Rangers at Blue Jackets: too close to call

The model sees this one as a coin flip with a slight lean toward the Blue Jackets. Nothing separates these two teams by much, and the case for each side runs through a different part of the game.

The Blue Jackets arrive with three days of rest, against one for the Rangers. That is the cleanest edge in the model, and it has nothing to do with how either team plays.

Shot attempts are where Columbus's case is strongest. The Blue Jackets have held 49.9% of them at five-on-five this season, and the Rangers sit at 44.9%. Expect the territorial tilt to favor the Blue Jackets.

The Rangers answer on expected goals, the model's measure of chance quality. They lead that split at 52.7%, and the Blue Jackets sit at 50.2%, which is why this stays close.

The Rangers are the hotter side, 6-3-0 over their 9 games so far. The Blue Jackets are the cooler one, at -5 in goal differential over 5 games so far.

Some of that Rangers heat may not hold. Their shooting-and-save luck figure, known as PDO, sits at 1.020, running hot, so expect some regression.

In goal, D. Garand has a 91.5% save percentage over 5 games and has saved 0.2 more goals than that workload should have allowed. C. Talbot's save percentage is 82.2% over 2 games, and the workload has produced 2.5 more goals allowed than it should have.

The model gives the Blue Jackets a 53% chance. Read that as a lean, not a verdict. Rest and shot attempts point toward Columbus, while the Rangers' expected goals and their recent results keep it open.

Data: NHL API, Backcheck expected goals model