Prove it. I am a data scientist and expert in that field. I am also a lifelong hockey fanatic. I have run the regressions myself and the correlation is absolutely there. I don’t even know how people could even begin to think otherwise. It’s a baseline for the advanced stats/metrics that have been developed over the last five years. It started with the correlation of shots to wins, then a derivative of that was created to correlate possession to shots. Now scenario based possession, including competition quality, situational variables, etc are being added in. It’s on the verge of a big data play in which virtual cognition will start to build algorithms that can create an inifinite number of scenarios compared to the few we judge now.

I absolutely understand advanced hockey analytics and absolutely understand there are outliers and sometimes they have a good reason, and sometimes they don’t. But trying to deny that shot differential isn’t the base of it all is ridiculous and just can’t be proven.

Based on that, Tanners analysis is spot on. The coyotes dominated in all of the baseline stats/metrics and if they continue that same rate they will be a playoff team.
- Dahlmanyotes

It's not about the science, its about the application. Based on your post here, you're way too focused on the science. So many factors go in to shot metrics. Game situations, injury situations, strategy, even systems and style of play.

Tanner used shot metrics in a small sample size of a game. As has been pointed out to you, last season Carolina led the league in CF% and did not make the playoffs. If you're as informed as you think you are, you'd make the statement that if they continue to play like they'll have a good shot to make the playoffs rather than they will be a playoff team. Analytics available to fans just aren't that good. What you call outliers are not outliers at all.

Most of what I said here is general because I don't have time right now to post more in depth, but I'll be happy to continue this conversation further on this thread later on today.

Chicago is undefeated in regulation, With points in every game played. I predict that if they continue this, They will be a playoff team.
I am also an expert at things.....See!
- camfor

Baffling. Just baffling. Please pick up a the Statistics for Dummies book and a coffee before you comment. Your comment doesn’t even make sense.

It's not about the science, its about the application. Based on your post here, you're way too focused on the science. So many factors go in to shot metrics. Game situations, injury situations, strategy, even systems and style of play.

Tanner used shot metrics in a small sample size of a game. As has been pointed out to you, last season Carolina led the league in CF% and did not make the playoffs. If you're as informed as you think you are, you'd make the statement that if they continue to play like they'll have a good shot to make the playoffs rather than they will be a playoff team. Analytics available to fans just aren't that good. What you call outliers are not outliers at all.

Most of what I said here is general because I don't have time right now to post more in depth, but I'll be happy to continue this conversation further on this thread later on today.
- MJL

I will ABSOLUTELY concede that I shouldn’t have said they “will” be a playoff team. The proper sentence is that they will most likely be a playoff team. This actually backs up what I have been saying though. There are outliers and the coyotes could be an outlier. But if you are building a team you want to give yourself the highest probability of success. Nothing is a guarantee, just a probability. And the probability of making the playoffs is MUCH higher if you have a positive shot differential on the season. The probability of making the playoffs is MICH higher when you are a positive possession team.

Is it guaranteed? Of course not. But when evaluating a team you play the probabilities and that is exactly what Tanner did. If the coyotes continue to have the shot and possession metrics that they had in game 1 and 2, they will have one of the highest probabilities in the league to make the playoffs.

It’s not even a debate. It’s as simple as statistical math and hockey analytics gets. I am baffled that people are even trying to argue it. And left laughing and shaking my head at some of these silly rebuttals that you and others are backing in to?!?!?

What is embarrassing is that you seem to think correlation = causation.
- ruttager17

Sick burn! I would love for you to explain yourself because I feel like after a quick Wikipedia lesson you’d be fumbling your ay through a nonsense response.

Of course I know the difference between correlation and causation. Clearly you don’t based on that being your rebuttal to my statement.

It’s as simple as math/stats/hockey analytics get. If the coyotes continue to play the way they did the first two games and post the same or similar shot/possession metrics then they will have one of the largest probabilities to make the playoffs. There is no way anyone can debate that...it’s just math. Just like if you smoke for 30 years you have a higher probability to get lung cancer than I do. Doesn’t mean I won’t get lung cancer. Doesn’t guarantee you will. But if your a third party forced to place a bet, you’d be an idiot to bet on me. But in your case you would and you’d be singing some toddler song about how “correlation does not equal causation” as you did it.

