Why pay attention to game winning goals?

[plusandminus]

Since this is the history section, I might add that you are welcome to view this from a historical viewpoint. For example, when was game winning goals introduced as a more or less official stat?

Anyway, why is attention being payed to game winning goals?

Why not first goals instead? If there wasn't some guy scoring a team's first goal in a game, the team in question would not win it. If noone stepped forward to score a team's first goal, the team would end the season with 0 wins.
If a team wins 3-1. Why focus on the guy who scored the 2nd goal? Wouldn't the 1st goal be as important? If you win 4-3, well the 4th goal is great, but without the three other goals, there might not have been a win.

Why not list some other stat instead? Like number of goals forward a player have been on the ice on. Wouldn't that be an interesting stat? Or the player's +/- during even strength play only. Or the number of penalties a player has drawn (resulting in PPs), or the numbers of penalties a player has caused (resulting in SHs).

(Speaking of unimportant stats, I would think the mentioned penalty stat would be at least as important as the arbitrary giveaways and takeaway stat. The penalty stat would not be very arbitrary. The guy who gets 2 mins (or 5 mins) gets either 1 or number of penalty minutes (2 or 5) added to his "causing a SH situation" stat. The guy he interfered, or slashed, etc. gets a number added to his "creating a PP" stat. In cases like too many men on the ice, or puck out, things might get more complicated but probably can still be solved.)

Click to read more...

 

[Ohashi_Jouzu*]

I guess it's an early and yet un-evolved attempt to measure a player's "clutch" factor. Definitely loses a bit of meaning when you just happen to score your team's 3rd goal in a 7-2 game, but the majority of games are tighter scores, I guess.

I'd like to see hockey track a stat similar to basketball's new "clutch" scoring stat (points scored when teams are separated by 5 or less points); maybe make it points scored when the game is tied or separated by 1 goal, and maybe restrict it to even strength. Similarly, I'd like the NHL to track goalies' SV% in the same situations. None of it is "perfect", but it'd be interesting to see if it's possible to get any measure of a player's ability/contribution in tight games.

 

[BraveCanadian]

Since this is the history section, I might add that you are welcome to view this from a historical viewpoint. For example, when was game winning goals introduced as a more or less official stat?

Anyway, why is attention being payed to game winning goals?

Why not first goals instead? If there wasn't some guy scoring a team's first goal in a game, the team in question would not win it. If noone stepped forward to score a team's first goal, the team would end the season with 0 wins.
If a team wins 3-1. Why focus on the guy who scored the 2nd goal? Wouldn't the 1st goal be as important? If you win 4-3, well the 4th goal is great, but without the three other goals, there might not have been a win.

Why not list some other stat instead? Like number of goals forward a player have been on the ice on. Wouldn't that be an interesting stat? Or the player's +/- during even strength play only. Or the number of penalties a player has drawn (resulting in PPs), or the numbers of penalties a player has caused (resulting in SHs).

(Speaking of unimportant stats, I would think the mentioned penalty stat would be at least as important as the arbitrary giveaways and takeaway stat. The penalty stat would not be very arbitrary. The guy who gets 2 mins (or 5 mins) gets either 1 or number of penalty minutes (2 or 5) added to his "causing a SH situation" stat. The guy he interfered, or slashed, etc. gets a number added to his "creating a PP" stat. In cases like too many men on the ice, or puck out, things might get more complicated but probably can still be solved.)

Some good points here.

So much of what makes a team or even an individual successful in hockey is not captured in the stats collected - especially up until recently when much more is being captured.

I do believe that some players do rise up in important situations but, as you said, where do we draw the line on which particular situations are the important ones?

I also believe that many players receive the "clutch player" label based more upon opportunity being available to them (usually a first stringer who is put out when the team needs a goal or when defending a lead for example) and people remembering successes over failures than any extra gear the player found.

Game winning goals is a very poor measure of value. As you said, it doesn't matter if you scored the winning goal in a 6-0 game or a 2-1 game according to that stat.

Secondly, as an example, Mats Sundin has more game winning goals than Messier, Gretzky, Yzerman, Joe Sakic, Mike Bossy etc.. I don't even have to add anything to that.

 

[Iain Fyffe]

It is an attempt to measure "important" goals. A poor attempt, mind you.

I don't think there's any evidence that scoring a high proportion of GWG is a repeatable skill. That is, players who score a lot of GWG, season to season, tend just to be those that score a lot of goals on good teams.

 

[WingsFan95]

Better than nothing.

Also scoring your team's first goal when it took 4-3 to win makes it seem a lot less important also.

I know there was a posting that showed most goals scored when at even strength, down by 1 and down by 2.

