Speculation: Justin Holl Discussion Thread

[Buds17]

All depends on what you mean by top pairing - in terms of matchups it's always been the Muzz-Holl (and now brodie-Holl) that gets the top line matchups and the heavy dzone starts.

True. I should have mentioned that TOI and ES pairings aren't necessarily going to be one and the same. I suppose Keefe could opt to ease Muzzin back into the lineup on the third pairing. For a variety of reasons though, a Rielly-Brodie-Muzzin left side D doesn't sound like it could work out well for long.

 

[tmlfan98]

True. I suppose Keefe could opt to ease Muzzin back into the lineup on the third pairing. For a variety of reasons though, a Rielly-Brodie-Muzzin left side D doesn't sound like it could work out for long.

Tampa has been running a left side of Hedman-McDonagh-Sergachev for years.

 

[Zybalto]

Top ten NHL dmen for XGA/60 REL since January 1st (min 400+ 5v5 minutes and 17+ 5v5 minutes a night, 73 dmen total) :
1. Morrisey
2. Doughty
3. Holl
4. Matheson
5. Siegenthaler
6. Orlov
7. Braun
8. Provorov
9. Anderson
10. Ceci

Top ten NHL dmen for GA/60 REL since January 1st: (min 400+ 5v5 minutes and 17+ 5v5 minutes a night, 73 dmen total) :
1. Ceci
2. Holl
3. Heskenin
4. Dobson
5. Maatta
6. Suter
7. Fowler
8. Dahlin
9. Matheson
10. Jones

Looking at usage, Holl is rocking team D MVP numbers for both expected and actual results this calendar year. Ceci making the board on a 2nd straight team confused as they had heard he was bad but looks good most nights (just like he was here). Holl is the same type of guy as he has a week or two of great play but as soon as he has a questionable game in super tough usage, the folks in the peanut gallery start saying "SEE!!, I TOLD YOU HE WAS BAD!!!".

 

[tmlfan98]

Top ten NHL dmen for XGA/60 REL since January 1st (min 400+ 5v5 minutes and 17+ 5v5 minutes a night, 73 dmen total) :
1. Morrisey
2. Doughty
3. Holl
4. Matheson
5. Siegenthaler
6. Orlov
7. Braun
8. Provorov
9. Anderson
10. Ceci

Top ten NHL dmen for GA/60 REL since January 1st: (min 400+ 5v5 minutes and 17+ 5v5 minutes a night, 73 dmen total) :
1. Ceci
2. Holl
3. Heskenin
4. Dobson
5. Maatta
6. Suter
7. Fowler
8. Dahlin
9. Matheson
10. Jones

Looking at usage, Holl is rocking team D MVP numbers for both expected and actual results this calendar year. Ceci making the board on a 2nd straight team confused as they had heard he was bad but looks good most nights (just like he was here). Holl is the same type of guy as he has a week or two of great play but as soon as he has a questionable game in super tough usage, the folks in the peanut gallery start saying "SEE!!, I TOLD YOU HE WAS BAD!!!".

If you factor in HDGA/60 Rel and HDGF% Rel with the same minutes per game filters (over a full season sample though) he's top 10 in both as well:

HDGA/60 Rel:
1. Artem Zub
2. Matt Roy
3. Darnell Nurse
4. Miro Heiskanen
5. Zach Whitecloud
6. Justin Holl
7. Ivan Provorov
8. Thomas Chabot
9. Anton Stralman
10. Adam Pelech

HDGF% Rel:
1. Artem Zub
2. Jamie Oleksiak
3. Brent Burns
4. Jaccob Slavin
5. Matt Roy
6. Zach Whitecloud
7. Dante Fabbro
8. Anton Stralman
9. Justin Holl
10. Charlie McAvoy

 

[zeke]

Tampa has been running a left side of Hedman-McDonagh-Sergachev for years.

Each having a very similar deployment as Rielly-Muzzin-Sandin get.

 

[tmlfan98]

Each having a very similar deployment as Rielly-Muzzin-Sandin get.

Where does the Brodie-Holl pair stand if Rielly-Muzzin-Sandin is your left side?

 

[IPS]

If you factor in HDGA/60 Rel and HDGF% Rel with the same minutes filters he's top 10 in both as well:

HDGA/60 Rel:
1. Artem Zub
2. Matt Roy
3. Darnell Nurse
4. Miro Heiskanen
5. Zach Whitecloud
6. Justin Holl
7. Ivan Provorov
8. Thomas Chabot
9. Anton Stralman
10. Adam Pelech

HDGF% Rel:
1. Artem Zub
2. Jamie Oleksiak
3. Brent Burns
4. Jaccob Slavin
5. Matt Roy
6. Zach Whitecloud
7. Dante Fabbro
8. Anton Stralman
9. Justin Holl
10. Charlie McAvoy

Atleast those ones can actually find a couple of defensemen who are legit the best defensively in the league in Slavin, Pelech and Mcavoy.