The same is true for the coyotes and their performance through the first two games.

If you really don’t understand what he was saying, then I don’t know what to say. When you look at the whole game, the coyotes were the better team. There was a two minute stretch where the stars were and it resulted in a loss. Statistics prove that out. I am baffled that you are not able to understand that and actually wasted time trying to question it in this board.

I guess I assume most of the people on here are adults, but I should remember you may be a 12 year old still learning the basics of math.
- Dahlmanyotes

Actually, I'm a 48 year old, who has covered hockey professionally for 20 years, seen thousands of NHL, AHL, junior and European games and understands the game itself pretty damn well even though I'm your basic beer league 4th liner when I've attempted to play it.

When you look at the whole game, the Coyotes were trailing 3-0 for the majority of it. Also if you ask Tocch, whom I've known for years, he'd tell you a) that hockey games can be won or lost in a short segment and, just as important, b) the teams that win consistently know how to win many different ways (including in such games) whereas teams that don't win have endless rationales and excuses. A Corsi victory isn't a victory unless you put more pucks in the net. Shot quality matters more but, by the same token, "expected goals" aren't goals when you can't finish and a player with a high "points per 60 mins" doesn't necessarily warrant more ice time nor would he be likely to better help his team win if he got it.

Yes, there is a process and, yes, shot metrics are part of that process. But I watched that entire Stars-Coyotes game and Tanner's metrics argument was a joke. The Coyotes had their moments but, once the Stars got the bang-bang-bang outburst and Arizona was in chase-the-game mode, the Stars' control of the game never felt in much jeopardy regardless of what the metrics said.

Hockey has a math element, of course, but isn't math per se. To steal a quote from Mark Howe, "In hockey, if A is greater than B and B is greater than C, it doesn't always mean is A is greater than C." In math, that's the case 100 percent of the time. In hockey, Team A could sweep B in a season series and B could sweep C, yet C sweeps A.

Prove it. I am a data scientist and expert in that field. I am also a lifelong hockey fanatic. I have run the regressions myself and the correlation is absolutely there. I don’t even know how people could even begin to think otherwise. It’s a baseline for the advanced stats/metrics that have been developed over the last five years. It started with the correlation of shots to wins, then a derivative of that was created to correlate possession to shots. Now scenario based possession, including competition quality, situational variables, etc are being added in. It’s on the verge of a big data play in which virtual cognition will start to build algorithms that can create an inifinite number of scenarios compared to the few we judge now.

I absolutely understand advanced hockey analytics and absolutely understand there are outliers and sometimes they have a good reason, and sometimes they don’t. But trying to deny that shot differential isn’t the base of it all is ridiculous and just can’t be proven.

Based on that, Tanners analysis is spot on. The coyotes dominated in all of the baseline stats/metrics and if they continue that same rate they will be a playoff team.
- Dahlmanyotes

I've been down this road with these guys so many times. It doesn't matter that you're right beyond any reasonable doubt - they do not care. They only care about bothering you, so they'll go off on tangents about one little thing which they completely take out of context.

It's not about the science, its about the application. Based on your post here, you're way too focused on the science. So many factors go in to shot metrics. Game situations, injury situations, strategy, even systems and style of play.

Tanner used shot metrics in a small sample size of a game. As has been pointed out to you, last season Carolina led the league in CF% and did not make the playoffs. If you're as informed as you think you are, you'd make the statement that if they continue to play like they'll have a good shot to make the playoffs rather than they will be a playoff team. Analytics available to fans just aren't that good. What you call outliers are not outliers at all.

Most of what I said here is general because I don't have time right now to post more in depth, but I'll be happy to continue this conversation further on this thread later on today.
- MJL

What I find hilarous is no one ever says "it was only one game and a score of 3-2 is a really small sample size" .........

If you really wanted to know who the best team was, you wouldn't even want to know the score because it only subverts your objectivity. I mean, if one game of possession stats is too short of a sample (which it is) then goals are up to ten times worse of a stat to go on.

The point is that you do get information out of shot-attempts because you can build ten times the sample in one game as you can from goals. Sure there are things like shot quality etc. but the fact is that over a game, it's incredibly unlikely that you could significantly hold possession while not getting any high quality chances.

And shot attempts don't factor in goaltending which is incredibly random. Last year, the Hawks, the Canes and the Flames would all have been excellent teams had they received even league average goaltending.

