it's actually 0.477 as a Ranger coming into this season
Currently .467 GPG as a Ranger.
Goals Per Game in Franchise History
1. Pat Verbeek 0.58
2. Mike Gartner* 0.54
3. Walt Poddubny 0.51
4. Pierre Larouche 0.49
5. Bill Cook* 0.48 6. Marian Gaborik 0.47
7. Jaromir Jagr 0.45
8. Don Murdoch 0.44
9. Phil Esposito* 0.44
10. Tomas Sandstrom 0.43
His scoring touch has been a little inconsistent to date this season but I am willing to go look at his entire time here. Sure, he has beaten up on some lesser opponents but you don't score that many goals against just them. He did score a hat trick against the big, bad Bruins. I have seen some people say he only scores against bad teams so, I finally went and looked at the numbers. I think the "just scores against the bad teams" is a misnomer:
Last year...5 vs Isles, 4 vs Flyers, 4 vs Bruins, 3 vs Sens, 3 vs Canes, 3 vs Leafs (2 or less vs rest of the league)
Against the top 15 teams (by winning %), he scored 21 of his 41 goals.
2011....in decidedly down year, 9 of his 22 came against top 15 teams.
2010...21 of his 42 came against top 15 teams. 4 vs Caps (best team based on winning %), 4 vs Isles, 3 vs Pens.
Um. OK. GWGs are hardly a perfect stat. Maybe not even a meaningful one.
I'd like to see a "close and late" statistic in hockey like we have in baseball. Tell me how often a guy produces points in a tied/one goal game and late in the action. It's not perfect either, but it would be far more meaningful that GWG. You can pick up a GWG in a 5-0 blowout.
I'd like to see a "close and late" statistic in hockey like we have in baseball. Tell me how often a guy produces points in a tied/one goal game and late in the action. It's not perfect either, but it would be far more meaningful that GWG. You can pick up a GWG in a 5-0 blowout.
He is what he is and many nights his style of play mimics a disinterested soft non skater but the numbers don't lie. He did have surgery so it's possible he'll be even more streaky than normal.
I'd like to see a "close and late" statistic in hockey like we have in baseball. Tell me how often a guy produces points in a tied/one goal game and late in the action. It's not perfect either, but it would be far more meaningful that GWG. You can pick up a GWG in a 5-0 blowout.
Last year I did a breakdown of each players goals scored according to the score in the game (after the goal).
GoalSig is the sum of all the goals discounted by a factor of 1/1.25^n, where n is a based on the score of the game. For goals that put the team in the lead, n is equal to the lead before the goal is scored. For goals that tie the game or see the team trailing, n is equal to the deficit after the goal. So for goals that:
Tie the game, or put the team up 1: n=0, and the goal is valued at 1
Cut the lead to 1, or put the team up 2: n=1 and the goal is valued at 0.8
Cut the lead to 2, or put the team up 3: n=2 and the goal is valued at 0.64
Cut the lead to 3, or put the team up 4: n=3 and the goal is valued at 0.512
I'd like to see a "close and late" statistic in hockey like we have in baseball. Tell me how often a guy produces points in a tied/one goal game and late in the action. It's not perfect either, but it would be far more meaningful that GWG. You can pick up a GWG in a 5-0 blowout.
Does that make the GWG any less valuable than in a 1-0 shutout? Unless the GWG in the latter came in the final minutes of regulation or OT, how can you assume either is more important? When you score the first goal of the game there's no way to know how the rest of the contest will go.
Last year I did a breakdown of each players goals scored according to the score in the game (after the goal).
GoalSig is the sum of all the goals discounted by a factor of 1/1.25^n, where n is a based on the score of the game. For goals that put the team in the lead, n is equal to the lead before the goal is scored. For goals that tie the game or see the team trailing, n is equal to the deficit after the goal. So for goals that:
Tie the game, or put the team up 1: n=0, and the goal is valued at 1
Cut the lead to 1, or put the team up 2: n=1 and the goal is valued at 0.8
Cut the lead to 2, or put the team up 3: n=2 and the goal is valued at 0.64
Cut the lead to 3, or put the team up 4: n=3 and the goal is valued at 0.512
And so on.
I would have counted the N=2,3+ at a much lower rate. Teams that are up by 2 goals tend to win (and I'm being conservative) 80%+ of the time. I would pull in those stars to determine goal value. You should look at how much the goal changes the outcome of the game
Does that make the GWG any less valuable than in a 1-0 shutout? Unless the GWG in the latter came in the final minutes of regulation or OT, how can you assume either is more important? When you score the first goal of the game there's no way to know how the rest of the contest will go.
Does that make the GWG any less valuable than in a 1-0 shutout? Unless the GWG in the latter came in the final minutes of regulation or OT, how can you assume either is more important? When you score the first goal of the game there's no way to know how the rest of the contest will go.
True.
I agree any tie breaking goal is big.
Any stat that helps us filter away the 3-0 or 4-0 goals in games that finish 5-2 or 5-3 will help us more accurately determine, which players consistently tend to perform in close games when we need guys who can make the difference when the pressure is on and the open ice is usually at more of a premium.
Last year I did a breakdown of each players goals scored according to the score in the game (after the goal).
GoalSig is the sum of all the goals discounted by a factor of 1/1.25^n, where n is a based on the score of the game. For goals that put the team in the lead, n is equal to the lead before the goal is scored. For goals that tie the game or see the team trailing, n is equal to the deficit after the goal. So for goals that:
Tie the game, or put the team up 1: n=0, and the goal is valued at 1
Cut the lead to 1, or put the team up 2: n=1 and the goal is valued at 0.8
Cut the lead to 2, or put the team up 3: n=2 and the goal is valued at 0.64
Cut the lead to 3, or put the team up 4: n=3 and the goal is valued at 0.512
And so on.
Thanks a lot for doing all this work.
What was your rationale for the values you assigned to the goals?
It seems like the chart creates a value for goals scored with extra wight for the importance. It may be interesting just to use it to dumb it down a bit and track percentage of a players goals that are tie-breaker goals or goals when the team is down or up a goal and divide that by the players total goals.
It may help us get a clearer idea, which players tend to contribute most when goals are needed the most.
Thought both he and Richards were better last night. I like Gabby a lot I really do, just wish he was more visible. No discrediting his offensive skill, but when he's invisible it drives me nuts. He's just not dynamic at all, which is ok if he's being elusive and effective.
Last year I did a breakdown of each players goals scored according to the score in the game (after the goal).
GoalSig is the sum of all the goals discounted by a factor of 1/1.25^n, where n is a based on the score of the game. For goals that put the team in the lead, n is equal to the lead before the goal is scored. For goals that tie the game or see the team trailing, n is equal to the deficit after the goal. So for goals that:
Tie the game, or put the team up 1: n=0, and the goal is valued at 1
Cut the lead to 1, or put the team up 2: n=1 and the goal is valued at 0.8
Cut the lead to 2, or put the team up 3: n=2 and the goal is valued at 0.64
Cut the lead to 3, or put the team up 4: n=3 and the goal is valued at 0.512
And so on.
Interesting stuff. How did you calculate the goal values?
What was your rationale for the values you assigned to the goals?
I just used the formula 1/X^n and tried to find an X value that gave values that seemed reasonable. Very unscientific.
Quote:
Originally Posted by The Sweetness
It seems like the chart creates a value for goals scored with extra wight for the importance. It may be interesting just to use it to dumb it down a bit and track percentage of a players goals that are tie-breaker goals or goals when the team is down or up a goal and divide that by the players total goals.
That would be easy to do. I'll get that to you soon.