If Cody Ceci is ranking top-10 in your stats that are supposed to reflect defensive proficiency, you should probably take a second look.

 

[tmlfan98]

Atleast those ones can actually find a couple of defensemen who are legit the best defensively in the league in Slavin, Pelech and Mcavoy.

If Cody Ceci is ranking top-10 in your stats that are supposed to reflect defensive proficiency, you should probably take a second look.

High danger metrics are the most predictive when it comes to playoff success and it's not even close. Slavin, Pelech and McAvoy are playoff type players and their high danger metrics prove it.

 

[IPS]

High danger metrics are the most predictive when it comes to playoff success and it's not even close. Slavin, Pelech and McAvoy are playoff type players and their high danger metrics prove it.

So where do any of Tampa's D fit in on this picture the last couple years?

 

[tmlfan98]

So where do any of Tampa's D fit in on this picture the last couple years?

As a team Tampa has always rated very highly in these high danger defensive metrics.

2019-20:
7.27 HDSA/60 (5th)
1.13 HDGA/60 (5th)
58.94 HDGF% (2nd)

2020-21:
7.59 HDSA/60 (14th)
1.12 HDGA/60 (4th)
58.12 HDGF% (4th)

2021-22:
7.45 HDSA/60 (7th)
1.34 HDGA/60 (12th)
53.33 HDGF% (8th)

A bit of a drop off this year, probably fatigue from back to back cup runs.

 

[zeke]

Where does the Brodie-Holl pair stand if Rielly-Muzzin-Sandin is your left side?

That's the question.

 

[zeke]

High danger metrics are the most predictive when it comes to playoff success and it's not even close.

Not actually true.

Slavin, Pelech and McAvoy are playoff type players and their high danger metrics prove it.

These guys have excellent metrics across the board.

 

[tmlfan98]

Not actually true.

Do you have any proof to support that there are more predictive metrics when it comes to playoff success than high danger metrics?

Because in my case, superior high danger metrics have crazy accuracy in predicting the winner of a playoff series. Some good background info on this can be found in points 15-17 of this old 32 thoughts article: 31 Thoughts: Why Blackhawks must publicly address sexual assault allegations

 

[Buds17]

Tampa has been running a left side of Hedman-McDonagh-Sergachev for years.

Valid point. I'd just be uncertain about how the rest of our blueline shakes out under those circumstances.

 

[tmlfan98]

Valid point. I'd just be uncertain about how the rest of our blueline shakes out under those circumstances.

With the group we have at the moment, if everyone was healthy I'd probably go with

Rielly-Lyubushkin
Brodie-Holl
Muzzin-Sandin/Liljegren

 

[JT AM da real deal]

Stats mean nothing .. especially regular season stats .. hockey people look at raw skills .. Holl is 2nd slowest Leaf defenseman in a race from goal line to goal line (Sandin is slowest) .. Holl has lowest skill level of all our defense (not sure about Lub because have not seen him at camp) .. you guys are spreadsheet types and are ruled like Dubie on these numbers which mean NOTHING to us guys who played .. we assess plays and skills which we see and know from personal prior experiences playing da game at junior levels and beyond into pro hockey .. but one thing i notie about Holl is when he starts to think he is Orr he causes our entire team issues .. when he plays according to his skill level and stays back he can play on a 3rd pair on a decent team and a 7th on a Cup contender team .. right now we are completely missing a 2nd pair to be a Cup contender and likely do not have a tender who can win a Cup (maybe Woll we will see soon enough)

 

[zeke]

Do you have any proof to support that there are more predictive metrics when it comes to playoff success than high danger metrics?

Because in my case, superior high danger metrics have crazy accuracy in predicting the winner of a playoff series. Some good background info on this can be found in points 15-17 of this old 32 thoughts article: 31 Thoughts: Why Blackhawks must publicly address sexual assault allegations

You are the one making a claim. It's up to you to prove it. Every simple calculation I've ever seen has shown that the "high danger" stat, especially the one on naturalstattrick.com, is fairly useless.

Note that even in those points 15-17 they note that the next year TB wasn't any good in that stat and still won, and of course that rhe Leafs were one of the best in that stat last year.

 

[tmlfan98]

You are the one making a claim. It's up to you to prove it. Every simple calculation I've ever seen has shown that the "high danger" stat, especially the one on naturalstattrick.com, is fairly useless.

Note that even in those points 15-17 they note that the next year TB wasn't any good in that stat and still won, and of course that rhe Leafs were one of the best in that stat last year.

Prove the bolded.

Using the last 3 seasons, defending champs Tampa (18th, far from "not any good" as you claim) beat Montreal (2nd) in last years final. The Leafs ranked 3rd overall last year in xHDGA/60 but were still inferior to their first round opponent. In the 2020 finals Tampa (1st) beat Dallas (4th). In the 2019 finals St. Louis (2nd) beat Boston (3rd).