But if you don't ignore goaltending when you evaluate the team you'll never ever get any usable conclusions. But we don't totally ignore goalies, we just average out their stats/ For example, we know that if a goalie is posting a save percentage well above league average, that it will regresses and vice versa.

But the dumbest thing is that "analytics available to fans are not good." That is just incredibly wrong. I know you're basing it off one Bob McKenzie quote, but you might also want to realize that Bob has a long history of being against analytics and his network has lost so much, if not all, of their previous authority because of analytics.

I mean, just look at the work people like Tyler Dellow, Dom C, Sean Tierney, Ian Tulloch, Manny whose last name I forget and Steve Burch are doing with these "bad stats," anyways I hate that I get sucked in talking to people who will never take in new information or consider how wrong they actually are. m

Actually, I'm a 48 year old, who has covered hockey professionally for 20 years, seen thousands of NHL, AHL, junior and European games and understands the game itself pretty damn well even though I'm your basic beer league 4th liner when I've attempted to play it.

When you look at the whole game, the Coyotes were trailing 3-0 for the majority of it. Also if you ask Tocch, whom I've known for years, he'd tell you a) that hockey games can be won or lost in a short segment and, just as important, b) the teams that win consistently know how to win many different ways (including in such games) whereas teams that don't win have endless rationales and excuses. A Corsi victory isn't a victory unless you put more pucks in the net. Shot quality matters more but, by the same token, "expected goals" aren't goals when you can't finish and a player with a high "points per 60 mins" doesn't necessarily warrant more ice time nor would he be likely to better help his team win if he got it.

Yes, there is a process and, yes, shot metrics are part of that process. But I watched that entire Stars-Coyotes game and Tanner's metrics argument was a joke. The Coyotes had their moments but, once the Stars got the bang-bang-bang outburst and Arizona was in chase-the-game mode, the Stars' control of the game never felt in much jeopardy regardless of what the metrics said.

Hockey has a math element, of course, but isn't math per se. To steal a quote from Mark Howe, "In hockey, if A is greater than B and B is greater than C, it doesn't always mean is A is greater than C." In math, that's the case 100 percent of the time. In hockey, Team A could sweep B in a season series and B could sweep C, yet C sweeps A.
- bmeltzer

Bill, all due respect, but we've all seen thousands of games, this just reads like every intermission on Sportsnet with Kypreos cracking jokes about the Corsi Cup. Probabilities are what we're talking about and if you replay this game a 1000 times, the Coyotes win most of the time.


There is no such thing as a "corsi" victory - it's simple probability. Who won? That's important at the end of the season, but what I'm concerned with on a game-to-game basis is who should have won. 'How is the team playing?' is far more relevant to their long term success than 'who won?"

I see what you're saying, I even understand your reservations, but the thing is, the guys who write for the Athletic, the guys who are like twitter hockey stats amateurs, and the guys on youtube videos talking at conferences have overcome these objections years ago.

The Arizona #Coyotes have assigned 2018 5th overall pick Barrett Hayton back to his Junior club in the OHL.

I crunched some numbers, Corsi'd my Fenwick and concluded through 2 games of the season, they have a shooting % of 0.

Now if we multiply by 5 and carry the Luck, my prediction is they'll get a goal at some point before 82 games are done

I crunched some numbers, Corsi'd my Fenwick and concluded through 2 games of the season, they have a shooting % of 0.

Now if we multiply by 5 and carry the Luck, my prediction is they'll get a goal at some point before 82 games are done
- ChonDerry

Not just a goal, but they will be a playoff team if they continue their trend of lots of shots and possession but no goals to show for it. Anyone who doesn't bother to actually watch the games, but simply look up the fancy stats can tell you the same. It's just science.

Bill, all due respect, but we've all seen thousands of games, this just reads like every intermission on Sportsnet with Kypreos cracking jokes about the Corsi Cup. Probabilities are what we're talking about and if you replay this game a 1000 times, the Coyotes win most of the time.


There is no such thing as a "corsi" victory - it's simple probability. Who won? That's important at the end of the season, but what I'm concerned with on a game-to-game basis is who should have won. 'How is the team playing?' is far more relevant to their long term success than 'who won?"

