This is a bit wonky, if you ask me because I think the Wings’ Zetterberg, Datsyuk, and Lidstrom are below the average of clutch points because EVERYONE ELSE on the Wings is scoring the clutch goals.  I agree with Babcock in that he keeps pushing all season long for other people to score goals.  He doesn’t want too much pressure on Zetterberg, Datsyuk, Lidstrom, and even Holmstrom to do all the scoring.

I wonder if there’s a list of the distribution of GWGs scored by the Wings.

Posted by SYF from Las Vegas, NV on 02/05/08 at 12:36 PM ET

Here is your GWG list, SYF.

Posted by Paul from Motown Area on 02/05/08 at 12:39 PM ET

It’s a very interesting concept, but based on the sketchy description provided RE: what constitutes “clutch” vs. “soft” goals, assists, and points (as well as their respective values), it seems like mitigating factors such as total goals scored by the team could skew the ratings in the favor of players whose teams score less. In other words, a goal scorer on a low-scoring team is inherently going to score more clutch/important goals than one on a high-scoring team. I wonder if there are factors built in to balance the results?

Posted by YzermanZetterberg on 02/05/08 at 12:49 PM ET

Thanks, Paul! 

Holmstrom and Datsyuk do lead the Wings in GWGs but after them, there are 12 other guys scoring GWGs.  Hell, Lilja has two GWGs.

One other thing:  I find it striking that on that list the guys on top of those lists are on losing teams and lead the team in scoring…until you get to the list of defensemen where Lidstrom and Rafalski are at the top two spots.

I’m not making any sense right now and I don’t know why.

LOL!

Posted by SYF from Las Vegas, NV on 02/05/08 at 12:57 PM ET

Another point…it also appears that this system only provides offensive rankings. Isn’t playing strong defense “clutch” too?

I would really be interested to see the values assisgned to individual items (i.e., how much a goal is worth if you’re down by one goal in the third period vs. the first goal in a scoreless first period, etc.).

Posted by YzermanZetterberg on 02/05/08 at 12:57 PM ET

Oh, boy.

An argument could be made that a clutch player actually scores fewer goals than a non-clutch player, as a perception of being “clutch” is scoring disproportionately in key situations, and less in more “average” situations.

Somehow you need to control for the fact that some players score an enormous proportion of their team’s goals (Ovechkin), so of course more of their goals will be “clutch” since they account for (seemingly) 90% of the offense generated by their team.

Several solid players on one team (Lidstrom, Datsyuk, Zetterberg) will dilute the “clutchness” so there is a bias toward the brightest light on a team without much else to go to scoring-wise.

It doesn’t account for the difference some players have in regular season vs. playoff performance, or who comes up big in the more significant games.  Is it more “clutch” to score a tying goal in the third period against a non-playoff team in January when your team has a 10 point lead in the division, or to score the goal in the first period to push your team to a three-goal lead in a game agianst a divisional rival in late March?

Despite all the hand-waving in the affirmative, does “clutch” even exist?

Now my brain is tired. 

(And those are just the concerns I have right off the hop.)

Posted by Baroque from Michigan on 02/05/08 at 01:01 PM ET

That’s precisely right, ‘Roque.  Hence, I’m skeptical about the validity of this or any methodology to put a number on a nebulous quantity.

[looks at what he just wrote and goes cross-eyed]

What the hell did I just say?  I’m confused by my own writing now.  I need more Cranberry/Pomegranate juice.

Posted by SYF from Las Vegas, NV on 02/05/08 at 01:20 PM ET

Using their criteria, Wings players don’t stand a chance of being “clutch” players. In addition to what’s already been mentioned, consider these elements that Dr. Lakra weighs heavily in his (her?) assessment.

Goal Differential at the time the goal is scored: The Wings outscore the opposition by a huge margin compared to the rest of the league, so they’re almost always ahead in a game. Hard to score a “clutch” tying or tie-breaking goal if you’re always ahead.