Career wise it looks like this:

Brett Hull-24
Wayne Gretzky-24

With Wayne playing 200 more.

Then you go to Sakic and Claude Lemieux at 19.

You look at those 4 alone you get mixed signals.

Glen Anderson is at 17 while Mario Lemieux is at 11.

It tells you that only good-great players have those compiled, but then again only good-great players lay enough games to compile big numbers in any stat.

In fairness though if you consider Claude and Nieuwendyk Hall of Famers the only guys in the Top 13 ( 28 players ) not a Hall of Famer is Stephane Richer, Bobby Smith and active players Drury and Marleau.

 

[seventieslord]

Iain is right - it's a really poor attempt to adjudge clutch play. The problem is, there are some very clutch goals that don't prove to be game winners. And there are some very blah goals that do prove to be game winners. The GWG is very, very approximate, to the point that it tells us nothing.

 

[doktor2d]

A game-winning goal is important. The NHL, however, doesn't measure GWG correctly. Whoever decided that it's the goal that is one more than the opponent was a lunatic.

Baseball's GWRBI had it right ... it should be the goal that puts your team ahead for good.

Seriously ... Your team goes up 5-0, then the other team rallies with 3 goals. The guy who scored the fourth goal ... in a 3-0 game ... is now deemed clutch or game-winning? That's just stupid. In that game, the goal that put you up 1-0 is the most clutch goal for the winning team.

Similarly, goalies' wins and losses are judged in the NHL in a sub situation. So you can have a goalie come in relief with his team losing 4-0. If his team outscores the opponent 4-1 the rest of the way, still trailing 5-0, then 5-1, 5-2, 5-3, 5-4 ... the relief goalie gets the loss because the fifth goal is deemed the "winner". Stupid.

Game-winning goals have some meaning. The NHL, however, has no idea what a GWG is.

 

[plusandminus]

It is an attempt to measure "important" goals. A poor attempt, mind you.

I don't think there's any evidence that scoring a high proportion of GWG is a repeatable skill. That is, players who score a lot of GWG, season to season, tend just to be those that score a lot of goals on good teams.

How do you think one could get the answer to that?
If I understand you right, one should e.g. create a winning goal percentage, where one divides game winning goals by total goals scored. Different players will then have differently high values. Do you mean that players' values would not consistently be above (or below) average, season after season, but rather look a bit random?

Even if one could consistently see that certain players had above average percentage season after season, I still would rather look at other stats instead.

Also, I wouldn't just want to look at goals, but also assists, i.e. game winning points. And why settle with game winning, when one could also include game tying goals. To me, this kind of thinking is rather arbitrary.

I'd much more prefer to simply see total goals forward. That is a simple stat to understand even for the casual fan, and it would also be easy to compare with points (or goals) scored to see how involved players are in the goals forward. Showing total goals against, however, would be pretty useless and probably misleading, as it would sort of "punish" guys for doing much penalty killing. In that case, one could rather just show the number of PK minutes per game for the player.

 

[plusandminus]

Some good points here.

Thank you.

So much of what makes a team or even an individual successful in hockey is not captured in the stats collected - especially up until recently when much more is being captured.

I do believe that some players do rise up in important situations but, as you said, where do we draw the line on which particular situations are the important ones?

I also believe that many players receive the "clutch player" label based more upon opportunity being available to them (usually a first stringer who is put out when the team needs a goal or when defending a lead for example) and people remembering successes over failures than any extra gear the player found.

Game winning goals is a very poor measure of value. As you said, it doesn't matter if you scored the winning goal in a 6-0 game or a 2-1 game according to that stat.

Secondly, as an example, Mats Sundin has more game winning goals than Messier, Gretzky, Yzerman, Joe Sakic, Mike Bossy etc.. I don't even have to add anything to that.

Yes, I too think people remember successes over failures. I think a player can play pretty bad, and perhaps being critisized heavily by commentators during games, and then he suddenly - after missing open net 4 times earlier in the game - scores the game winner. (I think this often is true for some purely offensively minded, goal scoring minded, players. They may be -2 in the game, but still manage to be a hero due to their scoring ability.)
When it comes to goalies, it might be the opposite. I know otherwise skilled goalie Tommy Salo is still being remembered and taunted by Swedes, for allowing a long shot from a Belarus player many years ago. He is much more remembered for that, than for e.g. when Sweden defeated Finland 0-0 1-0 in a two game World Championship final.

