NBA Over/Under Predictions - February 4, 2026
Today's Picks Summary
| Game | Vegas Total | Pick | Predicted Total | Confidence |
|---|---|---|---|---|
| DEN @ NYK | 226.3 | UNDER 226.3 | 222.3 | HIGH |
| BOS @ HOU | 215.5 | UNDER 215.5 | 212.5 | HIGH |
| NOP @ MIL | 222.6 | UNDER 222.6 | 219.5 | MEDIUM |
| MIN @ TOR | 225.4 | SKIP | - | - |
| OKC @ SAS | 223.1 | SKIP | - | - |
| MEM @ SAC | 231.2 | SKIP | - | - |
| CLE @ LAC | 223.3 | SKIP | - | - |
Best Bets of the Day
DEN @ NYK UNDER 226.3
Confidence: HIGH
Key Factors:
- BOTH teams on B2B — strongest UNDER signal in the system (-6 to -8 pts)
- Denver played @ DET last night (away B2B with travel)
- New York played @ WAS last night (home B2B)
- First dual-B2B game in our tracking history
- Aaron Gordon OUT (~14 PPG) + Cameron Johnson OUT (~16 PPG) — DEN missing key scorers
- NYK is a slow-pace team — pace alignment supports UNDER
- Denver 100% OVER trend (3 games) but B2B overrides trend per guidelines
- Dual B2B is the most reliable UNDER signal we have (untested but theoretically strongest)
- Boston on away B2B — played @ DAL last night, now traveling to Houston
- Away team B2B = -4 to -5 pts per guidelines
- Houston is a slow-pace team → triggers Auto-UNDER rule (away B2B + slow pace opponent)
- Boston is our most reliable team: 0% OVER rate (3 games), 100% prediction accuracy (3/3)
- Houston neutral at 50% O/U rate, 50% prediction accuracy (2 games)
- Tatum OUT all season (already priced into line — no additional adjustment)
- B2B away factor is 100% accurate in our system (1/1)
- Already the lowest total on the board at 215.5 — Vegas acknowledges the defensive matchup
- Primary factor: DUAL B2B — Both Denver and New York played last night. This is the first time in our tracking history both teams are on a B2B. Per guidelines, both teams B2B = -6 to -8 pts combined impact
- Denver traveled from Detroit, adding fatigue on top of the B2B
- New York is a slow-pace team, which supports the UNDER direction
- Denver is a fast-pace team but B2B fatigue dampens pace output
- Denver has gone OVER in all 3 tracked games (100% rate) but the dual B2B is unprecedented and overrides trend data
- NYK is neutral at 50% O/U rate
- DEN: Aaron Gordon OUT (calf strain) — ~14 PPG scorer removed
- DEN: Cameron Johnson OUT (ankle) — ~16 PPG scorer removed
- DEN: Combined ~30 PPG of scoring absent from Denver's rotation
- NYK: Karl-Anthony Towns Questionable (knee) — if he sits, massive further reduction
- No B2B for either team — primary edge finder absent
- Both teams are neutral pace — no pace signal
- Minnesota has 100% OVER rate but only 1 tracked game, and our prediction on that game was wrong (0% accuracy)
- Toronto has zero tracking data — completely unproven in our system
- Per guidelines, never assign HIGH on unproven teams
- Without B2B, pace alignment, or reliable team data, there's no edge here
- Julius Randle status uncertain — adds unpredictability
- Primary factor: BOS away B2B — Boston played in Dallas last night and now travels to Houston. Away B2B = -4 to -5 pts per guidelines
- Auto-UNDER trigger: Away team B2B + slow pace opponent (Houston) = automatic UNDER per guidelines
- Boston is our most reliable team in the system: 0% OVER rate across 3 games and 100% prediction accuracy (3/3 correct)
- Houston is a slow-pace team, which compounds the fatigue UNDER signal
- B2B away factor has been 100% accurate in our system (1/1)
- Tatum has been OUT all season — his absence is fully priced into the line
- Already the lowest total on the board at 215.5, confirming Vegas sees a low-scoring game
- Houston is neutral at 50% O/U rate and 50% prediction accuracy
- BOS: Jayson Tatum OUT (Right Achilles repair) — season-long absence, fully priced in
- BOS: Roster otherwise relatively healthy for a B2B
- HOU: No major injuries reported
- Primary factor: MIL home B2B — Milwaukee played Chicago last night. Home team B2B = -3 to -4 pts per guidelines
- Both teams are neutral pace — no pace signal