Period of game: The Wings are #1 in scoring 1st period goals. They’re #2 in scoring 2nd period goals. I assume those are “non-clutch” periods in this ranking. Guess the Wings will just have to sit back and let teams get a lead so they can prove they’re “clutch” performers.

Strength of opposing team: Based on their record, every team the Wings play is “weaker” than they are.

Defense of opposing team: Ditto!

This “system” is so flawed it isn’t funny. As others have said, it’s a heck of a lot easier to be a “clutch” player on a bad team. So I guess Hank and Pavel and Nick will just have to settle for simply being important “non-clutch” cogs in a great wheel (pun intended).

Posted by OlderThanChelios from Grand Rapids on 02/05/08 at 02:14 PM ET

Oddly enough, that makes perfect sense to me.

Scary.

Part of the problem at least is how to define “clutch” in the first place.  Some people use it to describe a player who is likely to deliver a goal in a key situation because he is so good anyway that he is likely to deliver a goal at any point (to me that isn’t clutch, it’s just really good) - and some people use it to describe a player who is more likely to deliver when the stakes are high than he is at any other time (and to me, that is either lucky or the result of a spotty work ethic so he only really cares if someone is watching - otherwise the game doesn’t much matter to him).

I think clutch situations exist, but not clutch players.

By that I mean that a tying goal scored in the dying seconds of a game against a team that is tied with you in the standings is definietly a clutch goal - high stakes, intense pressure, important implications, a high level of defense, whatever way you want to define it.

I just don’t think it is possible to define certain players as “clutch players” because they uniquely deliver goals in just those situations at a disproportionate rate to the goals they score at other times during the game (all other factors, ice time, opposition, etc. being equal - which they aren’t but assume for a minute).

In the closing minutes of a game, the players on the ice will be the only ones with an opportunity to be “clutch”, and that will be a mixture of players the coach trusts because they are simply good players, players who have been playing very well lately, who happens to be stuck on the ice because a team can’t get in a change, etc.

Someone like Zetterberg is more likely to score a “clutch” goal in that situation than, say, Brett Lebda, but he is a) more likely to score a goal at any point, period and b) more likely to be on the ice in the closing minutes.

It’s expected for someone like Mats Sundin to score a goal late in a game, and less expected for Nik Antropov to do the same.

So if Antropov scores, does that make him more clutch than Sundin?

Really long way of saying that perception colors our views of who is clutch or not, when most of the time the players thought of as clutch should just be thought of as excellent players, period.

I have a hard time with any statistics that have little-to-zilch predictive power.

Posted by Baroque from Michigan on 02/05/08 at 02:20 PM ET

I just realized how dang long that comment was.  Sorry! 

Posted by Baroque from Michigan on 02/05/08 at 02:22 PM ET

I posted a mini-manifesto about what was wrong over at the blog where this originated from. I’ll refrain from pasting it here, but feel free to check it out.

“Clutch” is the idea that you are better in high-pressure or important situations than you are in normal situations. This is difficult, if not impossible, to measure in a free-form game like hockey. If hockey were turn-based, like baseball, this would be easier to at least try and quantify.

To measure “clutchness”, someone would have to tally up all of the ice time every player in the league has played this year. That would have to be broken down into “situational ice time”. While there are many ways to break it down, I would propose the following categories:

- Tie game (most “clutch”)
- Trailing by 1 (second-most “clutch”)
- Trailing by 2 (third-most “clutch”)
- Leading by 1 (fourth-most “clutch”)
- Everything else (“normal” situation)

Once that was done, we’d know how much ice time each player in the league played in the above five scenarios. From there, we’d have to break down the number of goals and assists that each player tallied in each of these five situations.

Here is where it looks like this blog went wrong—it stopped here. You have to keep going. Right now, these statistics would reward those players on teams that are in tie games or trailing the most often. To adjust this, the number of points scored in each of the five categories would have to be divided for each player to come up with the number of points each player scores in each situation per unit of time (let’s say, a minute).