I see what you're saying, I even understand your reservations, but the thing is, the guys who write for the Athletic, the guys who are like twitter hockey stats amateurs, and the guys on youtube videos talking at conferences have overcome these objections years ago.
- James_Tanner

James you literally just started watching hockey this decade and have in no way seen thousands of games. You aren't even close to being in the same league as a guy like Mr. Meltzer with the experience he has. Watching highlights on Sports center, does not equal watching an entire game.

Actually, I'm a 48 year old, who has covered hockey professionally for 20 years, seen thousands of NHL, AHL, junior and European games and understands the game itself pretty damn well even though I'm your basic beer league 4th liner when I've attempted to play it.

When you look at the whole game, the Coyotes were trailing 3-0 for the majority of it. Also if you ask Tocch, whom I've known for years, he'd tell you a) that hockey games can be won or lost in a short segment and, just as important, b) the teams that win consistently know how to win many different ways (including in such games) whereas teams that don't win have endless rationales and excuses. A Corsi victory isn't a victory unless you put more pucks in the net. Shot quality matters more but, by the same token, "expected goals" aren't goals when you can't finish and a player with a high "points per 60 mins" doesn't necessarily warrant more ice time nor would he be likely to better help his team win if he got it.

Yes, there is a process and, yes, shot metrics are part of that process. But I watched that entire Stars-Coyotes game and Tanner's metrics argument was a joke. The Coyotes had their moments but, once the Stars got the bang-bang-bang outburst and Arizona was in chase-the-game mode, the Stars' control of the game never felt in much jeopardy regardless of what the metrics said.

Hockey has a math element, of course, but isn't math per se. To steal a quote from Mark Howe, "In hockey, if A is greater than B and B is greater than C, it doesn't always mean is A is greater than C." In math, that's the case 100 percent of the time. In hockey, Team A could sweep B in a season series and B could sweep C, yet C sweeps A.
- bmeltzer

I don’t disagree with anything you said (other than you saying Tanners analysis was bad). Nothing about your argument proves James or myself wrong. OF COURSE good teams find ways to win. OF COURSE a game can be won or lost in a moments notice. And maybe a team is particularly good at avoiding 2 minute meltdowns. But none of that disproves the actual argument we are having. In fact it doesn’t even speak to it, which is why this convo is frustrating.

The Coyotes out shot and out possessed their opponents two straight games. If that continues to be true over the course of the season they will have a high probability chance to be in the playoffs. Along the way they very well may lose games in a two minute interval despite winning the metrics. And they also might win games that they were by and large losing from a metric perspective.

BUT as long as their metrics look like they did these first two games, they absolutely will have a high probability to make the playoffs.

It’s hinestly not even an argument. It’s just basic math. There is no way to disprove it, no matter how many head coaches you know personally.

I don’t disagree with anything you said (other than you saying Tanners analysis was bad). Nothing about your argument proves James or myself wrong. OF COURSE good teams find ways to win. OF COURSE a game can be won or lost in a moments notice. And maybe a team is particularly good at avoiding 2 minute meltdowns. But none of that disproves the actual argument we are having. In fact it doesn’t even speak to it, which is why this convo is frustrating.

The Coyotes out shot and out possessed their opponents two straight games. If that continues to be true over the course of the season they will have a high probability chance to be in the playoffs. Along the way they very well may lose games in a two minute interval despite winning the metrics. And they also might win games that they were by and large losing from a metric perspective.

BUT as long as their metrics look like they did these first two games, they absolutely will have a high probability to make the playoffs.

It’s hinestly not even an argument. It’s just basic math. There is no way to disprove it, no matter how many head coaches you know personally.
- Dahlmanyotes

I don't think your equations take into account the nuisances of the games. By that I mean, if the Coyotes get scored on quickly and the other team sits back in a defensive posture, or rolls back their top players icetime, by in large not really caring if the Coyotes possess the puck more or continually shoot from non dangerous areas (because they lack talent) are they still in the best position to win?

I don't think your equations take into account the nuisances of the games. By that I mean, if the Coyotes get scored on quickly and the other team sits back in a defensive posture, or rolls back their top players icetime, by in large not really caring if the Coyotes possess the puck more or continually shoot from non dangerous areas (because they lack talent) are they still in the best position to win?
- LeftCoaster

Am I being punked? I feel like everyone is just messing with me to get me riled up!

