Boston Bruins Advanced Stat Thread

[Gee Wally]

Please try to keep the deep dives and discussions in here.

(if you arent interested in this topic use ignore , thanks’

Click to read more...

 

[DominicT]

Since the trade proposal thread was hijacked by analytics, can I hijack this thread for trade talk?

j/k

 

[Gee Wally]

No

 

[Kegs]

No

Fun sucker

 

[Over the volcano]

Anyone seen an analysis of the effects zone starts have on shots for and against? On the face of it I’d assume getting 20% of your zone starts in the offensive end would be pretty limiting but I’m just guessing.

 

[DominicT]

As good as their goaltending numbers are, it's the Bruins' 5-on-5 defense that has the analytics community dumbstruck.

"Honestly, I suspect the public numbers are underrating them a bit defensively," Filipovic said.

He notes there are some disparities in "goals saved above expected" by Ullmark on sites such as Evolving Hockey (25.1), Sportlogiq (14.1) and Money Puck (19.2). Filipovic thinks the Bruins' goaltending probably nets out on the lower end of that scale, based on everything else the numbers tell us about their defense, including where they give up shots.

Micah Blake McCurdy of Hockey Viz creates shot heat maps for every team. The Bruins' show them as getting great goaltending, but also creating a great defensive environment for the netminders to thrive in.

"They basically choke off everything around the net and through the middle of the ice," Filipovic said, while wanting to take nothing away from Ullmark's season. "Even when you occasionally break through that shell and get into a good spot to shoot from, they're so quick to contest with active sticks in lanes that the quality of the shot itself has largely been neutralized. It's about as good of a defensive environment as you can get to play in as a goalie."

The Bruins are a hard team to score against. They're a team that can score at will. There's a reason they're a Stanley Cup favorite -- just ask their opponents. -

Breaking down how the Bruins have been so dominant -- and how long it can last

With Boston winning at a record pace, we look at how the team was built, what's behind its success, what foes are saying and whether it can take the Cup.

www.espn.com

 

[Gonzothe7thDman]

in only 5 games. Sadly expected goals don't equal real goals.

It does show how well they are playing though.

I'm making the # up but basically it says if that player plays 100 games like that, they will win their matchups the vast majority of the time.

But with the playoffs being a small sample size statistical anomalies are going to happen.

 

[Smitty93]

It does show how well they are playing though.

I'm making the # up but basically it says if that player plays 100 games like that, they will win their matchups the vast majority of the time.

But with the playoffs being a small sample size statistical anomalies are going to happen.

What's curious with Grzelcyk is that he's played 59 playoff games, where he's got a 54.5% xGF%, but just a 36.9% GF%. Compare that to Carlo and McAvoy who have a similar amount of games played, but have xGF%/GF% of 49.8%/47.1% and 54.0%/50.9%, respectively.

Now maybe the "he's small so he sucks in the playoffs" argument is correct, but you can't really confirm that without seeing how each goal was scored and his relevance to the play.

It's clear he's been unlucky. Both Carlo and McAvoy have PDO's just below 1 (0.987 and 0.992, respectivel), but Grzelcyk's is 0.950. It seems impossible for that kind of terrible luck to continue.

 

[LouJersey]

It does show how well they are playing though.

I'm making the # up but basically it says if that player plays 100 games like that, they will win their matchups the vast majority of the time.

But with the playoffs being a small sample size statistical anomalies are going to happen.

What's curious with Grzelcyk is that he's played 59 playoff games, where he's got a 54.5% xGF%, but just a 36.9% GF%. Compare that to Carlo and McAvoy who have a similar amount of games played, but have xGF%/GF% of 49.8%/47.1% and 54.0%/50.9%, respectively.

Now maybe the "he's small so he sucks in the playoffs" argument is correct, but you can't really confirm that without seeing how each goal was scored and his relevance to the play.

It's clear he's been unlucky. Both Carlo and McAvoy have PDO's just below 1 (0.987 and 0.992, respectivel), but Grzelcyk's is 0.950. It seems impossible for that kind of terrible luck to continue.