Speaking of Swedes, don't you think Mats Sundin was simply great when playing for them? Definitely a clutch player, scoring many memorable and important goals, including like 3-4 and 4-4 with less than a minute left vs Finland, or the assist to Lidstrom on the 2006 Olympics game winning goal. I think he would be the kind of player who could very well have scored important goals in Stanley Cup finals, had he played in better surroundings. I would actually interpret your last sentence as a case for, rather than against, game winning goals. And of course Gretzky and Bossy scored lots of game winning goals, they were the top guns for dynasties that used to win most their games. Same with Yzerman's Detroit and eventually Sakic's Colorado. Poor Sundin, however, didn't have it that easy. He had to struggle hard to even get wins, but when the win did come, he usually had contributed. I wonder where he would rank if one could divide "game winning goals" by "games won".

 

[Bear of Bad News]

How do you think one could get the answer to that?
If I understand you right, one should e.g. create a winning goal percentage, where one divides game winning goals by total goals scored. Different players will then have differently high values. Do you mean that players' values would not consistently be above (or below) average, season after season, but rather look a bit random?

You would also need to net out the level of opportunity for players (that is, players on good teams have more opportunities for GWG simply by the nature of being on a good team).

 

[Scott1980]

Don't worry. Guy Lafleur proved that tying goals are more remembered than winnings goals!

 

[plusandminus]

Better than nothing.

Also scoring your team's first goal when it took 4-3 to win makes it seem a lot less important also.
I know there was a posting that showed most goals scored when at even strength, down by 1 and down by 2.

Career wise it looks like this:
Brett Hull-24 Wayne Gretzky-24
With Wayne playing 200 more.
Then you go to Sakic and Claude Lemieux at 19.
You look at those 4 alone you get mixed signals.
Glen Anderson is at 17 while Mario Lemieux is at 11.

It tells you that only good-great players have those compiled, but then again only good-great players lay enough games to compile big numbers in any stat.
In fairness though if you consider Claude and Nieuwendyk Hall of Famers the only guys in the Top 13 ( 28 players ) not a Hall of Famer is Stephane Richer, Bobby Smith and active players Drury and Marleau.

I know at least C.Lemieux and Drury had a "clutch" label.

I do understand that your stat sort shows "who are the best guys at scoring when having to score in order to equalize or get within one goal". That may be indicative of some kind of character.

But why would those goals really be more important? Scoring 4-2 could be just as important. It's a tight 3-2 game where the trailing team fights hard to equalize, but then player X scores 4-2, which has a major influence on the game. The trailing team loses energy, and the leading team gets energy and added calmness.

 

[plusandminus]

You would also need to net out the level of opportunity for players (that is, players on good teams have more opportunities for GWG simply by the nature of being on a good team).

Yes, if comparing the whole league. And that could be complicated as the two affects each other. Good team mates makes it easier for player to receive a win, while good player makes it easier for team to get a win. Then other factors might also affect things, like it being perhaps easier to be THE "your role is to score important goal" guy on a team, rather than playing on a team where several guys tend to share the burden.

But as a start, one could look within teams.

 

[vadim sharifijanov]

in basketball, they have a useful stat for "crunch time scoring." i think it means the last five minutes of a game, if the score of the game is within five points.

so that would be useful if translated to hockey. the first goals of a game are also really important. i'd like to see a stat or that.

 

[reckoning]

First goals have been kept by the NHL for several. NHL.com has them here under the column labeled "1G":

http://www.nhl.com/ice/playerstats.htm?fetchKey=20112ALLSASAll&sort=firstGoals&viewName=goals

 

[Iain Fyffe]

Do you mean that players' values would not consistently be above (or below) average, season after season, but rather look a bit random?

It's a very noisy number. Goal-scoring ability in general, the overall quality of your team, and the relative frequency with which you are used in tied games all have an effect on it, and can create illusions.

 

[SaxMan]

The sad thing is an empty net goal could end up being a game winning goal. If you're up by 1 and score an empty net but the opposing team manages to score before the game ends.

 

[FootKnight]

First goals have been kept by the NHL for several. NHL.com has them here under the column labeled "1G":

http://www.nhl.com/ice/playerstats.htm?fetchKey=20112ALLSASAll&sort=firstGoals&viewName=goals

Yeah, I was going to say that the NHL does actually track a lot of this stuff, you just have to get off of Summary sheet on the stats page. There's a ton of stuff in there once you find that Report drop down.

You can get goals by situation (leading/trailing by 1, 2, 3+, tied, by period, etc) for individuals players if you go to their player page and click on Splits.

 

[Iain Fyffe]

Baseball's GWRBI had it right ... it should be the goal that puts your team ahead for good.

