Playoff Performers Voting Record - Canadiens1958

[sr edler]

The goal of hockey isn't defense - it's to ensure your team scores more than opposition.

Yes, and good defense plays a big part in that. How are you going to score more than your opponent if your defense sucks?

 

[Canadiens1958]

Explained

But is Dallas Smith more dominant/efficient than Orr or Robinson?

Solely by the calculation you offered earlier, he certainly is.

Explained already. Any five man unit comes with its own efficiencies/inefficiences that contribute to the whole.Bobby Orr was the motor while Dallas Smith was the most efficient d-man amongst the various pairing options.

 

[ImporterExporter]

Plus-minus has some (positive) value, and it does contain some information. Problems occur when people try to use it to draw conclusions that make no sense (like comparing the two in a ratio and calling that some measure of "dominance").

If all you have to go on are points and plus-minus, and you use the two to reach the conclusion that Player A (50 points, +25) is better than Player B (80 points, +25), then you're using the statistics wrong. It's not the statistics' fault if you call Player A more "dominant".

This is also the problem with most "advanced" statistics as well - it's not the statistics' fault, but people use them incorrectly (usually to support whatever they believed all along) and the statistics get blamed. For what it's worth, plus-minus is probably hockey's original "advanced stat".

Useless is too broad and dismissive (see: people railing against wins and GAA for goalies). Requires context and understanding of nuance. Misused or overused specific info/metrics will produce a garbage product regardless of the situation. Out right disregarding of it in totality is not dissimilar to being married to the metric itself...

Well technically speaking I did say it was "one of the most useless", not THE most useless

I agree that there is "some" use for it, but as you say, most folks don't know HOW to use it without it looking foolish as a whole. Unless you can actually re-watch every time that a player is on the ice for goals for/against, it's very hard to hard to draw any relevant conclusions.

There are countless times where a player gets a +/- and the only reason for it, is he was simply on the ice. Had no impact on the play in question. A goalie can knock the puck into his own net at ES, and unfortunately, the 5 guys skating get dinged.

 

[Doctor No]

Explained already. Any five man unit comes with its own efficiencies/inefficiences that contribute to the whole.Bobby Orr was the motor while Dallas Smith was the most efficient d-man amongst the various pairing options.

So at what point did the game of hockey change significantly enough for Dallas Smith to not be more dominant than Orr or Robinson, but for Orr and Robinson to be more dominant than Gretzky?

You're the one that offered up the comparison. Why can you offer a calculation that "proves" Orr and Robinson to be more dominant than Gretzky, but then dismiss the same calculation when it's used on Dallas Smith?

Sorry for the repeated questions; I'm legitimately trying to understand your reasoning here.

 

[Canadiens1958]

Dominance

So at what point did the game of hockey change significantly enough for Dallas Smith to not be more dominant than Orr or Robinson, but for Orr and Robinson to be more dominant than Gretzky?

You're the one that offered up the comparison. Why can you offer a calculation that "proves" Orr and Robinson to be more dominant than Gretzky, but then dismiss the same calculation when it's used on Dallas Smith?

Sorry for the repeated questions; I'm legitimately trying to understand your reasoning here.

Hockey is fluid so there is no point in time. Just as you cannot pinpoint a point in time when Dominik Hasek became a dominant goalie.

Certain players are expected to dominate - the ones in question, Howe, Lemieux, Crosby, Beliveau etc. Other teammates are expected to support. That is why the Penguins hve been succesful lately finding the appropriate supporting players for Crosby and Malkin with the likes of Sheary and Guentzel.

That is what the calculation shows. That someone expects a Dallas Smith level player to dominate that is their perogative.

Use the calculation wisely for each player level and you will get a solid picture.

 

[ESH]

Yes, and good defense plays a big part in that. How are you going to score more than your opponent if your defense sucks?

The whole point is that all you need to do to win is outscore the opponent to win, and Gretzky did that better than anyone. Defense and offense both contribute to outscoring the opposition, but why does it matter that Gretzky isn't an especially notable defensive player when he was so dominant offensively that the the ice was tilted for his team anyway?

