NSBA4 Season-Record Prop Markets — Pricing
Four season-long “record” props priced off historical NSBA1–3 data (193 games) and Week-1 NSBA4 results. Every market is a season-maximum / extreme-value quantity: we’re pricing the chance the season’s single biggest value clears a line. Season length assumed at 6 rounds total (14 teams, 7 games/round) — ~35 games / 70 team-scores remaining after Week 1.
Summary — recommended single O/U lines
| Market | Banked so far | Rec. O/U line | Over % | Note |
|---|---|---|---|---|
| Largest margin of victory | 118 | 119.5 | ~67% | Live 2:1 market |
| Biggest player scoring swing | — (1 game in) | 37.5 | ~38% | Lumpy at 36 |
| Highest single-team score | 138 | 159.5 | ~52% | Best balance (pick’em) |
Margin and team-score markets also work well as multi-bucket products; the player-swing market is best as buckets due to its discreteness.
1. Largest Margin of Victory banked: 118
The question: will the season’s single largest margin of victory finish over/under 119.5? We already have a 118 (Ferroasumin 138–20 Gid, Week 1), sitting right on the line with ~35 games left.
Historic context
| Season | Max margin (all) | Max (non-forfeit) |
|---|---|---|
| NSBA1 2022 | 150 | 150 |
| NSBA2 2023 | 178 (forfeit) | 94 |
| NSBA3 2025 | 160 | 160 |
Per-game rate of a 120+ blowout: NSBA1 1.3%, NSBA2 0% (real), NSBA3 8%. The model: P(over) = 1 − (1−p)N.
Pricing (N≈35 remaining games)
Over 119.5 ≈ 67% / Under ≈ 33% — a genuinely live line (earlier ~88% figure used a wrong 50-game count). As American odds with vig: roughly Over −200 / Under +175.
Bucket alternative (season-max margin)
| Bucket | Fair prob | Fair dec |
|---|---|---|
| Under 120 (no new blowout) | 34% | 2.97 |
| 120–135 | 20% | 4.92 |
| 135–150 | 16% | 6.07 |
| 150–165 | 12% | 8.71 |
| 165+ (open-ended) | 18% | 5.54 |
Soft spot: the 165+ leg is model-driven (exponential tail). Only the NSBA2 178–0 forfeit ever reached it. Shade toward ~12–14%, keep open-ended, load extra vig, and decide forfeit treatment up front.
2. Biggest Individual Player Scoring Swing banked: 0
The question: the largest game-to-game point range posted by any single player — their highest single-game tossup total minus their lowest (e.g. 32 one game, −4 another → 36).
Historic context
Season-max player range: NSBA1 40, NSBA2 48, NSBA3 44. But those came from players with 7–13 games. NSBA4 gives each player only ~6 games, which compresses the swing. Nothing is banked yet (everyone has 1 game).
Model (bootstrap of real game-logs, sampled to a 6-game season, ~50 players)
| Threshold | P(season-max range ≥ X) |
|---|---|
| ≥ 28 | ~99% |
| ≥ 32 | ~90% |
| ≥ 36 | ~72% |
| ≥ 40 | ~38% |
| ≥ 44 | ~22% |
| ≥ 48 | ~13% |
Median ≈ 36. Points come in multiples of 4, so the distribution spikes hard at 36 — a single O/U is inherently lumpy. Never set a line on a multiple of 4.
Recommended bucket market
| Bucket | Fair prob | Fair dec |
|---|---|---|
| ≤ 32 | 28% | 3.57 |
| 34–38 (the 36 lump) | 34% | 2.94 |
| 38–42 | 16% | 6.25 |
| 42–46 | 9% | 11.1 |
| 46+ (open-ended) | 13% | 7.69 |
Single-line alternative: O/U 37.5 → Over ~38% / Under ~62% (Over +150 / Under −185 with vig). Soft spot: 46+ is least data-anchored; shade toward ~10%.
3. Highest Single-Game Team Score banked: 138
The question: the highest point total posted by any one team in any single game. We’ve banked 138 with ~70 team-scores left to come.
Historic context
Season-max single-team score: NSBA1 170, NSBA2 178 (forfeit), NSBA3 162. Real ceiling ~170 (a game is ~23 tossups, so scores are near-hard-capped).
Pricing — the cleanest market of the four
| Line | Over | Under |
|---|---|---|
| Over 149.5 | ~77% | ~23% |
| Over 159.5 | ~52% | ~48% |
| Over 169.5 | ~25%* | ~75% |
O/U 159.5 is essentially pick’em — price it Over −110 / Under −110.
Bucket alternative
| Bucket | Fair prob | Fair dec |
|---|---|---|
| Under 150 (no big jump) | ~23% | 4.33 |
| 150–160 | ~26% | 3.98 |
| 160–170 | ~26% | 4.70 |
| 170+ (open-ended) | ~22% | 3.28 |
*Caveat on 170+: the raw exponential model says ~30%, but that’s inflated — only two team-scores in history ever reached 170, and the 178 was a forfeit. Real non-forfeit 170+ is ~1 in 386 per team-game, putting the true bucket nearer ~18%; 180+ is essentially forfeit-only (≈0%). Decide forfeit treatment — it’s the biggest swing on the top bucket.
Resolution rules (shared template)
All four markets settle on a season maximum at the end of the NSBA4 regular season (Rounds 1–6), on finalized box scores. Playoffs excluded unless stated. All O/U lines set on half-points so no result can push.
- Largest margin of victory — max of |score_a − score_b| across all games.
- Biggest player scoring swing — for each player, (highest single-game tossup total) − (lowest); settles on the max across players. Tossup points only; negatives count; player needs ≥ 2 counting games.
- Highest single-team score — max of any one team’s final game score.
- Forfeits/walkovers — recommend excluding credited forfeit scores from all four (this is the single most impactful rule; it keeps the 178–0-type result out of the tail).
Key modeling caveats
- Season length is the biggest lever. All probabilities assume ~35 games / 70 team-scores remaining (6 rounds, 14 teams). Confirming the exact bye structure would tighten every number.
- Exponential tails overstate extremes. Used for margin and (rejected for) team-score; team scores and player swings are bounded by game structure (~23 tossups), so empirical tails are safer at the top end.
- Forfeit treatment drives every top bucket. Settle it before booking.
- Data: 193 historic games (NSBA1–3), 5 NSBA4 Week-1 games. Tail data is thin (~6 games above margin 115), so deep-tail legs carry model risk — price extra vig there.