I would love whoever does these stats on a game-by-game basis to put out a video saying "These are the shots that according to my model should have been goals."
No way would that video consistently match the number of goals metrics say.
I'd like to see that, too.
I'm guessing those models look at the location of the shot attempts and at the movement of the puck that sets them up. I'd bet they don't even try to quantify the quality of the player's shot.
Sabres had plenty of good chances last night, but they didn't shoot as well as they usually do. Lyon made a few nice saves. Also, a lot of "muffins" and "Mitts couldn't finish", etc.
With their normal shooting, they score three easy. In an eighty-two game season, sometimes it just goes that way. The trick is to be consistent enough over the whole year where you don't have "must-win" games just to make the playoffs. Because hockey is too much like roulette to rely on the outcomes from a single game. That's why playoffs are best-of-seven.
Top-4 D (and normal off-season adjustments) should be able to get them there.