NHL Stats

The official source for NHL Stats including skaters, goalies, teams stats and more.

www.nhl.com

12 years worth.

 

[MarchysNoseKnows]

NHL Stats

The official source for NHL Stats including skaters, goalies, teams stats and more.

www.nhl.com

12 years worth.

Lou - the guys above are trying to explain the ridiculous anomaly you’re posting about. Could you try to give them a shot and listen to them?

 

[LouJersey]

Lou - the guys above are trying to explain the ridiculous anomaly you’re posting about. Could you try to give them a shot and listen to them?

I'm not saying it isn't, what I am saying is the numbers are the numbers for both and the truth is somewhere in the middle...I think it's some bad luck and he is at a physical disadvantage sometimes in the play-offs. It can be both is all I have ever tried to say.

 

[MarchysNoseKnows]

I'm not saying it isn't, what I am saying is the numbers are the numbers for both and the truth is somewhere in the middle...I think it's some bad luck and he is at a physical disadvantage sometimes in the play-offs. It can be both is all I have ever tried to say.

But in the advanced stat thread to answer two well thought out and reasonable responses with another post of the plus/minus stat is kind of off base no?

One defender turns the puck over right to a guy in the slot wide open. Goalie makes a ridiculous save.

Other defender defends his ass off, forces an unscreened muffin from the point and the goalie lets it in somehow.

Which defender played better on those plays? Plus minus says the former. xGF says the latter.

 

[Smitty93]

But in the advanced stat thread to answer two well thought out and reasonable responses with another post of the plus/minus stat is kind of off base no?

One defender turns the puck over right to a guy in the slot wide open. Goalie makes a ridiculous save.

Other defender defends his ass off, forces an unscreened muffin from the point and the goalie lets it in somehow.

Which defender played better on those plays? Plus minus says the former. xGF says the latter.

No, no, it's okay. Because I think that data can speak to the issue. When I look at it, the GA looks bad, but the GF is a lot more noticeable. That makes sense, because under Cassidy in the playoffs (5 on 5), the Bruins scored only 90% of their expected goals for, but gave up 106% of their expected goals against. Let's compare that to their regular season 5 on 5 play, where they scored 101% of their expected goals for, and gave up 101% of their expected goals against. (Wow, turns out large sample sizes make the data even out, huh? Who would have guessed?)

Anyway, there's two ways to look at it, either playoffs are a small sample size, or Cassidy was a bad playoff coach (honestly, not going to get an argument against the latter from me).

So I think the thing is, you've got two options. You can say that Grzelcyk just can't play the style the playoffs require, and bench/replace him. OR, you can do the smarter thing, which is review the tape and data, and then figure out what adjustments can be made so that your 3rd best defenseman (which he absolutely is) can play as well as he can in the regular season.

Because that's really the greatest value of data. You interpret it in order to optimize future decisions. That's the case in every industry, including professional sports. And I refuse to believe that Grzelcyk is a lost cause in the playoffs. He has every skill necessary to provide offensive value, and I have a firm belief that the only part of playing defense that isn't fixable is having poor hockey IQ, and I have zero concerns about Grzelcyk there.

***all data above taken from naturalstattrick.com***

 

[GatorMike]

For what it's worth when it comes to Grzelcyk's career playoff numbers...

xGF: 36.42
GF: 24
Difference: -12.42

xGA: 30.36
GA: 41
Difference: -10.64

On-Ice Shooting Percentage: 5.05%
On-Ice Save Percentage: 90.0%

So, when Grzelcyk is on the ice 5v5 in the playoffs, the Bruins tend have better/more scoring chances, but they convert them at an exceptionally low rate, and the other team converts their chances at a really high rate. I have a hard time blaming Grzelcyk for the team's inability to finish their chances, but it's certainly plausible that he shares a great deal of responsibility for the fact that they haven't been able to keep the puck out of their net.

But if Gryz is the root of all that ails the Bruins come playoff season, then you'd expect the number of scoring chances and high danger scoring chances increase dramatically as well, right? Here's a look at his defensive metrics over the past three seasons... (Natural Stat Trick only allows you to view three years at a time, and I'm far too lazy to add this all up.)