 

[Doctor No]

Hockey is fluid so there is no point in time. Just as you cannot pinpoint a point in time when Dominik Hasek became a dominant goalie.

Certain players are expected to dominate - the ones in question, Howe, Lemieux, Crosby, Beliveau etc. Other teammates are expected to support. That is why the Penguins hve been succesful lately finding the appropriate supporting players for Crosby and Malkin with the likes of Sheary and Guentzel.

That is what the calculation shows. That someone expects a Dallas Smith level player to dominate that is their perogative.

Use the calculation wisely for each player level and you will get a solid picture.

If the calculation doesn't work, then it doesn't work. You can't "use the calculation wisely".

Your algorithm appears to be "use the calculation when it suits your preconceived notions", and at that point, the calculation loses 100% of its meaning.

If you truly believe that Orr and Robinson are more dominant, then say *that*. Don't give us a formula that's flimsier than a broken stick blade.

 

[Plural]

Using calculation wisely sounds more like an excuse for using flimsy statistical analysis as a base to build biased narratives.

 

[Canadiens1958]

Really

If the calculation doesn't work, then it doesn't work. You can't "use the calculation wisely".

Your algorithm appears to be "use the calculation when it suits your preconceived notions", and at that point, the calculation loses 100% of its meaning.

If you truly believe that Orr and Robinson are more dominant, then say *that*. Don't give us a formula that's flimsier than a broken stick blade.

No different than grading exams in math for the various senior high school levels. Same calculation, usually three to four levels in a senior class. However the interpretation is different for post high school purposes. A 95% in the pre-university stream and a 95% in the practical stream have to be interpreted wisely and accordingly. Not an algorithm issue. Just a question of wisely applying the results.

Same in hockey.

 

[Michael Farkas]

I guess if I were to make a counterpoint, though I'm unsure if I want to, it would be: there are no statistics that paint a 100% accurate, clear and consistent picture of such a broad scope. So why would this use of a statistic be expected to?

Moreover, statistics are the residual of the story, they don't make the story. The story comes first, always. So there's a story and there's some supporting statistical evidence in this case. Trying to extrapolate the evidence used in the Gretzky/Robinson example and apply it to others doesn't seem relevant.

I mean, that's kind of the business we're in here, eh? Look at the Andreychuk HHOF discussion...one guy loves the idea of 640 goals, everyone else says "yeah, but compared to..." or "because of..." and there's a story. The story overrides. In some instances, some of us used the fact that there were so many PP goals as a negative even...but it's very likely that in a discussion about Tomas Holmstrom's place in history among bottom six forwards that PP contributions would be viewed as a major positive.

That's a weak, rambling example as I recover from a weekend-long, tequila-fueled bender, but the underlying point is that evidence of this nature has a natural proximity clause, so to speak. We don't have any one-size-fits-all stats for anything, why does C1958's have to perform at that level...?

 

[Doctor No]

No different than grading exams in math for the various senior high school levels. Same calculation, usually three to four levels in a senior class. However the interpretation is different for post high school purposes. A 95% in the pre-university stream and a 95% in the practical stream have to be interpreted wisely and accordingly. Not an algorithm issue. Just a question of wisely applying the results.

Same in hockey.

Following your analogy, you appear to be grading your students' exams based on how you felt about them at the start of the semester.

If the formula doesn't work, then the formula doesn't work. And if the formula only "works" some of the time, then the formula doesn't work.

 

[Doctor No]

So why would this use of a statistic be expected to?

That's pretty much my point. His evidence that Orr and Robinson were more dominant is (at this point) entirely based on this ratio that he developed.

If Orr and Robinson were more dominant than Gretzky (and perhaps they were), then surely he can come up with better than that, don't you think?

 

[ImporterExporter]

"People in glass houses sink ships"

This is how I feel the discussion about +/- is going.

 

[Canadiens1958]

How?

Following your analogy, you appear to be grading your students' exams based on how you felt about them at the start of the semester.

If the formula doesn't work, then the formula doesn't work. And if the formula only "works" some of the time, then the formula doesn't work.

How would the future final exam results be known at the start of the semester?