Numbers absolutely take in to account nuance. Well technically they are actually overriding the nuances to develop a more accurate forecast. It is simply saying that regardless of the nuance, there is a broader truth.

For example: Person A and B are lifelong smokers. The statistics are clear...by being lifelong smokers they are lowering their probably lifespan. Now lets add nuance and say “well person A also eats fast food everyday, and person B eats healthy and goes to the gym everyday.” Of course that is interesting info and suggests that Person B will live longer than person A. BUT THAT NUANCE DOESNT CHANGE THE BROADER TRUTH THAT THEY ARE BOTH LOWERING THEIR LIFESPAN BY SMOKING.

Of course game situations are valuable nuances, and as big data explodes these will be introduced more and more to advanced hockey stats and will make forecasts more and more accurate. But these nuances will never (mathematically can not EVER) change or eliminate the broader truth about shots and possession correlating to wins.

I don't think your equations take into account the nuisances of the games. By that I mean, if the Coyotes get scored on quickly and the other team sits back in a defensive posture, or rolls back their top players icetime, by in large not really caring if the Coyotes possess the puck more or continually shoot from non dangerous areas (because they lack talent) are they still in the best position to win?
- LeftCoaster

See you're missing out on the fact that every game has anomalies. But they are just noise. Yes, once in a while a team will put 30 bad shots on net, make it look like they had possession and lose a game they deserved to lose. BUT the stats show that these things happen rarely, which is why large sample sizes are important.

To but it more simply, things like quality of linemates, quality of competition, shot quality and actual zone time matter more the smaller the sample size.

Once you have a large sample size, you can safely ignore this stuff.

You could of course adjust for it and try to get a better number (this is what WAR and GAR attempt to do) but if you're just concerned with general probabilities, shot attempts are very valuable.

And don't make the mistake of thinking that even 'stat guys' don't adjust for things. No one was calling Carolina the best team in the NHL even though they were the best possession team. We knew with Scott Darling in net and few true finishers on their roster, they'd not be the best team.

The problem is, is that when anti-stats people argue against stats, they usually do so while citing worse stats, like goals or wins, i think because they don't realize those are also statistics.

Bottom line, is that stats are just a predictive tool. People look at one wrong answer and think they don't work, but they forget that they're just coming up with probabilities.

Remember that in the most lopsided NHL match-up, the worst team vs the best team, the worst team still has a 40% chance to win. So actually making predictions in this salary cap world is nearly impossible. Pocket Twos can beat Pocket Aces every once in a while. You still want to get your money in with the best hand every time.

Will the Coyotes make the playoffs?

After reading all of this, you can just tell that Tanner doesn’t even watch the games.

Especially when discussing the Yotes first 2 games.

What I find hilarous is no one ever says "it was only one game and a score of 3-2 is a really small sample size" .........

If you really wanted to know who the best team was, you wouldn't even want to know the score because it only subverts your objectivity. I mean, if one game of possession stats is too short of a sample (which it is) then goals are up to ten times worse of a stat to go on.


- James_Tanner

If you're writing a blog and analyzing what happened in a game, then no, one game is not too short of a sample size to go on. It's the only think to go on. You're making a strawman argument here. We're looking at who was the best team in this game.

The point is that you do get information out of shot-attempts because you can build ten times the sample in one game as you can from goals. Sure there are things like shot quality etc. but the fact is that over a game, it's incredibly unlikely that you could significantly hold possession while not getting any high quality chances.


- James_Tanner

Of course you get information out of shot attempts for a game. You just don't know how to use it and interpret it properly.

And shot attempts don't factor in goaltending which is incredibly random. Last year, the Hawks, the Canes and the Flames would all have been excellent teams had they received even league average goaltending.


- James_Tanner

Again, making a strawman. We're talking about analyzing the Yotes/Stars game. You don't understand the nuances of the game. Nothing is random on the ice. To ignore and eliminate the role of the goaltender on the outcome of any game is ridiculously stupid. A goalie can be the team's best player in a game, make key saves at big times of a game, and carry a team to victory. Actually happens a lot. That's what you don't understand. It's making key plays in the game. Not all shots are created equal. Not all saves are. Hockey comes down to making plays at the right time. Goals matter, not shot attempts. One team may just be throwing a bunch of low percentage shots at the net hoping to get lucky while the other team is actually possessing the puck more and generating much better quality chances. Shot metrics aren't really possession.