Matt Grzelcyk Playoffs (2020-22)
Scoring Chances Allowed/60: 25.43
High Danger Scoring Chances Allowed/60: 9.13
High Danger Goals Allowed/60: 1.83
Expected Goals Allowed/60: 2.07
On-Ice Save Percentage: 88.3%
Goals Allowed/60: 3.13

Matt Grzelcyk Regular Season (2020-22)
Scoring Chances Allowed/60: 21.72
High Danger Scoring Chances Allowed/60: 7.76
High Danger Goals Allowed/60: 0.94
Expected Goals Allowed/60: 1.80
On-Ice Save Percentage: 93.3%
Goals Allowed/60: 1.72

So, yeah... Grzelcyk isn't quite as effective defensively in the playoffs. And you shouldn't expect him to be. By definition, the quality of opposition will improve as the Bruins aren't playing the likes of Buffalo and Ottawa once they reach the post-season. What stands out to me is that while the rate of scoring chances and xGA goes up slightly, the team's save percentage and goals allowed craters. That indicates to me that for whatever reason, a lot of shots have gone in that shouldn't have when Grzelcyk's been on the ice.

By comparison, here are Charlie McAvoy's numbers, again over the past three seasons:

Charlie McAvoy Playoffs (2020-22)
Scoring Chances Allowed/60: 28.15
High Danger Scoring Chances Allowed/60: 9.91
High Danger Goals Allowed/60: 1.26
Expected Goals Allowed/60: 2.27
On-Ice Save Percentage: 91.4%
Goals Allowed/60: 2.53

Charlie McAvoy Regular Season (2020-22)
Scoring Chances Allowed/60: 22.81
High Danger Scoring Chances Allowed/60: 8.44
High Danger Goals Allowed/60: 1.1
Expected Goals Allowed/60: 1.92
On-Ice Save Percentage: 92.2%
Goals Allowed/60: 2.04

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

 

[Gonzothe7thDman]

For what it's worth when it comes to Grzelcyk's career playoff numbers...

xGF: 36.42
GF: 24
Difference: -12.42

xGA: 30.36
GA: 41
Difference: -10.64

On-Ice Shooting Percentage: 5.05%
On-Ice Save Percentage: 90.0%

So, when Grzelcyk is on the ice 5v5 in the playoffs, the Bruins tend have better/more scoring chances, but they convert them at an exceptionally low rate, and the other team converts their chances at a really high rate. I have a hard time blaming Grzelcyk for the team's inability to finish their chances, but it's certainly plausible that he shares a great deal of responsibility for the fact that they haven't been able to keep the puck out of their net.

But if Gryz is the root of all that ails the Bruins come playoff season, then you'd expect the number of scoring chances and high danger scoring chances increase dramatically as well, right? Here's a look at his defensive metrics over the past three seasons... (Natural Stat Trick only allows you to view three years at a time, and I'm far too lazy to add this all up.)

Matt Grzelcyk Playoffs (2020-22)
Scoring Chances Allowed/60: 25.43
High Danger Scoring Chances Allowed/60: 9.13
High Danger Goals Allowed/60: 1.83
Expected Goals Allowed/60: 2.07
On-Ice Save Percentage: 88.3%
Goals Allowed/60: 3.13

Matt Grzelcyk Regular Season (2020-22)
Scoring Chances Allowed/60: 21.72
High Danger Scoring Chances Allowed/60: 7.76
High Danger Goals Allowed/60: 0.94
Expected Goals Allowed/60: 1.80
On-Ice Save Percentage: 93.3%
Goals Allowed/60: 1.72

So, yeah... Grzelcyk isn't quite as effective defensively in the playoffs. And you shouldn't expect him to be. By definition, the quality of opposition will improve as the Bruins aren't playing the likes of Buffalo and Ottawa once they reach the post-season. What stands out to me is that while the rate of scoring chances and xGA goes up slightly, the team's save percentage and goals allowed craters. That indicates to me that for whatever reason, a lot of shots have gone in that shouldn't have when Grzelcyk's been on the ice.

