Today's games 2026-10-17 · 2 games
EDM @ UTA
EDM 41%59% UTA
Vegas lines (DK)
ML: EDM +124 · UTA -148
Puck: EDM +1.5 (-205) · UTA -1.5 (+170)
O/U 6.5 (o -122 / u +102)
Drivers
xG edge away (3%) · goalie edge home
🎯 Model's bets 1h before puck (16:35)
ML UTA -142 C+ · PL EDM +1.5 (-205) B
PIT 🔁 @ VGK
🔁 PIT on B2B — COL→VGK (~608 mi) · 🏔️ altitude leg: COL 5,280 ft
PIT 35%65% VGK
Vegas lines (DK)
ML: PIT +180 · VGK -218
Puck: PIT +1.5 (-130) · VGK -1.5 (+110)
O/U 6.5 (o -122 / u +102)
Goalies
Sergei Murashov (Confirmed) · Adin Hill (Confirmed)
Drivers
xG edge away (3%) · away on B2B
🎯 Model's bets 1h before puck (12:35)
ML VGK -205 B · PL PIT +1.5 (-135) B-
Best bets ranked by model-vs-market gap, with the reasoning
Season P&L: -5.2u (3-8 on 11 graded best bets · 1u flat stakes at logged DK lines)
#1 of 6 — NYI +114 @ TOR C+
Model prices NYI at 59%; DK's +114 implies 45%. That's a 13.8% gap between our number and the market's.
Line hasn't budged since 08:35 (+114) — the gap has sat there all day. Double-digit gap: this is the model's strongest kind of disagreement.
#2 of 6 — VAN +310 @ TBL D-
Model prices VAN at 33%; DK's +310 implies 23%. That's a 9.5% gap between our number and the market's.
Line hasn't budged since 08:35 (+310) — the gap has sat there all day. A solid-sized gap, comfortably over the bar.
#3 of 6 — DET -135 @ SJS B
Model prices DET at 64%; DK's -135 implies 55%. That's a 8.8% gap between our number and the market's.
Line has moved from -135 (08:35) to -148 (18:35) — the market has drifted, widening this gap. A solid-sized gap, comfortably over the bar.
#4 of 6 — PIT 🔁 +170 @ VGK D-
Model prices PIT at 43%; DK's +170 implies 36%. That's a 7.2% gap between our number and the market's.
The xG inputs lean slightly, not enough to decide anything on their own. Goaltending is a wash in the model. PIT is on the second night of a back-to-back (~608 mi COL→VGK, altitude leg (COL 5,280 ft)) — the pick is winning despite the rest spot, not because of it. You're fading home ice, which the model prices modestly — most of this gap is the market overcharging for it. In net: Sergei Murashov (Confirmed), Adin Hill (Confirmed). Line has moved from +195 (08:35) to +170 (12:35) — the market has drifted, widening this gap. A solid-sized gap, comfortably over the bar.
#5 of 6 — SEA +154 @ FLA D-
Model prices SEA at 45%; DK's +154 implies 38%. That's a 6.9% gap between our number and the market's.
Line hasn't budged since 08:35 (+154) — the gap has sat there all day. Just over the 5-point bar — a marginal qualifier, sized accordingly in your head.
#6 of 6 — WPG +154 @ CAR D-
Model prices WPG at 43%; DK's +154 implies 38%. That's a 5.4% gap between our number and the market's.
Line has moved from +154 (08:35) to +170 (14:35) — the market has drifted, widening this gap. Just over the 5-point bar — a marginal qualifier, sized accordingly in your head.
Ranked by the gap between model win probability and DK no-vig implied probability — rank is the size of the disagreement, the grade chip is the raw win probability. Experimental — the board needs ~100 graded plays before these mean anything.
Value board model's confident disagreements with Vegas
💎 NYI@TOR away — model 59% C+ vs DK +114 (45%) · gap 14%
💎 VAN@TBL away — model 33% D- vs DK +310 (23%) · gap 10%
💎 SJS@DET home — model 64% B vs DK -135 (55%) · gap 9%
💎 PIT@VGK away — model 43% D- vs DK +170 (36%) · gap 7% 🔁PIT B2B (~608 mi COL→VGK 🏔️COL 5,280 ft)
💎 SEA@FLA away — model 45% D- vs DK +154 (38%) · gap 7%
💎 CAR@WPG home — model 43% D- vs DK +154 (38%) · gap 5%
Value board record: 3-8 (27% · avg prob 46%)
Closing line value: line moved toward our number in 1/4 plays (avg +0.8pp)
Season P&L: -5.2u on 11 graded plays (3-8, 1u flat stakes at logged DK lines)
Only games where our win prob beats DK's no-vig implied prob by 5+ pts, ranked by gap. Dogs and favorites both qualify — confidence is the size of the disagreement. Experimental until ~100 graded plays.