 

[Canadiens1958]

False

That's pretty much my point. His evidence that Orr and Robinson were more dominant is (at this point) entirely based on this ratio that he developed.

If Orr and Robinson were more dominant than Gretzky (and perhaps they were), then surely he can come up with better than that, don't you think?

The ratio has been around for years.

 

[Doctor No]

How would the future final exam results be known at the start of the semester?

It's your analogy, I'm just interpreting it.

 

[Doctor No]

The ratio has been around for years.

Yes, the ability to divide one statistic by another statistic has been around for years.

That doesn't make the calculation reasonable, or consistent, or proper. See Dallas Smith.

Your chosen ratio doesn't *do* anything, so it really doesn't matter how long it's been around.

 

[Canadiens1958]

Peace

Yes, the ability to divide one statistic by another statistic has been around for years.

That doesn't make the calculation reasonable, or consistent, or proper. See Dallas Smith.

Your chosen ratio doesn't *do* anything, so it really doesn't matter how long it's been around.

Your reality. Peace.

 

[Doctor No]

Your reality.

If this was intended as an insult, rest assured that I'm comfortable living in this reality.

Have a good day.

 

[Canadiens1958]

?????

If this was intended as an insult, rest assured that I'm comfortable living in this reality.

Have a good day.

???? as in your experiences.

 

[Hawkey Town 18]

I'm confused on this percentage metric...

Player A scores 1 ES point and is a +1 every game: 82 points, 100%
Player B scores 2 ES points and is a +1 every game: 164 points, 50%

In terms of winning hockey games both players contributed equally yet you would say player A was more efficient or more dominant?

Dallas Smith 1970-71 +94 scoring 45 points or over 208%. Is he more efficient than the above 3?

Game is played in units of 3, 4 or five. So Dallas Smith would have been the ideal first partner for Bobby Orr. Still have a +30 difference so that would serve as an indicator that Orr had other partners.

What about the hypothetical Player A vs. Player B part of my post?...that is more important than the Dallas Smith example. It shows 2 players who are contributing equally to their teams ability to win hockey games yet one looks twice as good as the other using your metric (let's also assume they play the same role on their teams). How can you justify using a metric that does this?

 

[Canadiens1958]

Hypothetical

What about the hypothetical Player A vs. Player B part of my post?...that is more important than the Dallas Smith example. It shows 2 players who are contributing equally to their teams ability to win hockey games yet one looks twice as good as the other using your metric (let's also assume they play the same role on their teams). How can you justify using a metric that does this?

But they are hypotheticals who demonstratably are not playing the same role.

 

[Dennis Bonvie]

Aaaaand the last flurry of posts bickering about +/- goes to show how irrelevant it is as a stat.

Why bother folks? Anyone who follows hockey closely and understands the nuances of the game knows it is one of the most useless metrics available.

I would have to disagree with that.

 

[Hawkey Town 18]

But they are hypotheticals who demonstratably are not playing the same role.

A valid metric should be able to work for a realistic hypothetical. If the numbers are too big then cut them in half, the point remains the same.

By playing the same role I don't mean playing the same style I mean playing the same role for their team. Those 2 players could easily both be 1st line centers on teams that try to line match power vs. power.


EDIT: The point is, when using your metric it is possible to have 2 players equally contributing to their teams' ability to win games while one player looking twice as good. It is not a valid metric to use.

 

[Canadiens1958]

Game Situation

A valid metric should be able to work for a realistic hypothetical. If the numbers are too big then cut them in half, the point remains the same.

By playing the same role I don't mean playing the same style I mean playing the same role for their team. Those 2 players could easily both be 1st line centers on teams that try to line match power vs. power.


EDIT: The point is, when using your metric it is possible to have 2 players equally contributing to their teams' ability to win games while one player looking twice as good. It is not a valid metric to use.

Obviously you have not found an actual seasonal or playoff situation altough plenty exist. Look at Art Ross results through the ages. Playoff scoring.

Hint I, would you say Al Langlois contributed equally to Doug Harvey to the Canadiens success in 1960 or was he carried by Doug Harvey?

Hint II, compare 1979-80 Wayne Gretzky and Marcel Dionne, both scored 137 points.