By comparison, here are Charlie McAvoy's numbers, again over the past three seasons:

Charlie McAvoy Playoffs (2020-22)
Scoring Chances Allowed/60: 28.15
High Danger Scoring Chances Allowed/60: 9.91
High Danger Goals Allowed/60: 1.26
Expected Goals Allowed/60: 2.27
On-Ice Save Percentage: 91.4%
Goals Allowed/60: 2.53

Charlie McAvoy Regular Season (2020-22)
Scoring Chances Allowed/60: 22.81
High Danger Scoring Chances Allowed/60: 8.44
High Danger Goals Allowed/60: 1.1
Expected Goals Allowed/60: 1.92
On-Ice Save Percentage: 92.2%
Goals Allowed/60: 2.04

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

Thanks for putting in the work I was too lazy to do.

Great stuff here and great breakdown.

 

[MarchysNoseKnows]

For what it's worth when it comes to Grzelcyk's career playoff numbers...

xGF: 36.42
GF: 24
Difference: -12.42

xGA: 30.36
GA: 41
Difference: -10.64

On-Ice Shooting Percentage: 5.05%
On-Ice Save Percentage: 90.0%

So, when Grzelcyk is on the ice 5v5 in the playoffs, the Bruins tend have better/more scoring chances, but they convert them at an exceptionally low rate, and the other team converts their chances at a really high rate. I have a hard time blaming Grzelcyk for the team's inability to finish their chances, but it's certainly plausible that he shares a great deal of responsibility for the fact that they haven't been able to keep the puck out of their net.

But if Gryz is the root of all that ails the Bruins come playoff season, then you'd expect the number of scoring chances and high danger scoring chances increase dramatically as well, right? Here's a look at his defensive metrics over the past three seasons... (Natural Stat Trick only allows you to view three years at a time, and I'm far too lazy to add this all up.)

Matt Grzelcyk Playoffs (2020-22)
Scoring Chances Allowed/60: 25.43
High Danger Scoring Chances Allowed/60: 9.13
High Danger Goals Allowed/60: 1.83
Expected Goals Allowed/60: 2.07
On-Ice Save Percentage: 88.3%
Goals Allowed/60: 3.13

Matt Grzelcyk Regular Season (2020-22)
Scoring Chances Allowed/60: 21.72
High Danger Scoring Chances Allowed/60: 7.76
High Danger Goals Allowed/60: 0.94
Expected Goals Allowed/60: 1.80
On-Ice Save Percentage: 93.3%
Goals Allowed/60: 1.72

So, yeah... Grzelcyk isn't quite as effective defensively in the playoffs. And you shouldn't expect him to be. By definition, the quality of opposition will improve as the Bruins aren't playing the likes of Buffalo and Ottawa once they reach the post-season. What stands out to me is that while the rate of scoring chances and xGA goes up slightly, the team's save percentage and goals allowed craters. That indicates to me that for whatever reason, a lot of shots have gone in that shouldn't have when Grzelcyk's been on the ice.

By comparison, here are Charlie McAvoy's numbers, again over the past three seasons:

Charlie McAvoy Playoffs (2020-22)
Scoring Chances Allowed/60: 28.15
High Danger Scoring Chances Allowed/60: 9.91
High Danger Goals Allowed/60: 1.26
Expected Goals Allowed/60: 2.27
On-Ice Save Percentage: 91.4%
Goals Allowed/60: 2.53

Charlie McAvoy Regular Season (2020-22)
Scoring Chances Allowed/60: 22.81
High Danger Scoring Chances Allowed/60: 8.44
High Danger Goals Allowed/60: 1.1
Expected Goals Allowed/60: 1.92
On-Ice Save Percentage: 92.2%
Goals Allowed/60: 2.04

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

I’d like to buy this post a beer.

The funny part about a thread called “Advanced Stats Thread” is it won’t have any “eye test” people posting it or for the most part even reading it.

 

[RussellmaniaKW]

Please try to keep the deep dives and discussions in here.

