[Loyal and True]

Originally Posted by saillias

Mr. Corsi linked this article to show how his graph is made. http://www.silversevensens.com/2014/...tions-with-pip

So what I gather... It's a team's Fenwick over their last 25 games.

Most notably, the model ignores special teams.

Winnipeg is worst in the league at giving power plays (251 times shorthanded) and gives up more power play shots (354 shots) and goals (49 goals) than every team but one. Winnipeg need strong Fenwick 5on5 just to stay in the games. This is why they have +3 goal differential.

Meanwhile Calgary gives far and away the fewest power plays (135 times shorthanded) and gives up fewest power play shots (166 shots) and 3rd fewest goals (27 goals). Calgary would have done even better if their PK save percentage was not below league average.

Power play shots against are way more costly than 5on5 shots against. Shooting percentage league wide is around 13% on PP versus 8% 5on5.

So this graph grossly overestimates Winnipeg's chances and underestimates Calgary's chances because:

- it ignores special teams
- special teams has a massive impact on goal differential
- goal differential impacts wins (obviously)
- wins impact playoff probabilities (obviously)