From here, we compare the number of points each player scores per unit of time in each situation back to the “normal”, “non-clutch” situation of when a player’s team is leading. Which ever player shows the biggest average increase in points per unit of time between the four “clutch” categories when compared to the “normal” category is the most “clutch” player.

Obviously, there are lots of problems. What about defense? That’s one that comes to mind. This sort of thing would be be nothing more than an offensive quality rating.

Here’s the rub, and the point that I’m trying to make. No matter what, the best players are the best players. The best players will not show greater improvement in “clutch” situations necessarily because the best players are consistent players that put out similar efforts and results regardless of the situation. So, such a study could reveal that Link Gaetz was the most “clutch” player of the 1991-‘92 season, and frankly, that would mean absolutely nothing. To anyone.

Posted by Nathan on 02/05/08 at 02:36 PM ET

I said:

Really long way of saying that perception colors our views of who is clutch or not, when most of the time the players thought of as clutch should just be thought of as excellent players, period.

Nathan said:

Here’s the rub, and the point that I’m trying to make. No matter what, the best players are the best players.

Great minds think alike, Nathan?

Or we’re both equally deranged.  Take your pick. 

Posted by Baroque from Michigan on 02/05/08 at 02:42 PM ET

Hey all… I helped out on the clutch factor so if you have any questions I’d be happy to answer them for you as best I can.

First off, I’d like to note that the stat is not separating good players from bad players, but is trying to produce a way to understand the goals/assists on a different level to see the differences between how particular players get those goals.

That being said, I’d like to address the ‘so flawed its funny’ statement. And add that Detroit players not being at the top of the list does not necessarily mean its broken.

Goal Differential at the time the goal is scored: The Wings outscore the opposition by a huge margin compared to the rest of the league, so they’re almost always ahead in a game. Hard to score a “clutch” tying or tie-breaking goal if you’re always ahead.

Good point. Clutch is (loosely) defined as scoring under pressure. It makes sense then that Detroit is not scoring many clutch goals and the players are not at the top of the list. If they’re not under pressure, then its tough to be considered clutch. Unless you can explain how a goal when a team is up 3 nothing in the third is clutch, I think you have to accept this.

Period of game: The Wings are #1 in scoring 1st period goals. They’re #2 in scoring 2nd period goals. I assume those are “non-clutch” periods in this ranking. Guess the Wings will just have to sit back and let teams get a lead so they can prove they’re “clutch” performers.

I’d not suggest that. But I still don’t think scoring in the first and having regular blow outs against the weaker competition is clutch or pressure performances.

Strength of opposing team: Based on their record, every team the Wings play is “weaker” than they are.
Defense of opposing team: Ditto!

Its relative to the league. If Detroit is playing a good team, they get the same bonuses Toronto gets.

This “system” is so flawed it isn’t funny. As others have said, it’s a heck of a lot easier to be a “clutch” player on a bad team. So I guess Hank and Pavel and Nick will just have to settle for simply being important “non-clutch” cogs in a great wheel (pun intended).

Hmm… again, I don’t think its saying ‘terrible team’ or ‘terrible player’. Its saying ‘team/player scores most of their goals in non-pressure situations’. One of the goals of the stat is to provide some reasons why some teams/players do great in the regular season, but falter under that pressure when the competition stiffens up. Some teams/players lack practice in pressure and falter for this reason. (ie: Ottawa). In a seven game series though against increasing competition, you are bound to hit serious pressure. Its why sports are awesome.

Again, the stat is not evaluating how good a player is, its trying to quantitatively show how teams play to give the stats more meaning. In the context of an MVP debate I think its totally relevant. A guy on a team of superstars is not really the MVP because his team would probably dominate without him. A guy on a team that should not be doing as well as they are, with one player pushing them over the top is an MVP, which tends to be the trend of the stat.

Any feedback would be great… check out the blog in the coming days where I believe the system will be more throughly explained.

Go Wings!

Posted by indiefan from Canuckada on 02/08/08 at 11:10 PM ET