(if you arent interested in this topic use ignore , thanks’

it's funny... I can think of several regular posters off the top of my head who certainly seem interested in this topic because any time you try to use even pretty widely adopted stats to argue for a player's value they are quick to completely disregard your argument and instead tell you how stupid your stats or "charts" are. Usually the response is a reasoned explanation of what the stats mean or why the information being conveyed is valuable.

These reasoned arguments are often met with wilful ignorance or outright disdain, but most often just a general unwillingness to even try to understand what these stats mean. After this back & forth a mod (usually you) will shuffle us off to a thread like this at which point the only people (with some exceptions) who show up to have a meaningful discussion are the people who already agree on the value of analytics. Meanwhile the people who started the shitstorm with their dismissive initial comments never bother to join this thread and engage in real discussion. rinse & repeat.

The point is that these arguments don't happen because guys like @MarchysNoseKnows are just dying to discuss analytics. They happen because they simply try to use the information available to them in a debate about players and someone else comes along and tells them how worthless that information is. how is someone even supposed to respond to that if they aren't allowed to defend their own hockey worldview?

So I guess my questions are these:
1. Why does it seem like people have free reign to make derisive comments about how stupid "fancy" stats are but as soon as someone tries to explain why they aren't, that's when the mods crack down?
2. do you think it's good for the board to isolate serious, thoughtful discussion about evaluating hockey players into its own thread that mostly gets visited by people who already agree?
3. the thread that usually gets "derailed" by these debates is the ongoing trade/rumors/speculation thread. Do you really think the discourse that is typically happening in that thread is worth preserving at the expense of the types of discussion you see just on the first page of this thread?
4. what exactly is the point of this thread? Is it a thread to discuss analytics as a concept? Or is it just the only thread where people are allowed to use analytics in discussions about players?

 

[BlackFrancis]

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

I've said it before, but the disparity between Grzelcyk's expected vs. actual has to be a combo of factors.

• One of those playoffs had that series against the Isles where Rask was friggin' crippled. Not sure how many Grzelcyk was on for, but defending with a statue behind you leaves you pretty much at the mercy of luck.
• Speaking of luck, 60 games isn't exactly small, but it's still in the range where someone could shoot 25% and no one would bat an eye - just wait out the eventual regression to the mean. As MarchyNose has been saying repeatedly, this stuff will probably normalize.
• That Carolina series last year, his xGF was fine, but he was struggling. Just could not get the puck out with the bum shoulder. If 10% of the game is getting the puck out and the shoulder isn't bother you as much the other 90%, I'd imagine the data isn't going to take a giant shit on your head. But if you can't get the puck out and your game isn't normally challenged in this area, the scrambling and frustration is surely following.
• 17% of his playoff games came in his second season, where he was okay but not great. Bruce also flipped his zone starts in the playoffs for some unfathomable reason.
• Someone already brought it up I think but Bruce's teams just farted out 5 on 5, especially in the playoffs. And Grzelcyk only plays 5 on 5. Not easy to get a check mark in the black column when the team is going balls out to win 0-0 until they get a PP attempt, which were more infrequent by the year, as refs and God hated Cassidy.

I'm sure there's more but I'm sleepy. And I'm being a dink, putting up mostly anecdote in a stats thread.

 

[BourqueFromORRk4277]

For what it's worth when it comes to Grzelcyk's career playoff numbers...

xGF: 36.42
GF: 24
Difference: -12.42

xGA: 30.36
GA: 41
Difference: -10.64

On-Ice Shooting Percentage: 5.05%
On-Ice Save Percentage: 90.0%

So, when Grzelcyk is on the ice 5v5 in the playoffs, the Bruins tend have better/more scoring chances, but they convert them at an exceptionally low rate, and the other team converts their chances at a really high rate. I have a hard time blaming Grzelcyk for the team's inability to finish their chances, but it's certainly plausible that he shares a great deal of responsibility for the fact that they haven't been able to keep the puck out of their net.

