Quote:
Originally Posted by MrMike
I mean I do believe the stats, but watching those Tag goals, they were mostly perfect one timer rockets . I’ll just continue not understanding advanced stats.
|
These are just models
There was an early goal that squeaked through, but at least 7 of 9 were not really particularly on the goalie
One main thing to remember is that advanced stats absolutely do not factor in shot placement, in any way. Given the same preceding events, a blast in to the goalie’s crest has the same xGF contribution as a shot that goes in off the iron, top corner
Secondly, the number of shots taken in a single game is far smaller than the number of shots on which these models are built, so statistical significance comes in to play. (When they say things like 93 percent confidence 19 times out of 20, games like this can be the other 1 out of 20)
Columbus had some egregious turnovers. Buffalo had time and space on a few, with some wicked shots. They also had really good puck movement on their 5 on 3… Merzlikins looked like he wasn’t particularly square when it went in, but it’s understandable when you see the several puck moves immediately prior
There was a goal over the shoulder from that dead puck area, right around the decision point, where he commits to the RVH and lives with the result. A generally accepted play, and perfectly placed shot
xGF is great because people understand what it is trying to say. But it’s far from perfect