Player Discussion J.T. Miller Part II

[silverfish]

How much worse was Kevin Hayes?

Much. I'd imagine some of this was compounded by them playing together.

Important to keep in mind that you can be mad at both of them at the same time, though

 

[Mac n Gs]

Much. I'd imagine some of this was compounded by them playing together.

Important to keep in mind that you can be mad at both of them at the same time, though

I am, but I am just more annoyed with Kevin Hayes

 

[RGY]

So Miller's advanced stats and regular stats both suck

 

[silverfish]

So Miller's advanced stats and regular stats both suck

*Since March 1, yes.

 

[Thirty One]

So Miller's advanced stats and regular stats both suck

And I'm not that fond of his face.

 

[True Blue]

This past post-season, Zibanejad had 9 points. Kreider, Miller and Hayes combined for 10.

Those 3 need to be better.

No doubt. But unlike the latter two, Kreider has a history of success in the NHL.

Yet, Nash was labeled a play-off choker for 1 mediocre post-season, based on the no of goals he scored.

It was not for one mediocre post-season. It was, and still is, due to the fact that in 77 games, he has 15 goals. Coming into these playoffs, he scored 11 goals in 61 playoff games for the Rangers.

 

[True Blue]

Because it's an arbitrary number with no context or analysis put into it

It's very easy to ignore actually

1 goal in 40 games is not arbitrary. I would say it spells a pretty clear picture.

I dream of a world where people stop reading into playoff boxcar stats to decide a player's quality

Yes, it would be far better to exist in fantasy land where extrapolated stats give you extrapolated amounts.

 

[silverfish]

1 goal in 40 games is not arbitrary. I would say it spells a pretty clear picture.

Yes, it would be far better to exist in fantasy land where extrapolated stats give you extrapolated amounts.

It's arbitrary because you're taking data across seasons and adding it up like it's all the same. It's not. Especially in a sample size as small as the playoffs.

In terms of "extrapolated" stats... do you also hate goals against average?

 

[True Blue]

It's arbitrary because you're taking data across seasons and adding it up like it's all the same. It's not. Especially in a sample size as small as the playoffs.

How many games played and how many goals scored is what I call binary. You either score or you do not score. That is not arbitrary.

In terms of "extrapolated" stats... do you also hate goals against average?

Not to be snide here, but:

Extrapolated: extend the application of (a method or conclusion, especially one based on statistics) to an unknown situation by assuming that existing trends will continue or similar methods will be applicable.


Not quite the same thing and you know it. Typically the goalie plays the whole game. There fore taking the number of goals that he allows multiplying it by 60 and then dividing the number by minutes played will not have nearly that much noise. However, that is different than taking what a skater does and extrapolating that to a /60 stat. That stat attempts to equalize all players due to discrepancy in actual time played on ice or mismatch in games played. Which is living in fantasy land. It can make a role player seem fantastic. It is a tool, hardly the end all and be all.

We already had a discussion where the per 60 stat was attempted to illustrate that Nash was a better playoff performer than Ovechkin and in fact, since the trade Nash was the best even strength scorer in the league. But let's not digress.

FYI: When looking at a goalie, save percentage is a much better indicator of the goalie's performance.

Lastly, I do not hate advanced stats. I understand them and believe that they have a place in hockey and should be looked upon. That said, what I cannot stand is the group think that one can evaluate strictly based on looking at them. One needs to be able to evaluate both. The best analysis is based on projections but also on reality. And the best ROI is brought back when projections are updated based on what is actually transpiring in the real world.

 

[Mandar]

It's arbitrary because you're taking data across seasons and adding it up like it's all the same. It's not. Especially in a sample size as small as the playoffs.

In terms of "extrapolated" stats... do you also hate goals against average?

The playoffs are always going to be a limited sample size per year. If no sample size is big enough due to these parameters, then no player should be criticized for their production, correct?

By this thinking, here is no such thing as a bad playoff performer due to production (I guess other analytics would be useless as well due to sample size across seasons).

 

[silverfish]

How many games played and how many goals scored is what I call binary. You either score or you do not score. That is not arbitrary.

So then someone coudl theoretically find all the games that JT Miller scored in during the regular season and say: "Wow, JT Miller scored 22 goals in 22 games!".

What's the difference?

Not to be snide here, but:

Extrapolated: extend the application of (a method or conclusion, especially one based on statistics) to an unknown situation by assuming that existing trends will continue or similar methods will be applicable.