But if Gryz is the root of all that ails the Bruins come playoff season, then you'd expect the number of scoring chances and high danger scoring chances increase dramatically as well, right? Here's a look at his defensive metrics over the past three seasons... (Natural Stat Trick only allows you to view three years at a time, and I'm far too lazy to add this all up.)

Matt Grzelcyk Playoffs (2020-22)
Scoring Chances Allowed/60: 25.43
High Danger Scoring Chances Allowed/60: 9.13
High Danger Goals Allowed/60: 1.83
Expected Goals Allowed/60: 2.07
On-Ice Save Percentage: 88.3%
Goals Allowed/60: 3.13

Matt Grzelcyk Regular Season (2020-22)
Scoring Chances Allowed/60: 21.72
High Danger Scoring Chances Allowed/60: 7.76
High Danger Goals Allowed/60: 0.94
Expected Goals Allowed/60: 1.80
On-Ice Save Percentage: 93.3%
Goals Allowed/60: 1.72

So, yeah... Grzelcyk isn't quite as effective defensively in the playoffs. And you shouldn't expect him to be. By definition, the quality of opposition will improve as the Bruins aren't playing the likes of Buffalo and Ottawa once they reach the post-season. What stands out to me is that while the rate of scoring chances and xGA goes up slightly, the team's save percentage and goals allowed craters. That indicates to me that for whatever reason, a lot of shots have gone in that shouldn't have when Grzelcyk's been on the ice.

By comparison, here are Charlie McAvoy's numbers, again over the past three seasons:

Charlie McAvoy Playoffs (2020-22)
Scoring Chances Allowed/60: 28.15
High Danger Scoring Chances Allowed/60: 9.91
High Danger Goals Allowed/60: 1.26
Expected Goals Allowed/60: 2.27
On-Ice Save Percentage: 91.4%
Goals Allowed/60: 2.53

Charlie McAvoy Regular Season (2020-22)
Scoring Chances Allowed/60: 22.81
High Danger Scoring Chances Allowed/60: 8.44
High Danger Goals Allowed/60: 1.1
Expected Goals Allowed/60: 1.92
On-Ice Save Percentage: 92.2%
Goals Allowed/60: 2.04

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

If "Bad Luck" also includes brain farts, turnovers, and defensive lapses by a teammate while you just happen to be on the ice with them, then this agrees with my eye test.

 

[LouJersey]

No, no, it's okay. Because I think that data can speak to the issue. When I look at it, the GA looks bad, but the GF is a lot more noticeable. That makes sense, because under Cassidy in the playoffs (5 on 5), the Bruins scored only 90% of their expected goals for, but gave up 106% of their expected goals against. Let's compare that to their regular season 5 on 5 play, where they scored 101% of their expected goals for, and gave up 101% of their expected goals against. (Wow, turns out large sample sizes make the data even out, huh? Who would have guessed?)

Anyway, there's two ways to look at it, either playoffs are a small sample size, or Cassidy was a bad playoff coach (honestly, not going to get an argument against the latter from me).

So I think the thing is, you've got two options. You can say that Grzelcyk just can't play the style the playoffs require, and bench/replace him. OR, you can do the smarter thing, which is review the tape and data, and then figure out what adjustments can be made so that your 3rd best defenseman (which he absolutely is) can play as well as he can in the regular season.

Because that's really the greatest value of data. You interpret it in order to optimize future decisions. That's the case in every industry, including professional sports. And I refuse to believe that Grzelcyk is a lost cause in the playoffs. He has every skill necessary to provide offensive value, and I have a firm belief that the only part of playing defense that isn't fixable is having poor hockey IQ, and I have zero concerns about Grzelcyk there.

***all data above taken from naturalstattrick.com**

But in the advanced stat thread to answer two well thought out and reasonable responses with another post of the plus/minus stat is kind of off base no?

One defender turns the puck over right to a guy in the slot wide open. Goalie makes a ridiculous save.

Other defender defends his ass off, forces an unscreened muffin from the point and the goalie lets it in somehow.

Which defender played better on those plays? Plus minus says the former. xGF says the latter.