So if you look at the last 3 seasons the only "outlier" if you even want to call them that is 2021 Tampa who were the defending champs coming off a long playoff run the previous season and finished 1st in xHDGA/60 the year before. What exactly about high danger metrics are useless again?

 

[zeke]

Prove the bolded.

Using the last 3 seasons, defending champs Tampa (18th, far from "not any good" as you claim) beat Montreal (2nd) in last years final. The Leafs ranked 3rd overall last year in xHDGA/60 but were still inferior to their first round opponent. In the 2020 finals Tampa (1st) beat Dallas (4th). In the 2019 finals St. Louis (2nd) beat Boston (3rd).

So if you look at the last 3 seasons the only "outlier" if you even want to call them that is 2021 Tampa who were the defending champs coming off a long playoff run the previous season and finished 1st in xHDGA/60 the year before. What exactly about high danger metrics are useless again?

The defending champs ranked 18th. That's pretty clear.

Where are you getting those numbers btw - they don't match up with any high danger numbers I can find.

For example, the 2020 Stars were very mediocre from the high danger stat I can find, while performing elitely via xGA.

 

[tmlfan98]

The defending champs ranked 18th. That's pretty clear.

Where are you getting those numbers btw - they don't match up with any high danger numbers I can find.

For example, the 2020 Stars were very mediocre from the high danger stat I can find, while performing elitely via xGA.

I crunch them myself because xHDGA/60 doesn't exist anywhere I can find on public sites. To calculate it yourself you have to factor out goaltending errors.

And in terms of your claim about the 2020 Stars high danger defense being mediocre, it is simply false. The 2019-20 the Stars ranked 4th in the league in HDGA/60 and 4th in xHDGA/60 as well. Their high danger defense was elite that season.

 

[zeke]

I crunch them myself because xHDGA/60 doesn't exist anywhere I can find on public sites. To calculate it yourself you have to factor out goaltending errors.

And in terms of your claim about the 2020 Stars high danger defense being mediocre, it is simply false. The 2019-20 the Stars ranked 4th in the league in HDGA/60 and 4th in xHDGA/60 as well. Their high danger defense was elite that season.

Wait you're calculating your own stats?

I have to take your word for it?

I thought we were talking about publicly available HDCA chances.

For example, the Stars that year were 13th in high danger chances against, and 6th in expected goals against.

 

[123offtheglass]

Wait you're calculating your own stats?

I have to take your word for it?

I thought we were talking about publicly available HDCA chances.

For example, the Stars that year were 13th in high danger chances against, and 6th in expected goals against.

 

[tmlfan98]

Wait you're calculating your own stats?

I have to take your word for it?

I thought we were talking about publicly available HDCA chances.

For example, the Stars that year were 13th in high danger chances against, and 6th in expected goals against.

To the bolded, yes I calculate them myself by factoring out goaltending errors. You don't have to take my word for it if you don't want to.

Looking at the whole picture at 5v5 for the 2020 Stars:

They ranked 6th in HDSA/60, 4th in HDGA/60, 4th in xHDGA/60 and 12th in HDCA/60
Using regular xGA metrics, they ranked 12th in SA/60, 2nd in GA/60, 6th in xGA/60 and 18th in CA/60

So basically, you are clinging to their HDCA/60 ranking just outside the top 10 to suggest they were a mediocre high danger defensive team in general, when they ranked no lower than 6th in the other 3 major high danger defensive metrics.

 

[keonsbitterness]

Wait you're calculating your own stats?

I have to take your word for it?

I thought we were talking about publicly available HDCA chances.

For example, the Stars that year were 13th in high danger chances against, and 6th in expected goals against.

Why is that worse than taking the word of a public site stat run by people you also don't know?

 

[zeke]

To the bolded, yes I calculate them myself by factoring out goaltending errors. You don't have to take my word for it if you don't want to.

Looking at the whole picture at 5v5 for the 2020 Stars:

They ranked 6th in HDSA/60, 4th in HDGA/60, 4th in xHDGA/60 and 12th in HDCA/60
Using regular xGA metrics, they ranked 12th in SA/60, 2nd in GA/60, 6th in xGA/60 and 18th in CA/60

So basically, you are clinging to their HDCA/60 ranking just outside the top 10 to suggest they were a mediocre high danger defensive team in general, when they ranked no lower than 6th in the other 3 major high danger defensive metrics.

I don't know what to say - but your numbers are not anything I've seen anywhere before, nor have you explained how you put them together. I just don't know how to evaluate your numbers.

Based on the data that's available, I haven't seen any evidence that "high danger chances" is a good stat in either end. Maybe you have developed a better stat - but you should probably publish it and explain it - it might even make you rich.