I understand that idea - this is a goal that greatly affects your team's chances of winning the game, in a game that your team won. The problem is that it still relies on what happens after the gaol is scored to determine what it will be in the end. If your team gives up the lead, it cannot be a GWG; but if the other team does not score for the rest of the game (for example), it is.

The problem is the retrospective evaluation of an event that's required. If it's an important goal, you should be able to know it right away, and not have to wait until the end of the game to make that determination.

A better idea might be to track go-ahead goals, since they put your team ahead, and you know at the time what the goal is. But of course a game-tying goal has the same value in this sense as a go-ahead goal. You go from likely losing to a tied state, as opposed to going from a tied state to likely winning. So if you want to track GAG (go-ahead goals) you'll also have to include GTG (game-tying goals), since they're the same value in terms of affecting your team's chances in the game.

 

[rocketpop]

I'd rather see a statistic like "tie-breaking goals".

The 4th goal of the night is hardly the "clutch" if you go up 8-0 in the 1st and the losing team scores 3 late in the 3rd.

Tie breaking goals makes a bit more sense--and first goals of the night would always be considered tie-breaking. If it is 3-3 and someone scores to make it 3-4--that'd be another tie breaking goal.

 

[lextune]

First goals have been kept by the NHL for several. NHL.com has them here under the column labeled "1G":

http://www.nhl.com/ice/playerstats.htm?fetchKey=20112ALLSASAll&sort=firstGoals&viewName=goals

On this subject is the amazing playoff record that Crosby set in '09.

Virtually nothing defines a playoff game more than the first goal scored, (especially these days).

And in the '09 playoffs Crosby scored the the opening goal of six games. Even Wayne, Mario, Kurri and the like, in their playoff goal scoring glory, never put together that kind of performance.

Not incidentally, Pittsburgh won 5 out of those 6 games. And the one loss saw Crosby score all three of his team's goals in a one goal loss 4-3 to the Caps, (the dueling hat tricks night with Ovy).

 

[DJ Man]

I think I recall a stat called "Important Goals" being introduced in the 1960s.

This included first goal of the game; a goal that put you ahead; and a tying goal.

But it also included an "insurance goal" (one that put you two ahead), and a "proximate goal" (one that closed the gap to one).

So ... it was pretty damned difficult to find an unimportant goal. These only occurred when you were far ahead or far behind, and in a low scoring era, almost never.

You can see why that didn't catch on.

 

[SaxMan]

I think I recall a stat called "Important Goals" being introduced in the 1960s.

This included first goal of the game; a goal that put you ahead; and a tying goal.

But it also included an "insurance goal" (one that put you two ahead), and a "proximate goal" (one that closed the gap to one).

So ... it was pretty damned difficult to find an unimportant goal. These only occurred when you were far ahead or far behind, and in a low scoring era, almost never.

You can see why that didn't catch on.

It's a good stat if you remove the insurance and proximate goal. It would be much better than game winning goals.

 

[reckoning]

The Hockey News used to have those "important points" listed as a stat in their yearbooks in the early 80s, but it usually ended up being about 60-70% of the players overall point totals. It would be very rare that you'd see an abnormally high or low percentage.

Proximity goals and insurance goals are something I'd be inclined to say should only count in the third period, but even that depends on the situation. If someone scores a proximity goal to cut a lead to one goal with five minutes left, it changes the complexion of the game. Likewise with insurance goals; Glenn Anderson's insurance goal in Game 7 of the '87 Final was hugely important, but I wouldn't want to see empty net goals scored with five seconds left counted as equal.

 

[doktor2d]

I understand that idea - this is a goal that greatly affects your team's chances of winning the game, in a game that your team won. The problem is that it still relies on what happens after the gaol is scored to determine what it will be in the end. If your team gives up the lead, it cannot be a GWG; but if the other team does not score for the rest of the game (for example), it is.

The problem is the retrospective evaluation of an event that's required. If it's an important goal, you should be able to know it right away, and not have to wait until the end of the game to make that determination.

A better idea might be to track go-ahead goals, since they put your team ahead, and you know at the time what the goal is. But of course a game-tying goal has the same value in this sense as a go-ahead goal. You go from likely losing to a tied state, as opposed to going from a tied state to likely winning. So if you want to track GAG (go-ahead goals) you'll also have to include GTG (game-tying goals), since they're the same value in terms of affecting your team's chances in the game.

Agreed. Go-ahead goals are better. However, if you're looking for a singular goal per game, then the baseball version of GWG is better than the hockey version.

But the "same value" that you mentioned isn't true. A GTG 2 minutes into the first period has very different value than a GTG with 2 second left in the third period.

You can give value to goals by when they occur and how they change the score by comparing it with average won-loss percentages that vary on time line and goal differential.