What is the purpose of the game winning goal statistic?

[Kiekko 67 Turku]

What is the purpose of the game winning goal statistic? Let's use the scoring line from game 4 of the Capitals-Knights finals series to examine this:
1st Period
09:54 WSH 1-0 T.J. Oshie (8) Evgeny Kuznetsov and Nicklas Backstrom
16:26 WSH 2-0 Tom Wilson (5) Evgeny Kuznetsov
19:39 WSH 3-0 Devante Smith-Pelly (6) Matt Niskanen and Alex Ovechkin
2nd Period
15:23 WSH 4-0 John Carlson (5) Evgeny Kuznetsov and T.J. Oshie
3rd Period
05:43 VEG 4-1 James Neal (6) Erik Haula and Colin Miller
12:26 VEG 4-2 Reilly Smith (4) Jon Marchessault and Luca Sbisa
13:39 WSH 5-2 Michal Kempny (2) Nicklas Backstrom and T.J. Oshie
18:51 WSH 6-2 Brett Connolly (6)
According to how the GWG is defined, the GWG is awarded to Smith-Pelly, as that gave the Capitals a 3-0 lead and the Knights ended up with 2 for the entire game.

1. We know it isn't meant to show which goal won the game, as Smith-Pelly's goal didn't win the game. It merely made a 2-0 undecided game a 3-0 undecided game.
2. We know it isn't meant to show which goal was the most clutch goal, as at the time of Smith-Pelly's goal, there was no way of knowing that Vegas would end up with 2 goals. Nor is there anything to show that Smith-Pelly's goal was any more important than any of the other Capitals goals.
3. Some people would argue, "if Smith-Pelly didn't score, then the Capitals wouldn't have won". That argument doesn't hold up, as if Smith-Pelly didn't score, then we don't know what would have happened. And if we assume all other scoring remained the same, then we end up with a 5-2 Capitals victory.

So I ask, what purpose does this statistic serve? Does the NHL know it has no merit? Perhaps it could just be a marketing gimmick. Any thoughts out there?

Click to read more...

 

[Canadiens1958]

Point of no return in a game is defined by the GWG.

 

[Kiekko 67 Turku]

Point of no return in a game is defined by the GWG.

Are you saying that when the Capitals went up 3-0 in the example used in the OP, that's the point in the game that the Knights chances of winning the game were reduced to 0%?

 

[Canadiens1958]

No. Simply that the Knights never got to three goals in that specific game.

 

[Kiekko 67 Turku]

No. Simply that the Knights never got to three goals in that specific game.

It doesn't seem like you're adding anything useful, as I already acknowledged that in the OP. Perhaps you can elaborate.

According to how the GWG is defined, the GWG is awarded to Smith-Pelly, as that gave the Capitals a 3-0 lead and the Knights ended up with 2 for the entire game.

 

[Doctor No]

The game-winning goal was part of an attempt by hockey to offer more statistics in a fashion similar to baseball in its infancy. Similarly with winning and losing goaltenders.

They both do exactly as they claim to do, and both are about as useful as the user deems them to be. Tying the above together, the game-winning goal is necessary to determine the winning (and losing) goaltenders in a game.

 

[Kiekko 67 Turku]

The game-winning goal was part of an attempt by hockey to offer more statistics in a fashion similar to baseball in its infancy. Similarly with winning and losing goaltenders.

That makes sense, as my understanding of the history of hockey is that it lacked a strong base of statistics which were present in other sports.

They both do exactly as they claim to do, and both are about as useful as the user deems them to be. Tying the above together, the game-winning goal is necessary to determine the winning (and losing) goaltenders in a game.

The game winning goal as defined by the statistic doesn't always win the game. The only situation I can think of where it actually wins the game is if it is scored in overtime. In the OP example, Smith-Pelly's goal didn't win the game, but just gave the Capitals a 3-0 lead late in the 1st period.

 

[Doctor No]

The game winning goal as defined by the statistic doesn't always win the game. The only situation I can think of where it actually wins the game is if it is scored in overtime. In the OP example, Smith-Pelly's goal didn't win the game, but just gave the Capitals a 3-0 lead late in the 1st period.

Agreed - I chose the wording that I used above very carefully. As they are both defined in the sport, "the game-winning goal is necessary to determine the winning (and losing) goaltenders in a game".

 

[Kiekko 67 Turku]

Agreed - I chose the wording that I used above very carefully. As they are both defined in the sport, "the game-winning goal is necessary to determine the winning (and losing) goaltenders in a game".

