Who Moves The Puck?
Tom Benjamin had a good post last year correcting a common mis-statement. People often refer to the Red Wings as a "puck-possession" team, but Benjamin argued that good hockey isn't so much "puck possession" but "puck position". Even the Red Wings will dump the puck into the Offensive Zone at times. There are many times in a game where a player will trade possession of the puck in order to move it much farther from his own goal.
OK how does this relate to the Thrashers and their struggles? Well last year the NHL started publishing play by play game sheets that list which players are on the ice at each stoppage of play. Vic Ferrari created a nice little program that scraps the data off so that we can see who comes off the ice after a stoppage and who goes onto the ice. This gives us a picture of "puck possession" on a player by player basis. Great hockey players (Datsyuk, Zetterberg, Ovechkin) move the puck down the ice from their Defensize Zone (DZ) to the Offensive Zone (OZ)
Why does this territorial aspect of hockey matter so much? Research by a number of Edmonton bloggers shows that teams that have the puck in the opposition end are more likely to out-shot the opposition and ultimately more likely to win the game. In summary, playing in the opposition end produces a shot advantage and shot advantage produces a goal advantage which produces a standing point advantage. Why do the Thrashers lose so many games? Because they get out chanced. Why do they get out chanced? Because they spend more of the game in their own end. You want to win more hockey games--play in the Offensive Zone.
So who is winning or losing the territorial game for the Atlanta Thrashers? To answer that question I compared the data start-of-shift data to the end-of-shift data. Specifically, I contrast puck location (the ratio of OZ/DZ) at the beginning of shift to the puck location (OZ/DZ) at the end of shifts. This gives us a measure of which players tended to move the puck down the ice from the Defensive Zone to the Offensive Zone.
Forwards Who Move The Puck Into the Opposition End at ES
+26 Reasoner
+26 Armstrong
+9 Kozlov
+7 Christensen
+4 Little
+3 Perrin
+3 Williams
+0 Sterling
-6 Thorburn
-16 White
-17 Boulton
-17 Kovlachuk
-26 Slater
Both the 3rd and 4th line are supposed to be able to play tough minutes against the other club's scoring lines. The 3rd line of Armstrong-Reasoner combo has been outstanding in terms of territory but the 4th line combo of Boulton and Slater are getting demolished. The puck is ending up back in the Thrashers zone a lot.
Among the scoring forwards the Kozlov-Little pairing is moving the puck down the ice, while White has some ugly numbers (which must have come before he was put together with Kozlov and Little).
When Kovalchuk has stepped onto the ice the Thrashers have been in their own end 110 times and in the opposition end 95 times. So at the start of shits he began in the "bad end" 15 times more than in the "good" end of the ice. At the end of Kovalchuk's shifts he was in the offensive zone just 81 times and in his own zone 113 times for a net difference of -32. So he started off in the hole 15 times and ended up in the hole 32 times. In other words on the puck is going the wrong direction. (Remember that puck location is the best predictor Winning!)
Again this is one example of why I don't consider Kovalchuk to be on the same plane as other superstars--his poor defense and carelessness with the puck hurts the team and hurts their chance of winning the game. The best players play in the other end because they make smart passes and put the puck in a position where their team is going to have success.
Here's another data point. Last year Hossa started out in the "bad" end of the ice 280 times more often than the "good" end, but at the end of shifts Hossa was only -72 which means he moved the puck from the Defensive Zone to the Offensive Zone +208. In contrast, Kovalchuk began in the "bad" more a net -28 times but by the end of his shift the puck was in the defensive zone -111 times, which means a net backwards shift of -83. When Hossa was out there the puck typically moved forward and when Kovalchuk was out the puck came backwards.
Defensemen Who Move the Puck to the Opposition End at ES
+10 Bogosian
+6 Schneider
+5 Enstrom
+3 Oystrick
-1 Valabik
-6 Hainsey
-13 Exelby
-19 Havelid
On defense we see that the much maligned Mathieu Schneider has been doing a decent job of getting the puck out of his end compared to other Thrashers defensemen. Exelby and Havelid have struggled. I'm always liked Nic Havelid but his recent struggles on the PK and suggest to me that the Thrashers should not offer him another multi-year deal after this season as he is now closer to age 40 than he is to 30.
OK how does this relate to the Thrashers and their struggles? Well last year the NHL started publishing play by play game sheets that list which players are on the ice at each stoppage of play. Vic Ferrari created a nice little program that scraps the data off so that we can see who comes off the ice after a stoppage and who goes onto the ice. This gives us a picture of "puck possession" on a player by player basis. Great hockey players (Datsyuk, Zetterberg, Ovechkin) move the puck down the ice from their Defensize Zone (DZ) to the Offensive Zone (OZ)
Why does this territorial aspect of hockey matter so much? Research by a number of Edmonton bloggers shows that teams that have the puck in the opposition end are more likely to out-shot the opposition and ultimately more likely to win the game. In summary, playing in the opposition end produces a shot advantage and shot advantage produces a goal advantage which produces a standing point advantage. Why do the Thrashers lose so many games? Because they get out chanced. Why do they get out chanced? Because they spend more of the game in their own end. You want to win more hockey games--play in the Offensive Zone.