Line movement DK moneyline trail, snapshotted every ~2h to puck drop
EDM@UTA (+1.2pp to home since 08:35)+114/-135 08:35 → +114/-135 12:35 → +114/-135 14:35 → +120/-142 16:35away/home moneyline at each snapshot
PIT@VGK (-3.0pp to home since 08:35)+195/-238 08:35 → +180/-218 10:35 → +170/-205 12:35away/home moneyline at each snapshot
Season record logged daily, graded on finals
Moneyline: 17-2 (89% · avg prob 57.6%) · +8.9u
Puck line: 12-7 (63%) · +2.1u
1u flat stakes at logged DK odds. Model's side pick per game, logged before puck drop and graded on finals. For calibration, not betting advice.
Puck-line pick logic corrected 2026-10-18: now takes the higher-cover side of the listed ±1.5 line (earlier picks compared the two teams' win-by-2+ rates, which mispriced the +1.5 side).
Recent games
| Date | Game | ML pick | | PL pick | | Close | Score |
|---|
| 10-15 | BUF@TOR | away 58% C+
-102 | W +1.0u | away +1.5 74% A
-245 | W +0.4u | — | 5-4 |
| 10-15 | NYI@OTT | home 52% C-
-162 | W +0.6u | away +1.5 68% B+
-198 | W +0.5u | — | 1-2 |
| 10-15 | SEA@TBL | home 58% C+
-278 | W +0.4u | away +1.5 65% B
-112 | W +0.9u | — | 3-4 |
| 10-15 | VAN@FLA | home 64% B
-298 | W +0.3u | away +1.5 61% B-
-105 | W +1.0u | — | 2-3 |
| 10-15 | NYR@NJD | home 52% C-
-130 | W +0.8u | away +1.5 68% B+
-218 | L -1.0u | — | 1-7 |
| 10-15 | PHI@DET | home 62% B-
-142 | W +0.7u | away +1.5 62% B-
-225 | W +0.4u | — | 1-2 |
| 10-15 | WPG@CHI | away 55% C
-125 | L -1.0u | home +1.5 68% B+
-238 | W +0.4u | — | 3-4 |
| 10-15 | SJS@NSH | home 59% C+
-130 | L -1.0u | away +1.5 64% B
-218 | W +0.5u | — | 3-2 |
| 10-15 | STL@MIN | home 56% C
-142 | W +0.7u | away +1.5 66% B+
-218 | W +0.5u | — | 4-5 |
| 10-15 | UTA@EDM | away 53% C-
-115 | W +0.9u | home +1.5 69% B+
-270 | L -1.0u | — | 4-2 |
| 10-15 | CGY@VGK | home 64% B
-285 | W +0.4u | away +1.5 61% B-
-118 | L -1.0u | — | 2-6 |
| 10-16 | PIT@COL | home 59% C+
-205 | W +0.5u | away +1.5 64% B
-142 | L -1.0u | ML PIT +170 · COL -205
PL +1.5 (-142) · -1.5 (+120) | 2-5 |
| 10-16 | BOS@ANA | away 53% C-
-108 | W +0.9u | home +1.5 69% B+
-285 | L -1.0u | ML BOS -112 · ANA -108
PL -1.5 (+240) · +1.5 (-298) | 4-2 |
| 10-17 | NJD@WSH | home 50% D+
-125 | W +0.8u | home 31% D-
+210 | W +2.1u | ML NJD -102 · WSH -118
PL +1.5 (-270) · -1.5 (+220) | 2-4 |
| 10-17 | CAR@WPG | away 57% C
-185 | W +0.5u | away 34% D-
+142 | W +1.4u | ML CAR -205 · WPG +170
PL -1.5 (+130) · +1.5 (-155) | 3-2 |
NHL Model v1
Expected goals (xG) is the efficiency core — the hockey analogue of EPA. Honest probabilities, walk-forward tested before wiring. Totals excluded after backtesting showed no signal.
P(home win) = sigmoid(0.133 − 4.177·xGdiff + 0.347·goalie − 0.175·B2Bhome + 0.324·B2Baway)
Walk-forward 2022–26 (n=3,854): 56.6% winner accuracy vs 54.3% always-home baseline. Puck-line cover = P(win) × historical 2+ goal margin rate (home 60.9%, away 58.9%). Inputs: current-season 5v5 xG% regressed to prior season, team goalie GSAA shrunk to prior until ~15 games, rest flags. Starting goalies shown from DailyFaceoff; model uses team goalie aggregate until starter-specific GSAA is backtested and approved.
Early-season caveats
October–November inputs are heavily regressed toward last season — expect conservative probabilities and more coin-flip games until ~15 games played. Goalie quality takes longest to stabilize.
Value board
Flags games where model prob beats DK no-vig implied prob by 5+ pts. Tracked as a forward experiment — the NFL backtest showed model-vs-market gaps were coin flips, so this stays experimental until ~100 graded plays. Closing line value (did the market move toward our number?) is the earliest honest signal.