I think the only difference of opinion we have is that you just want to totally discount his entire gf/ga thing as poor luck. I don’t see it that way and I’m guessing the bruins don’t either

As for the goals against looking bad it is. You’re looking at something expected and not what really happened. Why are the bruins interested in Chychrun or Gavrikov if they have to his guy already in house?

I’m thinking it all plays out better w Monty so hopefully the numbers line up

 

[LouJersey]

For what it's worth when it comes to Grzelcyk's career playoff numbers...

xGF: 36.42
GF: 24
Difference: -12.42

xGA: 30.36
GA: 41
Difference: -10.64

On-Ice Shooting Percentage: 5.05%
On-Ice Save Percentage: 90.0%

So, when Grzelcyk is on the ice 5v5 in the playoffs, the Bruins tend have better/more scoring chances, but they convert them at an exceptionally low rate, and the other team converts their chances at a really high rate. I have a hard time blaming Grzelcyk for the team's inability to finish their chances, but it's certainly plausible that he shares a great deal of responsibility for the fact that they haven't been able to keep the puck out of their net.

But if Gryz is the root of all that ails the Bruins come playoff season, then you'd expect the number of scoring chances and high danger scoring chances increase dramatically as well, right? Here's a look at his defensive metrics over the past three seasons... (Natural Stat Trick only allows you to view three years at a time, and I'm far too lazy to add this all up.)

Matt Grzelcyk Playoffs (2020-22)
Scoring Chances Allowed/60: 25.43
High Danger Scoring Chances Allowed/60: 9.13
High Danger Goals Allowed/60: 1.83
Expected Goals Allowed/60: 2.07
On-Ice Save Percentage: 88.3%
Goals Allowed/60: 3.13

Matt Grzelcyk Regular Season (2020-22)
Scoring Chances Allowed/60: 21.72
High Danger Scoring Chances Allowed/60: 7.76
High Danger Goals Allowed/60: 0.94
Expected Goals Allowed/60: 1.80
On-Ice Save Percentage: 93.3%
Goals Allowed/60: 1.72

So, yeah... Grzelcyk isn't quite as effective defensively in the playoffs. And you shouldn't expect him to be. By definition, the quality of opposition will improve as the Bruins aren't playing the likes of Buffalo and Ottawa once they reach the post-season. What stands out to me is that while the rate of scoring chances and xGA goes up slightly, the team's save percentage and goals allowed craters. That indicates to me that for whatever reason, a lot of shots have gone in that shouldn't have when Grzelcyk's been on the ice.

By comparison, here are Charlie McAvoy's numbers, again over the past three seasons:

Charlie McAvoy Playoffs (2020-22)
Scoring Chances Allowed/60: 28.15
High Danger Scoring Chances Allowed/60: 9.91
High Danger Goals Allowed/60: 1.26
Expected Goals Allowed/60: 2.27
On-Ice Save Percentage: 91.4%
Goals Allowed/60: 2.53

Charlie McAvoy Regular Season (2020-22)
Scoring Chances Allowed/60: 22.81
High Danger Scoring Chances Allowed/60: 8.44
High Danger Goals Allowed/60: 1.1
Expected Goals Allowed/60: 1.92
On-Ice Save Percentage: 92.2%
Goals Allowed/60: 2.04

As expected - McAvoy's defensive metrics also aren't as good in the playoffs. But you don't see the huge delta between Goals Allowed and Expected Goals Allowed. Over the past three Regular Seasons, Grzelcyk's defensive metrics are better than McAvoy's across the board. (Insert quality of competition argument) In the last three playoff runs, Grzelcyk's defensive shot quality metrics are still better than McAvoy's, but the results are much worse. I'm open to other suggestions, but at this point, bad luck/goalie play makes the most sense to me.

Fantastic post

 

[DKH]

Remember folks Cassidy and Monty different style in defensive zone

Ideally he’s used likeJames White and not featured

 

[LouJersey]

Remember folks Cassidy and Monty different style in defensive zone

Ideally he’s used likeJames White and not featured

Agree 100pct.