So who is winning or losing the territorial game for the Atlanta Thrashers? To answer that question I compared the data start-of-shift data to the end-of-shift data. Specifically, I contrast puck location (the ratio of OZ/DZ) at the beginning of shift to the puck location (OZ/DZ) at the end of shifts. This gives us a measure of which players tended to move the puck down the ice from the Defensive Zone to the Offensive Zone.
Forwards Who Move The Puck Into the Opposition End at ES
+26 Reasoner
+26 Armstrong
+9 Kozlov
+7 Christensen
+4 Little
+3 Perrin
+3 Williams
+0 Sterling
-6 Thorburn
-16 White
-17 Boulton
-17 Kovlachuk
-26 Slater
Both the 3rd and 4th line are supposed to be able to play tough minutes against the other club's scoring lines. The 3rd line of Armstrong-Reasoner combo has been outstanding in terms of territory but the 4th line combo of Boulton and Slater are getting demolished. The puck is ending up back in the Thrashers zone a lot.
Among the scoring forwards the Kozlov-Little pairing is moving the puck down the ice, while White has some ugly numbers (which must have come before he was put together with Kozlov and Little).
When Kovalchuk has stepped onto the ice the Thrashers have been in their own end 110 times and in the opposition end 95 times. So at the start of shits he began in the "bad end" 15 times more than in the "good" end of the ice. At the end of Kovalchuk's shifts he was in the offensive zone just 81 times and in his own zone 113 times for a net difference of -32. So he started off in the hole 15 times and ended up in the hole 32 times. In other words on the puck is going the wrong direction. (Remember that puck location is the best predictor Winning!)
Again this is one example of why I don't consider Kovalchuk to be on the same plane as other superstars--his poor defense and carelessness with the puck hurts the team and hurts their chance of winning the game. The best players play in the other end because they make smart passes and put the puck in a position where their team is going to have success.
Here's another data point. Last year Hossa started out in the "bad" end of the ice 280 times more often than the "good" end, but at the end of shifts Hossa was only -72 which means he moved the puck from the Defensive Zone to the Offensive Zone +208. In contrast, Kovalchuk began in the "bad" more a net -28 times but by the end of his shift the puck was in the defensive zone -111 times, which means a net backwards shift of -83. When Hossa was out there the puck typically moved forward and when Kovalchuk was out the puck came backwards.
Defensemen Who Move the Puck to the Opposition End at ES
+10 Bogosian
+6 Schneider
+5 Enstrom
+3 Oystrick
-1 Valabik
-6 Hainsey
-13 Exelby
-19 Havelid
On defense we see that the much maligned Mathieu Schneider has been doing a decent job of getting the puck out of his end compared to other Thrashers defensemen. Exelby and Havelid have struggled. I'm always liked Nic Havelid but his recent struggles on the PK and suggest to me that the Thrashers should not offer him another multi-year deal after this season as he is now closer to age 40 than he is to 30.
4 Comments:
These are really great stats, Falconer! Maybe you should send these to ASG for them to read! Not sure what effect it will have though just like your comments or attendance at Town Halls for the STHs.
I agree with your assessment of Kovy. Everyone talks about how great a scorer, but I would rather have someone who scores 20-30 goals, but plays well on defense too than a 50 goal scorer who doesn't play much defense. And that is what Kovy is more often than not. I wonder what would happen if Kovy left and wonder who we would get to replace him. The rumor mill has already started with him wanting to be traded, and then the rumors are denied.
I don't know what to think, but he is playing with less heart, he is taking fewer shots on goal than last year. I was at the St. Louis game and there were times he and a few others were not taking the shot when they should have or; they are shooting from too far back like Schneider does. Anyhow, just frustrated. I hope they beat Montreal. It would be nice, but unlikely as the game is in Montreal and Montreal is on a 2 game winning streak. But a win would be nice as it is my birthday!
By Anonymous, at 2:15 PM
As always, a great analysis of the stats.
Hockey really is a simple game:
Chance of Winning=(number of shots)(inverse of average shot distance),
in other words, get close to the net and shoot as often as possible.
By Anonymous, at 1:02 PM
Didn't you have a post a while back that more or less showed that Havelid's career has a pretty clear good year/bad year rotation going on? Two years ago we blamed it on the death of his mother. This year I just can't find an excuse...
By Jennifer, at 4:35 PM
Jennifer:
I like Havelid. He still makes a lot of solid plays many night. But I think he is aging and he is perhaps more of a 2nd pairing defender now rather than a 1st pairing defenseman. But most night he and Enstrom are matched against the top opposition players.
On a playoff caliber team Havelid-Enstrom are a fine 2nd D pairing, but not the top duo.
By The Falconer, at 5:41 PM
Post a Comment
<< Home