If true, then the winning and losing goaltenders can be fairly arbitrary. Let's say goaltender A plays 57 minutes and leaves with his team up 3-0. With goaltender B minding the net, his team quickly gives up 2 goals and it's 3-2. They then go up 4-2 and then goaltender B gives up another goal with the final being 4-3. Goaltender A played 57 minutes without giving up a single goal, while goaltender B played 3 minutes and gave up 3 goals. Yet the rule says goaltender B gets the win. It doesn't seem right to me. That raises the question of what value is added by having a goaltenders W-L record?

And why is it called a "game winning goal" if it didn't win the game? It's considerably misleading to apply such a label to a late 1st period goal which puts your team up 3-0.

 

[Doctor No]

Yes, you've made your point.

 

[Kiekko 67 Turku]

Yes, you've made your point.

My point was clear in the OP.

I'm not quite sure what your point was other than that hockey wanted to throw statistics out there to keep up with other sports. If they were going to do that, then why not come up with more meaningful stats such as +/- per minute played, points per minute played, save percentage, etc. rather than obtuse, arbitrary, misleading and meaningless stats like Game Winning Goal and Goaltenders W-L?

 

[Canadiens1958]

My point was clear in the OP.

I'm not quite sure what your point was other than that hockey wanted to throw statistics out there to keep up with other sports. If they were going to do that, then why not come up with more meaningful stats such as +/- per minute played, points per minute played, save percentage, etc. rather than obtuse, arbitrary, misleading and meaningless stats like Game Winning Goal and Goaltenders W-L?

Most of your suggestions were kept but not made public. or rate as default or semi default stats, especially in the one goalie era or one unit with limited substitution era.

GWGs were never a default stat.

Part of a descriptive Key Goal package that included, first goal of the game,winning goal, go ahead goal, clinching goal, tying goal, PPG, PKG, ESG often broken down by opponent, home and away.

 

[Kiekko 67 Turku]

Most of your suggestions were kept but not made public. or rate as default or semi default stats, especially in the one goalie era or one unit with limited substitution era.

That would make sense with goaltender's W-L record - especially if as you say, each team has only one goaltender per game. However, a goaltender's W-L record is 50% dependent on how many goals the team scores, something the goaltender has practically no control over. For example, let's say the Canucks usually score 6 to 8 goals a game but have a goaltender who typically gives up 4 to 5 goals a game. Meanwhile, the Ducks usually score only 1-2 goals a game, while their goaltender typically gives up 3 to 4 goals a game. The Ducks goaltender is better than the Canucks goaltender but is going to have a worse W-L record due to things outside his control.

Part of a descriptive Key Goal package that included, first goal of the game,winning goal, go ahead goal, clinching goal, tying goal,

Correct me if I'm wrong - first goal of the game, winning goal and clinching goal there can be only one per game. Tying goal and go ahead goal - there can be more than one per game. And how is winning goal different from clinching goal?

PPG, PKG, ESG often broken down by opponent, home and away.

I know PPG, but can you tell me about PKG and ESG? And why not use PPG per minute played rather than total PPG?

 

[Doctor No]

My point was clear in the OP.

I'm not quite sure what your point was other than that hockey wanted to throw statistics out there to keep up with other sports. If they were going to do that, then why not come up with more meaningful stats such as +/- per minute played, points per minute played, save percentage, etc. rather than obtuse, arbitrary, misleading and meaningless stats like Game Winning Goal and Goaltenders W-L?

Sorry, I thought you were looking for a conversation, not a monologue.

Carry on - enjoy the thread.

 

[Kiekko 67 Turku]

Sorry, I thought you were looking for a conversation, not a monologue.

Carry on - enjoy the thread.

Thanks! I'm actually interested in a conversation about this - ideally with the goal of exchanging viewpoints and ultimately arriving at what the NHL is trying to accomplish by tallying and publicizing the game winning goal statistic.

 

[Michael Farkas]

I think that was largely answered...it's an ancillary output stat that's simple to collect. Like most public statistics, it requires more refinement to attempt usefulness...

 

[Canadiens1958]

That would make sense with goaltender's W-L record - especially if as you say, each team has only one goaltender per game. However, a goaltender's W-L record is 50% dependent on how many goals the team scores, something the goaltender has practically no control over. For example, let's say the Canucks usually score 6 to 8 goals a game but have a goaltender who typically gives up 4 to 5 goals a game. Meanwhile, the Ducks usually score only 1-2 goals a game, while their goaltender typically gives up 3 to 4 goals a game. The Ducks goaltender is better than the Canucks goaltender but is going to have a worse W-L record due to things outside his control.


