NHL Best Bets, Predictions & Expert Picks Saturday, November 14, 2026

Analysis

Dallas enters this home matchup with a 10 and 6 record, while Buffalo sits at 8 and 7. That gives the Stars the better record, but the model isn't willing to turn that advantage into a side bet. Dallas is projected to win 55.0% of the time, yet the moneyline carries no model edge. The same problem shows up on the puck line and total. Dallas is modeled at minus 0.28, well short of the posted 1.5 goal requirement, and the projected total of 6.14 doesn't create value against 6.5. That leaves Thomas Harley's shot prop as the clearest way into this game. His projection is 2.26 shots on goal, below the 2.5 line, and the available price matters more than forcing a team wager. BetOnline's no vig number gives the under a fair chance of 61.0%, equivalent to fair odds of minus 156. Hard Rock is offering minus 120, which creates an 11.8% expected value edge against that sharper reference price. Dallas is also without Miro Heiskanen and Lian Bichsel. I won't pretend the supplied data tells us exactly how those absences change Harley's role, but they are part of the game context. The bet is simple: the projection sits under the number, and the market price is too generous.

Public Fade

Casual bettors will likely gravitate toward Dallas because the Stars have the better record and are playing at home. But the model sees no value on the side, while the prop price offers a real edge. Don't confuse Dallas being the better team with Harley getting over his shot line.

Analysis

Los Angeles owns a 5 and 10 record, while Calgary is 8 and 10, so neither team enters with a dominant profile. The model still gives the Kings a 61.4% chance to win, but that opinion doesn't translate into a betting edge on the moneyline. Los Angeles is projected at minus 0.64 against a 1.5 goal line, which isn't enough to justify laying the puck line. The total is another pass, with the model at 5.86 against a posted 5.5. Ryan Strome's shot prop is where the market gives us a better opportunity. His projection is 1.65 shots on goal, just above the 1.5 threshold. That projection alone is not a massive margin, but the price makes this play worth taking. BetOnline's no vig line puts the over at a fair 48.4% chance, with fair odds of plus 107. BetMGM is dealing plus 130, creating an 11.2% expected value edge. That is a meaningful difference for a prop that only needs two shots. Calgary is without Connor Zary, which is the only listed injury for this matchup. The supplied data doesn't quantify the lineup impact, so I won't manufacture one. The case is still strong enough without it. The market has Strome priced as if the over is less likely than a coin flip, while the sharper reference price says otherwise. Take the plus money.

Public Fade

The public may see Los Angeles's 5 and 10 record and assume this is a game to avoid entirely. That's fair for the side, since the model gives no edge there, but it misses the prop value. Strome doesn't need to carry the game for two shots to cash, and plus 130 is simply too high against the no vig number.

Analysis

This is the tightest game of the three from the model's perspective. Montréal owns a strong 12 and 3 record, while Colorado is also playing well at 11 and 6. Even with the Canadiens at home, the model gives Montréal only a 53.3% win probability. That explains why the moneyline is a pass. The puck line is even less appealing, with Montréal projected at minus 0.18 against a 1.5 goal requirement. The projected total is 6.35, which also falls short of creating an edge against 6.5. The available angle is Artturi Lehkonen to record over 0.5 points. His projection is 0.81 points, comfortably above the one point needed for the bet to cash. This isn't a huge model edge, so it's not a spot to overstate. Still, the price is on our side. BetOnline's no vig line gives the over a fair chance of 55.6%, translating to fair odds of minus 125. BetMGM is offering minus 120, which creates a 2.0% expected value edge. That is a smaller advantage than the other two props, but it is still the only listed bet with positive value in this matchup. Montréal's record may tempt bettors into laying a side, yet the model refuses to support that move. Colorado's 11 and 6 mark helps explain the caution. Keep this one controlled, but the over is the right play when the price beats the fair line.

Public Fade

The public will likely lean Montréal because the Canadiens are 12 and 3 and have home ice. But the model gives them only a 53.3% win probability, so the team market is priced too tightly. The better approach is taking Lehkonen over 0.5 points at minus 120, where the edge is small but real.