But if you don't ignore goaltending when you evaluate the team you'll never ever get any usable conclusions. But we don't totally ignore goalies, we just average out their stats/ For example, we know that if a goalie is posting a save percentage well above league average, that it will regresses and vice versa.


- James_Tanner

Not to make a nitpicking argument but who is we? I don't think any analytics expert would want you included with them with how badly you misunderstand them. Again, a goaltender's play in a game is vital to a team's success. Averaging out a goalies stats is not analyzing a game. You can't watch a game without looking at stats and come to a conclusion because you don't know the game.

But the dumbest thing is that "analytics available to fans are not good." That is just incredibly wrong. I know you're basing it off one Bob McKenzie quote, but you might also want to realize that Bob has a long history of being against analytics and his network has lost so much, if not all, of their previous authority because of analytics.


- James_Tanner

It's not incredibly wrong, it's actually right. The data that analytic websites have to them is crap. I've posted on this many times to you and you've yet to been able to refute it. You insinuating that it's based on one quote is disingenuous on your part. That was just offered to support my statement that Corsi is archaic as a stat. It is. You refuse to accept it.

Let me ask you a basic question. If the public stats we fans have available to us are so good, why are NHL teams hiring a lot of the experts on analytics to work for them and paying them a salary? Why not just log onto the websites and get the info for free? Think about that and get back to me. LOL

I mean, just look at the work people like Tyler Dellow, Dom C, Sean Tierney, Ian Tulloch, Manny whose last name I forget and Steve Burch are doing with these "bad stats," anyways I hate that I get sucked in talking to people who will never take in new information or consider how wrong they actually are. m
- James_Tanner

The science of the stats isn't bad. The data is. Garbage in, garbage out.


One final thought. I'll just start with the basics here in an attempt to teach you something about shot metrics. I'm going to provide a link to the game and in that link you'll find a game flow chart. Notice what happens in the 2nd period after the 3 goal events scored by Dallas. Mr. Meltzer, who by the way is an expert NHL analyst was right. Your analysis of that game is a joke.

https://www.naturalstattr...eason=20182019&game=20012

Bill, all due respect, but we've all seen thousands of games, this just reads like every intermission on Sportsnet with Kypreos cracking jokes about the Corsi Cup. Probabilities are what we're talking about and if you replay this game a 1000 times, the Coyotes win most of the time.


There is no such thing as a "corsi" victory - it's simple probability. Who won? That's important at the end of the season, but what I'm concerned with on a game-to-game basis is who should have won. 'How is the team playing?' is far more relevant to their long term success than 'who won?"

I see what you're saying, I even understand your reservations, but the thing is, the guys who write for the Athletic, the guys who are like twitter hockey stats amateurs, and the guys on youtube videos talking at conferences have overcome these objections years ago.
- James_Tanner

The teams executive staff, the coaches, the players are all more concerned with who did win then who should've win. You can generate all the shot chances you want. You've got to put the puck in the net to win. I guarantee you that statistically, the team that doesn't score a goal in a hockey game, doesn't have a very good chance of winning statistically. I can predict that a team not scoring a goal will not win.

You didn't even get who should've won correct. You just looked at total shot metrics and determined that instead of looking at what actually happened. It's called score effects! How can you call yourself an expert and challenge people when you don't get this?

They haven't overcome the objections, they've just decided to ignore them so they don't discredit their product.

The problem is, is that when anti-stats people argue against stats, they usually do so while citing worse stats, like goals or wins, i think because they don't realize those are also statistics.

Bottom line, is that stats are just a predictive tool. People look at one wrong answer and think they don't work, but they forget that they're just coming up with probabilities.


- James_Tanner

You have it wrong. Most people that you label as anti stat aren't really that. What they're against is guys like you who don't know how to use them, who over use them, and who don't understand that goals are actually more important than shot metrics and all it's derivatives. There are actually people who are actually more concerned with goals scored, goals prevented and actual winning and losing. Actual results.

If you want to play around on a spreadsheet and try and predict outcomes, hey more power to you. But in reality, predicting is complete crap and meaningless. NHL teams use analytics to help them identify what happens out on the ice and to look for strengths and weaknesses to improve themselves. Even then it's only a small part.