- New Orleans has 100% OVER rate but only 1 game, and our prediction was wrong (0% accuracy)
- Milwaukee is neutral at 50% O/U rate
- MIL: Giannis Antetokounmpo OUT (Right Calf strain) — 28+ PPG superstar
- MIL: Gary Harris OUT, Kevin Porter Jr. OUT, Taurean Prince OUT — depleted roster
- NOP: Zion Williamson likely limited, CJ McCollum status to monitor
- OKC is on an away B2B, which is an UNDER signal (-4 to -5 pts)
- However, OKC is a fast-pace team with 100% OVER trend (2 games) and 100% prediction accuracy
- San Antonio has 0% OVER rate (1 game) and 100% prediction accuracy — UNDER lean
- The B2B UNDER signal directly conflicts with OKC's fast pace and strong OVER identity
- This is the same pattern as the UTA @ IND SKIP from Feb 3 — B2B team with fast pace creates irreconcilable conflict
- Per guidelines, conflicting strong signals = SKIP
- Our system shows OVER picks at 75% accuracy vs UNDER at 45% — fading OKC's OVER identity feels wrong even with B2B
- No B2B for either team — primary edge finder absent
- Sacramento is a fast-pace team which provides a slight OVER lean
- Memphis has 100% OVER rate but only 1 game and 0% prediction accuracy
- Sacramento has 0% OVER rate but also only 1 game
- Both teams massively depleted by injuries and trades: - MEM: Ja Morant status uncertain, Desmond Bane status to monitor - SAC: De'Aaron Fox traded, DeMar DeRozan status to monitor
- Sacramento just lost De'Aaron Fox in a trade — roster in flux, replacement player chaos
- Per guidelines, both teams missing significant PPG combined without B2B = too unpredictable (replacement player effect)
- Highest total on the board at 231.2 — Vegas sees a shootout
- Conflicting trends between the two teams with tiny samples
- No B2B for either team
- Cleveland is a slow-pace team, providing a slight UNDER lean
- However, both teams trend OVER: CLE 100% (1 game), LAC 67% (3 games)
- Critical problem: both teams have 0% prediction accuracy — CLE 0/1, LAC 0/3
- Per rules, never assign HIGH on teams with 0% prediction accuracy
- LAC is our worst-predicted team alongside PHX at 0% accuracy across 3 games
- Major trade chaos: James Harden traded to CLE, Darius Garland traded to LAC today — completely new roster constructions
- Both teams integrating new players in their first game together — impossible to model
- Per guidelines, this level of roster upheaval without B2B = unpredictable
- B2B is the anchor for every pick — all 3 picks have B2B as the primary factor, per rule adjustment from Feb 3
- No star-absence-only UNDER picks (learned from PHX@POR 255 and MIL 131)
- NOP @ MIL capped at MEDIUM specifically because NOP has 0% prediction accuracy (1 game) — new rule prevents HIGH on unproven teams
- OKC @ SAS SKIP mirrors UTA @ IND SKIP pattern — B2B conflicting with fast pace/OVER identity
- CLE @ LAC SKIP due to same-day trade creating unmodelable rosters
- 4 SKIPs reflects quality-over-quantity approach — only bet when B2B provides
Predicted Total: 222.3 (4.0 pts under Vegas — at hard limit, justified by unprecedented dual B2B + injuries)
BOS @ HOU UNDER 215.5
Confidence: HIGH
Key Factors:
Predicted Total: 212.5 (3.0 pts under Vegas — justified by B2B + slow pace auto-UNDER trigger + BOS 100% prediction accuracy)
Game-by-Game Analysis
Game 1: Denver Nuggets @ New York Knicks
Vegas Total: 226.3
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | BOTH teams on B2B (DEN @ DET, NYK @ WAS) | Strong UNDER (-6 to -8) |
| Pace | DEN fast (100+ poss) / NYK slow (<97 poss) | Mixed (neutral) |
| Team O/U Trend | DEN: 100% over (3 games), NYK: 50% neutral | Mixed |
| Historical Accuracy | DEN: 67% correct (3), NYK: 100% correct (1) | Decent |
#### Analysis:
#### Injury Update:
The dual B2B is the anchor for this pick. Adding ~30 PPG of Denver injuries on top of the fatigue signal makes this the strongest UNDER case on the slate. Per rule adjustments, HIGH confidence UNDER requires B2B — this has B2B on BOTH sides. Both teams have >0% prediction accuracy, meeting the HIGH confidence requirement.
