GDT: Game 8: Columbus @ New Jersey, 5 PM EST

[majormajor]

Players are given ratings using an algorithm that weighs, by position, on-ice performance including xGF, xGA, and other stats. Dark Blue is defensive contribution, Royal blue offensive, sky blue individual, and lightest color is individual contribution.

I know there is an algorithm, I want to know what the algorithm is. If I see a guy has a 69% xG I don't expect him to have a bad game score, so I'd like to know the weights for the other factors.

 

[Cowumbus]

I know there is an algorithm, I want to know what the algorithm is. If I see a guy has a 69% xG I don't expect him to have a bad game score, so I'd like to know the weights for the other factors.

People pay money for it. Don’t think that they are giving it out. Although some people who watched the game aren’t surprised by the chart.

 

[Cowumbus]

Cole Sillinger leads the cumulative CBJ xG race with 1.78

 

[Jan]

And you know what individual corsi is, right? It's just shot attempts by that player. Sometimes guys don't shoot the puck. You will also have an individual zero xG if you don't shoot.

I'm not here to argue that Jack Roslovic is a great player or that he had a great game, but since the general tenor around here is that he can't win a draw and destroys every line he is on by not passing, well I'm going to take note of the fact that he did the opposite of that in this game. He won 9 of 10 draws and was on the ice for 69% of xG.

By the way - If you could tell me what the data sources are for that chart, we could discuss it, but since it is totally inscrutable, I'm going to ignore it when it runs counter to the publicly available stats. I can't use a graph if I don't know what it means.

I do not think that people do not see possibilities in Roslovic, it is just that he have some obviously flaws. Im my point of view two major flows. I tries to much himself (stick handling) and his passing play.
That destroys the over all picture.

He gets carried by Bjorkstrand.

Yes, but I only think that should be expected.
I think that if Roslovic would concentrate on his passing game, he would become an excellent player.

 

[Halberdier]

To use more words to better explain my point:

That he is playing like he is right now has happened before and is why Winnipeg decided it was ok to include him in the PLD trade.

context: after the trade, I was among the first to somewhat-but-only-somewhat sarcastically suggest that Jack was the best player in the deal.

Roslovic was most of the time okay/okayish on the Jets. Obviously it's harder for him to play as center now.

 

[Halberdier]

I know there is an algorithm, I want to know what the algorithm is. If I see a guy has a 69% xG I don't expect him to have a bad game score, so I'd like to know the weights for the other factors.

xGF% a nice stat, but pretty limited when compared to some of the most advanced ones. You are expected to have a quite good game with 69 xGF%, but you might just shoot 20 times on the chest of the goalie and then accidentally pass 3 times the puck for the opponent to score, having like -3 after the game.

Game score -type of stats typically incorporates also the actual results, goals for, goals against, drawn penalties etc.