 1255
Knockout Artist
VS
Couture
16-7
Elo 1248
Wrestler

The Light Heavyweight matchup features Vitor Belfort (15-9) taking on Randy Couture (16-7).

The Elo gap here is razor-thin — Belfort at 1255, Couture at 1248. On paper, this is as close to a coin flip as it gets. Couture has won 3 straight.

The style clash matters here: Belfort is most dangerous on the ground, constantly threatening chokes and joint locks, while Couture looks to close distance and put the fight on the mat. In our database, wrestlers own a 56% win rate against submission artists, giving Couture the stylistic edge.

A few statistical edges stand out. Couture throws significantly more leather — a 1.3 sig. strike per minute gap. Couture is far more active with takedowns, averaging 1.2 more per 15 minutes. Couture has tighter striking defense, making opponents miss more often.

The Pick: Randy Couture over Vitor Belfort. The model gives Couture a slight nod at 62% — this could easily go either way.

74%
Carlos Newton
Verissimo
1-1
Elo 976
VS
Newton
3-3
Elo 1023
Wrestler

The Welterweight matchup features Renato Verissimo (1-1) taking on Carlos Newton (3-3).

Newton carries a modest Elo edge (1023 to 976), the kind of gap that reflects a slightly better run of form rather than a talent chasm.

A few statistical edges stand out. Newton throws significantly more leather — a 0.5 sig. strike per minute gap. Newton is far more active with takedowns, averaging 2.2 more per 15 minutes. Verissimo has tighter striking defense, making opponents miss more often.

The Pick: Carlos Newton over Renato Verissimo. We're leaning Newton here at 74%, a solid but not overwhelming edge.

BJ Penn vs Matt Hughes

Welterweight
53%
Matt Hughes
Penn
12-12-2
Elo 938
All-Rounder
VS
Hughes
18-6
Elo 1273
Submission Artist

The Welterweight matchup features BJ Penn (12-12-2) taking on Matt Hughes (18-6). Hughes will look to use a 3-inch reach edge to control distance.

Hughes is rated at 1273 — 335 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 comfortable adjusting on the fly, mixing strikes and grappling as openings appear, while Hughes looks to close distance and put the fight on the mat. In our database, wrestlers own a 59% win rate against all-rounders, giving Hughes the stylistic edge.

A few statistical edges stand out. Penn throws significantly more leather — a 0.3 sig. strike per minute gap. Hughes is far more active with takedowns, averaging 1.4 more per 15 minutes. Hughes has tighter striking defense, making opponents miss more often.

The Pick: Matt Hughes over BJ Penn. This is essentially a pick'em. The model nudges toward Hughes at 53%, but there's almost nothing separating these two.

Frank Mir vs Wes Sims

Heavyweight
67%
Frank Mir
Mir
16-10
Elo 1252
Submission Artist
VS
Sims
0-2
Elo 823

The Heavyweight matchup features Frank Mir (16-10) taking on Wes Sims (0-2).

Mir is rated at 1252 — 429 points above Sims's 823. Gaps this large usually mean one fighter has been consistently beating better opponents.

A few statistical edges stand out. Mir throws significantly more leather — a 1.3 sig. strike per minute gap. Mir is far more active with takedowns, averaging 3.0 more per 15 minutes. Sims has tighter striking defense, making opponents miss more often.

The Pick: Frank Mir over Wes Sims. We're leaning Mir here at 67%, a solid but not overwhelming edge.

Lee Murray vs Jorge Rivera

Middleweight
65%
Jorge Rivera
Murray
0-0
Elo 1146
VS
Rivera
7-7
Elo 1174
Striker

The Middleweight matchup features Lee Murray (0-0) taking on Jorge Rivera (7-7).

The Elo gap here is razor-thin — Murray at 1146, Rivera at 1174. On paper, this is as close to a coin flip as it gets.

A few statistical edges stand out. Rivera throws significantly more leather — a 5.3 sig. strike per minute gap. Rivera is far more active with takedowns, averaging 0.0 more per 15 minutes. Murray has tighter striking defense, making opponents miss more often.

The Pick: Jorge Rivera over Lee Murray. We're leaning Rivera here at 65%, a solid but not overwhelming edge.

59%
Karo Parisyan
St-Pierre
19-2
Elo 2022
Wrestler
VS
Parisyan
8-3
Elo 1037
Wrestler

The Welterweight matchup features Georges St-Pierre (19-2) taking on Karo Parisyan (8-3). St-Pierre will look to use a 4-inch reach edge to control distance.

St-Pierre is rated at 2022 — 985 points above Parisyan's 1037. Gaps this large usually mean one fighter has been consistently beating better opponents. St-Pierre rides a 12-fight win streak into this one.

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. Parisyan throws significantly more leather — a 3.4 sig. strike per minute gap. Parisyan is far more active with takedowns, averaging 11.6 more per 15 minutes. St-Pierre has tighter striking defense, making opponents miss more often.

The Pick: Karo Parisyan over Georges St-Pierre. The model gives Parisyan a slight nod at 59% — this could easily go either way.

67%
Josh Thomson
Thomson
3-3
Elo 1217
Striker
VS
Franca
6-4
Elo 1107
Wrestler

The Lightweight matchup features Josh Thomson (3-3) taking on Hermes Franca (6-4). There's a 4-inch height gap favoring Thomson.

There's a real Elo separation here: Thomson at 1217 versus Franca at 1107. That 110-point gap typically reflects a meaningful difference in recent quality of competition and results.

Stylistically this is Thomson's striker game against Franca's wrestler approach. Thomson brings a versatile approach, while Franca 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. Thomson throws significantly more leather — a 2.2 sig. strike per minute gap. Thomson is far more active with takedowns, averaging 2.8 more per 15 minutes. Thomson has tighter striking defense, making opponents miss more often.

The Pick: Josh Thomson over Hermes Franca. We're leaning Thomson here at 67%, a solid but not overwhelming edge.

Matt Serra vs Jeff Curran

Lightweight
54%
Jeff Curran
Serra
7-6
Elo 1201
Knockout Artist
VS
Curran
0-2
Elo 792

The Lightweight matchup features Matt Serra (7-6) taking on Jeff Curran (0-2).

Serra is rated at 1201 — 409 points above Curran's 792. Gaps this large usually mean one fighter has been consistently beating better opponents.

A few statistical edges stand out. Serra throws significantly more leather — a 0.9 sig. strike per minute gap. Serra is far more active with takedowns, averaging 1.9 more per 15 minutes. Curran has tighter striking defense, making opponents miss more often.

The Pick: Jeff Curran over Matt Serra. This is essentially a pick'em. The model nudges toward Curran at 54%, but there's almost nothing separating these two.

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.