Predicted Total: 222.3 (4.0 pts under Vegas — at hard limit, justified by dual B2B + 30 PPG Denver injuries)
Pick: UNDER 226.3 | Confidence: HIGH | Factors Used: Dual B2B (-6 to -8 pts), Gordon OUT (-1 to -2 pts), Cameron Johnson OUT (-1 to -2 pts), NYK slow pace, B2B overrides Denver OVER trend
Game 2: Minnesota Timberwolves @ Toronto Raptors
Vegas Total: 225.4
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | Neither team on B2B | Neutral |
| Pace | MIN neutral / TOR neutral | Neutral |
| Team O/U Trend | MIN: 100% over (1 game), TOR: no data | N/A |
| Historical Accuracy | MIN: 0% correct (1), TOR: no data | Poor |
#### Analysis:
Pick: SKIP | Reason: No B2B, neutral pace, no Toronto data, Minnesota 0% prediction accuracy on 1 game. Insufficient factors to justify a pick.
Game 3: Boston Celtics @ Houston Rockets
Vegas Total: 215.5
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | BOS on away B2B (played @ DAL yesterday) | UNDER (-4 to -5 pts) |
| Pace | BOS neutral / HOU slow (<97 poss) | UNDER lean |
| Team O/U Trend | BOS: 0% over (3 games), HOU: 50% neutral | UNDER lean |
| Historical Accuracy | BOS: 100% correct (3), HOU: 50% correct (2) | BOS excellent |
#### Analysis:
#### Injury Update:
B2B + slow pace opponent + Boston's 100% UNDER/accuracy record = HIGH confidence. This hits every criteria: B2B primary factor, pace alignment, team trend alignment, within 3 pts of Vegas, and team has >0% prediction accuracy.
Predicted Total: 212.5 (3.0 pts under Vegas — justified by auto-UNDER trigger + BOS 100% accuracy)
Pick: UNDER 215.5 | Confidence: HIGH | Factors Used: BOS away B2B (-4 to -5 pts), HOU slow pace (auto-UNDER trigger), BOS 0% OVER rate (3 games), BOS 100% prediction accuracy, lowest total on board
Game 4: New Orleans Pelicans @ Milwaukee Bucks
Vegas Total: 222.6
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | MIL on home B2B (played vs CHI yesterday) | UNDER (-3 to -4 pts) |
| Pace | NOP neutral / MIL neutral | Neutral |
| Team O/U Trend | NOP: 100% over (1 game), MIL: 50% neutral | Mixed |
| Historical Accuracy | NOP: 0% correct (1), MIL: 50% correct (2) | Mixed |
#### Analysis:
#### Injury Update:
Critical lesson from Feb 3: Giannis OUT without fatigue → MIL scored 131 points (22.6 pts OVER Vegas). The replacement player effect is real. However, that was without B2B fatigue. Today, MIL is on a home B2B AND Giannis is out. Per our learnings, B2B + star absence is the reliable combo (fatigue compounds the star absence rather than triggering replacement player scoring spikes).