 

[GatorMike]

I've said it before, but the disparity between Grzelcyk's expected vs. actual has to be a combo of factors.

• One of those playoffs had that series against the Isles where Rask was friggin' crippled. Not sure how many Grzelcyk was on for, but defending with a statue behind you leaves you pretty much at the mercy of luck.
• Speaking of luck, 60 games isn't exactly small, but it's still in the range where someone could shoot 25% and no one would bat an eye - just wait out the eventual regression to the mean. As MarchyNose has been saying repeatedly, this stuff will probably normalize.
• That Carolina series last year, his xGF was fine, but he was struggling. Just could not get the puck out with the bum shoulder. If 10% of the game is getting the puck out and the shoulder isn't bother you as much the other 90%, I'd imagine the data isn't going to take a giant shit on your head. But if you can't get the puck out and your game isn't normally challenged in this area, the scrambling and frustration is surely following.
• 17% of his playoff games came in his second season, where he was okay but not great. Bruce also flipped his zone starts in the playoffs for some unfathomable reason.
• Someone already brought it up I think but Bruce's teams just farted out 5 on 5, especially in the playoffs. And Grzelcyk only plays 5 on 5. Not easy to get a check mark in the black column when the team is going balls out to win 0-0 until they get a PP attempt, which were more infrequent by the year, as refs and God hated Cassidy.

I'm sure there's more but I'm sleepy. And I'm being a dink, putting up mostly anecdote in a stats thread.

I think in some years, injuries have forced him into matchups that the Bruins would ideally like to keep him away from.

I also remember Jaro Halak crapping the bed in the 2nd round against Tampa one year.

 

[Gee Wally]

it's funny... I can think of several regular posters off the top of my head who certainly seem interested in this topic because any time you try to use even pretty widely adopted stats to argue for a player's value they are quick to completely disregard your argument and instead tell you how stupid your stats or "charts" are. Usually the response is a reasoned explanation of what the stats mean or why the information being conveyed is valuable.

These reasoned arguments are often met with wilful ignorance or outright disdain, but most often just a general unwillingness to even try to understand what these stats mean. After this back & forth a mod (usually you) will shuffle us off to a thread like this at which point the only people (with some exceptions) who show up to have a meaningful discussion are the people who already agree on the value of analytics. Meanwhile the people who started the shitstorm with their dismissive initial comments never bother to join this thread and engage in real discussion. rinse & repeat.

The point is that these arguments don't happen because guys like @MarchysNoseKnows are just dying to discuss analytics. They happen because they simply try to use the information available to them in a debate about players and someone else comes along and tells them how worthless that information is. how is someone even supposed to respond to that if they aren't allowed to defend their own hockey worldview?

So I guess my questions are these:
1. Why does it seem like people have free reign to make derisive comments about how stupid "fancy" stats are but as soon as someone tries to explain why they aren't, that's when the mods crack down?
2. do you think it's good for the board to isolate serious, thoughtful discussion about evaluating hockey players into its own thread that mostly gets visited by people who already agree?
3. the thread that usually gets "derailed" by these debates is the ongoing trade/rumors/speculation thread. Do you really think the discourse that is typically happening in that thread is worth preserving at the expense of the types of discussion you see just on the first page of this thread?
4. what exactly is the point of this thread? Is it a thread to discuss analytics as a concept? Or is it just the only thread where people are allowed to use analytics in discussions about players?

People typically are interested in the subject in thread titles. People have a variety of interests . Many dislike when those things get hijacked. The site tries to cater to as much as it can.

So much so it was realized many years ago this particular subject was a hot button issue for both ‘sides’. that HF devoted an entire forum for those that like to drill down into the numbers.

By The Numbers

Hockey Analytics... the Final Frontier. Explore strange new worlds, to seek out new algorithms, to boldly go where no one has gone before.

forums.hfboards.com

This board is doing no different. Providing alternatives.

Now, I believe you already knew the answer and know why you are doing this.
So having said that if you have any other issues with site moderation please do as you agreed to upon becoming a member and email the site via the *contact us* button.
Thanks.