NHL Hockey Predictions: On-Ice Stats

Hockey is a low-scoring sport where chance plays a major role: that’s precisely why in-depth analysis pays off. Pronovision relies on expected goals, goaltender performance (often the number one factor), power-play efficiency, and the schedule. The goal: to identify discrepancies between the odds and the actual probability on the moneyline, the puck line, and the totals.

The goalkeeper: the number one player

In the NHL, a hot goalie can single-handedly steal a game. The algorithm tracks save percentage, above-average saves (GSAx), and—most importantly—the announced starting goaltender, a detail that is sometimes confirmed late. A change in the starting goaltender radically alters the probability—and thus the value—of the bet.

• Percentage of stops and GSAx
• Proven starting goalkeeper
• Rest since the last game

Expected Goals and Power Play

Just like in soccer, xG reveals which team is truly creating danger, beyond the score. In five-on-five play, they predict future dominance. Power plays and penalty kills add another layer: a team with a strong power play facing a disciplined opponent has a measurable advantage in total goals.

Moneyline, puck line, and total

The moneyline predicts the winner, but in games that are often close, the puck line (a 1.5-goal spread) changes the equation: does the favorite have to win by two goals? The total goals, on the other hand, depend mainly on the goalies and the pace of the game. The algorithm compares all three for the same game and points you toward the most profitable bet.

Fatigue and Travel

The NHL features a packed schedule and long trips across the continent. A team playing back-to-back games—especially with its backup goalie—starts at a real disadvantage. These signs of fatigue, which the AI systematically factors in, are often underestimated by the market when initial odds are set.