Quote:
Originally Posted by Samonadreau
I would supplement charts and advanced stats with what i actually see in a live game. To strictly base your judgement 100% off of a chart is a little judge book by the cover. We must be watching different hockey games because i see poor goaltending and defensive breakdowns and poor coverage when i watch most of the games.
|
But how do you usefully compare the poor coverage with other teams' coverage? Every team makes a few mistakes a game.
How do you separate the bias of overweighing a "defensive breakdown that ends up in the back of the net" with a "defensive breakdown that the goalie stopped"?
You can't deny you have biases that affect your perception. We all do.
The data does not have biases. The data is not perfect. The charts aren't a complete story.
They especially won't tell you about things like screens, backdoor opportunities, context (breakaways / forechecking turnovers, deflections) but they do tell you a lot about what's happening
without the bias of projecting one poor defensive play onto every goal/chance against.
On the flip side, your biases won't always tell you when the goaltenders have been making an excess of great saves while the defense has been a technical mess. The data though? It will tell you that
last March, goaltending is what got us into the playoffs and Russell-Wideman were a horrible pairing.
If the Flames have 82.9% of the highest quality chances against as the average hockey team, and the best save percentage team in the league has 88.5% of the highest quality chances as the average team, that means the Flames have been doing
something right.
I mean, this is the shot chart of the Senators:
The shot chart tells you what the great save percentages mask.