UFC - Road to UFC 4.6: Predictions & Analysis
UFC - Road to UFC 4.6 lands on Friday, August 22, 2025 in Shanghai, Hebei, China with 2 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 |
|---|---|---|---|
| Bruna Brasil vs Shi MingWomen's Strawweight | Shi Ming | Toss-up | 50% |
| Nyamjargal Tumendemberel vs Terrance SaeteurnFlyweight | Nyamjargal Tumendemberel | Lean | 56% |
Fight-by-Fight Breakdown
Bruna Brasil vs Shi Ming
The Women's Strawweight matchup features Bruna Brasil (3-3) taking on Shi Ming (1-0). Brasil is the bigger frame at 5'6" with a 5-inch reach advantage.
Ming carries a modest Elo edge (997 to 959), the kind of gap that reflects a slightly better run of form rather than a talent chasm.
A few statistical edges stand out. Ming throws significantly more leather — a 0.2 sig. strike per minute gap. Ming is far more active with takedowns, averaging 0.1 more per 15 minutes. Ming has tighter striking defense, making opponents miss more often.
The Pick: Shi Ming over Bruna Brasil. This is essentially a pick'em. The model nudges toward Ming at 50%, but there's almost nothing separating these two.
Nyamjargal Tumendemberel vs Terrance Saeteurn
The Flyweight matchup features Nyamjargal Tumendemberel (0-1) taking on Terrance Saeteurn (0-0). Tumendemberel will look to use a 4-inch reach edge to control distance.
Tumendemberel is rated at 1049 — 187 points above Saeteurn's 862. Gaps this large usually mean one fighter has been consistently beating better opponents.
A few statistical edges stand out. Tumendemberel throws significantly more leather — a 1.7 sig. strike per minute gap. Saeteurn is far more active with takedowns, averaging 0.0 more per 15 minutes. Saeteurn has tighter striking defense, making opponents miss more often.
The Pick: Nyamjargal Tumendemberel over Terrance Saeteurn. The model gives Tumendemberel a slight nod at 56% — this could easily go either way.
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.