3 factors align: MIL home B2B (-3 to -4 pts), Giannis OUT (fatigue compounds absence), Milwaukee roster depletion. Capping at MEDIUM because NOP has 0% prediction accuracy (1 game) — per rules, can't assign HIGH on unproven teams.
Predicted Total: 219.5 (3.1 pts under Vegas)
Pick: UNDER 222.6 | Confidence: MEDIUM | Factors Used: MIL home B2B (-3 to -4 pts), Giannis OUT (compounded by B2B fatigue), MIL roster depletion, lesson applied: B2B + star absence = reliable UNDER combo
Game 5: Oklahoma City Thunder @ San Antonio Spurs
Vegas Total: 223.1
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | OKC on away B2B (played vs ORL yesterday) | UNDER (-4 to -5 pts) |
| Pace | OKC fast (100+ poss) / SAS neutral | Slight OVER |
| Team O/U Trend | OKC: 100% over (2 games), SAS: 0% under (1 game) | Conflicting |
| Historical Accuracy | OKC: 100% correct (1), SAS: 100% correct (1) | Both right |
#### Analysis:
Pick: SKIP | Reason: Direct conflict between OKC's away B2B (UNDER) and their fast pace + 100% OVER trend (OVER). Same pattern as UTA@IND SKIP on Feb 3. Conflicting strong signals with no clear resolution.
Game 6: Memphis Grizzlies @ Sacramento Kings
Vegas Total: 231.2
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | Neither team on B2B | Neutral |
| Pace | MEM neutral / SAC fast (100+ poss) | Slight OVER |
| Team O/U Trend | MEM: 100% over (1 game), SAC: 0% under (1 game) | Conflicting |
| Historical Accuracy | MEM: 0% correct (1), SAC: 100% correct (1) | Mixed |
#### Analysis:
Pick: SKIP | Reason: No B2B, both rosters in flux from trades and injuries, conflicting team trends on 1 game each, Memphis 0% prediction accuracy. Replacement player chaos without B2B = unpredictable.
Game 7: Cleveland Cavaliers @ Los Angeles Clippers
Vegas Total: 223.3
#### Factor Analysis:
| Factor | Finding | Impact |
|---|---|---|
| B2B Status | Neither team on B2B | Neutral |
| Pace | CLE slow (<97 poss) / LAC neutral | Slight UNDER |
| Team O/U Trend | CLE: 100% over (1 game), LAC: 67% over (3 games) | OVER lean |
| Historical Accuracy | CLE: 0% correct (1), LAC: 0% correct (3) | Both poor |
#### Analysis:
Pick: SKIP | Reason: Both teams 0% prediction accuracy, major trade today (Harden to CLE, Garland to LAC) creates completely new team compositions. No B2B. Impossible to model first-game-together dynamics.
Summary
Total Picks: 3 (0 OVER, 3 UNDER, 4 SKIP)
Best Bets (HIGH Confidence)
1. DEN @ NYK UNDER 226.3 — BOTH teams on B2B (first dual-B2B in our system). Denver missing Gordon + C. Johnson (~30 PPG). NYK slow pace. Strongest UNDER signal available. 2. BOS @ HOU UNDER 215.5 — BOS away B2B + HOU slow pace = Auto-UNDER trigger. Boston 100% prediction accuracy (3/3), 0% OVER rate. Most reliable team in our system.
Additional Picks (MEDIUM Confidence)
3. NOP @ MIL UNDER 222.6 — MIL home B2B + Giannis OUT. Lesson applied: B2B fatigue compounds star absence (unlike Feb 3 where Giannis OUT without B2B → 131 pts). Capped at MEDIUM due to NOP 0% accuracy.
Learnings Applied
Factors Used Today
| Factor | Times Used | Direction | |
|---|---|---|---|
| Dual B2B | 1 | UNDER | |
| B2B Away | 1 | UNDER | |
| B2B Home | 1 | UNDER | |
| Slow Pace Opponent | 2 | UNDER | |
| Star Scorer OUT | 2 | UNDER | |
| Team O/U Trend | 7 | Mixed | |
| Roster Depletion | 2 | UNDER | _Generated: 2026-02-04_ |