As a blogger, if you find yourself analyzing a game using stats completely, you're doing it wrong. Unless of course you didn't watch the game and or have no clue what you're looking at. The latter I know to be true in your case.

If you're writing a blog and analyzing what happened in a game, then no, one game is not too short of a sample size to go on. It's the only think to go on. You're making a strawman argument here. We're looking at who was the best team in this game.



Of course you get information out of shot attempts for a game. You just don't know how to use it and interpret it properly.



Again, making a strawman. We're talking about analyzing the Yotes/Stars game. You don't understand the nuances of the game. Nothing is random on the ice. To ignore and eliminate the role of the goaltender on the outcome of any game is ridiculously stupid. A goalie can be the team's best player in a game, make key saves at big times of a game, and carry a team to victory. Actually happens a lot. That's what you don't understand. It's making key plays in the game. Not all shots are created equal. Not all saves are. Hockey comes down to making plays at the right time. Goals matter, not shot attempts. One team may just be throwing a bunch of low percentage shots at the net hoping to get lucky while the other team is actually possessing the puck more and generating much better quality chances. Shot metrics aren't really possession.



Not to make a nitpicking argument but who is we? I don't think any analytics expert would want you included with them with how badly you misunderstand them. Again, a goaltender's play in a game is vital to a team's success. Averaging out a goalies stats is not analyzing a game. You can't watch a game without looking at stats and come to a conclusion because you don't know the game.



It's not incredibly wrong, it's actually right. The data that analytic websites have to them is crap. I've posted on this many times to you and you've yet to been able to refute it. You insinuating that it's based on one quote is disingenuous on your part. That was just offered to support my statement that Corsi is archaic as a stat. It is. You refuse to accept it.

Let me ask you a basic question. If the public stats we fans have available to us are so good, why are NHL teams hiring a lot of the experts on analytics to work for them and paying them a salary? Why not just log onto the websites and get the info for free? Think about that and get back to me. LOL



The science of the stats isn't bad. The data is. Garbage in, garbage out.


One final thought. I'll just start with the basics here in an attempt to teach you something about shot metrics. I'm going to provide a link to the game and in that link you'll find a game flow chart. Notice what happens in the 2nd period after the 3 goal events scored by Dallas. Mr. Meltzer, who by the way is an expert NHL analyst was right. Your analysis of that game is a joke.

https://www.naturalstattr...eason=20182019&game=20012
- MJL

Imagine caring this much

You have it wrong. Most people that you label as anti stat aren't really that. What they're against is guys like you who don't know how to use them, who over use them, and who don't understand that goals are actually more important than shot metrics and all it's derivatives. There are actually people who are actually more concerned with goals scored, goals prevented and actual winning and losing. Actual results.

If you want to play around on a spreadsheet and try and predict outcomes, hey more power to you. But in reality, predicting is complete crap and meaningless. NHL teams use analytics to help them identify what happens out on the ice and to look for strengths and weaknesses to improve themselves. Even then it's only a small part.

As a blogger, if you find yourself analyzing a game using stats completely, you're doing it wrong. Unless of course you didn't watch the game and or have no clue what you're looking at. The latter I know to be true in your case.
- MJL

MJL you legitimately don’t know what your talking about, and even worse you have tricked yourself in to thinking you do. Here is a decade old article that is irrefutable and goes directly against some of your claims. It will start your lessons:

http://hockeyanalytics.co...laws-of-hockey-analytics/

After you’ve soaked it in you can come back. Commenting on your comments is just not worth the time.

MJL you legitimately don’t know what your talking about, and even worse you have tricked yourself in to thinking you do. Here is a decade old article that is irrefutable and goes directly against some of your claims. It will start your lessons:

http://hockeyanalytics.co...laws-of-hockey-analytics/

After you’ve soaked it in you can come back. Commenting on your comments is just not worth the time.
- Dahlmanyotes

Here is what you can do. You can come back and prove that I'm wrong. I'm ready when you are. Good luck to you but remember this was your choice. Let's go.

It's also interesting that the link you posted is pretty much most of what I've been saying for years. So that means that you're pretty much talking out your ass when you state that I don't know what I'm talking about. You need to reassess and get yourself together and then come back.

Imagine caring this much
- PhillySportsGuy

You cared enough to post this. I know it burns you!