Correct me if I'm wrong - first goal of the game, winning goal and clinching goal there can be only one per game. Tying goal and go ahead goal - there can be more than one per game. And how is winning goal different from clinching goal?


I know PPG, but can you tell me about PKG and ESG? And why not use PPG per minute played rather than total PPG?

This is no going well. Regardless.

PPG = power play goals.
PKG = penalty killing goals.
ESG = even strength goals.

NHL teams dressed one goalie per game unless they were willing to dress one skater less until the 1964-65 season when two goalies were mandated.

Clinching goal produces at least a two goal win.

Comparing Home and Away records puts your 50% claim into the dubious column.

1917-1942 fall the NHL featured RS overtime without the loser point and the possibility of multiple overtime goals. GWGs as a result had two components - regulation and overtime.

Outlier examples do not work well. Avoid them.

 

[Kiekko 67 Turku]

I think that was largely answered...it's an ancillary output stat that's simple to collect. Like most public statistics, it requires more refinement to attempt usefulness...

Help me understand. What is the value in having an ancillary statistic which has no meaning but is used because it is simple to collect?

 

[Michael Farkas]

Yeah, extreme examples...though fun, aren't really helpful...

As a degenerate gambler, roughly half of all games (46-48% as of a couple years ago) are one-goal games...so the game-winning goal, as it were, seems to matter a little more than given credit for...which is none, as it stands.

 

[Michael Farkas]

Help me understand. What is the value in having an ancillary statistic which has no meaning but is used because it is simple to collect?

Same thing as save pct. No correlation to talent. No correlation to winning. But you love it (evidently). Saves don't win games. It's merely an averaging stat tied almost directly to shot volume against...we all draw the line somewhere. I don't care for either of them, you have chosen to split the difference...I'm not sure who is better off haha

 

[Kiekko 67 Turku]

Yeah, extreme examples...though fun, aren't really helpful...

As a degenerate gambler, roughly half of all games (46-48% as of a couple years ago) are one-goal games...so the game-winning goal, as it were, seems to matter a little more than given credit for...which is none, as it stands.

The example used in the OP isn't an extreme example. More than half of the goals which are credited as the GWG don't actually win the game. And a fairly good portion of GWGs occur prior to the third period - IOW, it was just padding an existing lead at the time it was scored. A lot of goals which don't end up being a GWG meet that same criteria.

A one goal game in which one team goes up 3-0 and ends up winning 3-2 causes the GWG to go to the 3rd goal for the winning team, which put them up 3-0. Help me understand what is so special about that goal such that it should be given this special status.

 

[Michael Farkas]

I'm not here to argue with you. You're not wrong. I don't value the stat. The public, generally, likes to feel like they're getting more insight than they actually do...so the NHL peddles out things over time that people think are interesting, but really don't tell a ton about the game: individual shot totals, save pct., corsi, etc. people get to gum on it for a while then realize that it's not worth (much of) a damn and then wait for the next thing...

You don't like GWGs...few do. It's one in a long line of empty or near-empty stats that are available to the public...

Maybe start the counter to GoFundMe, where you raise money to make things go away, call it Go****Yourself...I'd donate...

 

[Canadiens1958]

The example used in the OP isn't an extreme example. More than half of the goals which are credited as the GWG don't actually win the game. And a fairly good portion of GWGs occur prior to the third period - IOW, it was just padding an existing lead at the time it was scored. A lot of goals which don't end up being a GWG meet that same criteria.

A one goal game in which one team goes up 3-0 and ends up winning 3-2 causes the GWG to go to the 3rd goal for the winning team, which put them up 3-0. Help me understand what is so special about that goal such that it should be given this special status.

The ordering of the goals has to be accounted for and defined.

Regardless if the team does not score the third goal they do not win.

 

[Kiekko 67 Turku]

I'm not here to argue with you. You're not wrong. I don't value the stat. The public, generally, likes to feel like they're getting more insight than they actually do...so the NHL peddles out things over time that people think are interesting, but really don't tell a ton about the game: individual shot totals, save pct., corsi, etc. people get to gum on it for a while then realize that it's not worth (much of) a damn and then wait for the next thing...

You don't like GWGs...few do. It's one in a long line of empty or near-empty stats that are available to the public...

Maybe start the counter to GoFundMe, where you raise money to make things go away, call it Go****Yourself...I'd donate...

Sounds like you agree with my theory that it's just a marketing gimmick - but you use different words.

 

[Kiekko 67 Turku]

Regardless if the team does not score the third goal they do not win.

How do you know that?