Not quite the same thing and you know it. Typically the goalie plays the whole game. There fore taking the number of goals that he allows multiplying it by 60 and then dividing the number by minutes played will not have nearly that much noise. However, that is different than taking what a skater does and extrapolating that to a /60 stat. That stat attempts to equalize all players due to discrepancy in actual time played on ice or mismatch in games played. Which is living in fantasy land. It can make a role player seem fantastic. It is a tool, hardly the end all and be all.

We already had a discussion where the per 60 stat was attempted to illustrate that Nash was a better playoff performer than Ovechkin and in fact, since the trade Nash was the best even strength scorer in the league. But let's not digress.

GAA is a per 60 stat. Goalies don't always play 60 minutes. Sometimes they play less, sometimes they play more. If Henrik playes 58 minutes and gives up 2 goals, whose to say that he wouldn't have given up a 3rd in those next two minutes? His GAA is 2.07, but by god, behind this defense, he could've given up 3!

When people use per 60 stats as a metric, they aren't saying that if Nash played 60 minutes every game, he'd score whatever his G/60 is. It's a rate stat. Just like GAA. You're misunderstanding the intent of people when they use per 60.

More and more across HF there is such a divide between what a metric being used is actually showing, and what people who read that post think the poster is trying to show by using that metric.

The playoffs are always going to be a limited sample size per year. If no sample size is big enough due to these parameters, then no player should be criticized for their production, correct?

By this thinking, here is no such thing as a bad playoff performer due to production (I guess other analytics would be useless as well due to sample size across seasons).

Straight up ignoring the fact that it's also a sum across 4 years.

Y'all... if I'M defending JT, you know you're wrong.

 

[True Blue]

So then someone coudl theoretically find all the games that JT Miller scored in during the regular season and say: "Wow, JT Miller scored 22 goals in 22 games!".

Except that you are evaluating the entire season and not 22 games. So no.

GAA is a per 60 stat. Goalies don't always play 60 minutes. Sometimes they play less, sometimes they play more.

I understand what it is. However, a goalie plays muck closer to 60 minutes per game than not. Sometimes Henke gets pulled. Does not happen often. Sometimes there are overtime games. Again, most of the time he will play 60. Unlike a skater who AT BEST will play half the game. Or usually less than half and would be lucky to play a third.

If Henrik playes 58 minutes and gives up 2 goals, whose to say that he wouldn't have given up a 3rd in those next two minutes? His GAA is 2.07, but by god, behind this defense, he could've given up 3!

Which is why it is not a perfect stat and save percentage is much, much better.

When people use per 60 stats as a metric, they aren't saying that if Nash played 60 minutes every game, he'd score whatever his G/60 is. It's a rate stat. Just like GAA. You're misunderstanding the intent of people when they use per 60.

No I understand it just fine. I am not saying that they are stating that Nash plays 60 minutes. What is being done is that his stats are being extrapolated over 60 and compared to other players. Extrapolated results are far, far from perfect and should be judged with skepticism not as an absolute fact.

More and more across HF there is such a divide between what a metric being used is actually showing, and what people who read that post think the poster is trying to show by using that metric.

Reasonably certain when I was told that Nash is the league's best goal scorer at even strength that was exactly what was meant. The advanced stats crowd pretends that the detractors do not understand and are too stupid to see (this does not pertain to you). The reality is quite different.

Straight up ignoring the fact that it's also a sum across 4 years.

When playoff games begin to total a half a season, it is pretty meaningful.

 

[Amazing Kreiderman]

/60 stats are there to compare players despite not having the same TOI. It's not rocket science

 

[FLYLine27*]

Miller had a GREAT season. I dont care if he faded off in the end, he broke all his previous totals and took a huge step forward. Lets hope for another step forward next season.

 

[silverfish]

(Making up numbers because I'm too lazy to look it up)

Nash records 2.3 points per 60!

Fact (in this scenario): Nash records 2.3 points per 60
Analysis: Everything else derived from the above

Your beef is with analysis, and not fact. I continue to maintain that the analysis you think people are doing based off of per 60 metrics is not what they are doing.

 

[allrevvedup25]

Miller had a GREAT season. I dont care if he faded off in the end, he broke all his previous totals and took a huge step forward. Lets hope for another step forward next season.

This!! I don't want the Rangers to trade Miller. I want to see if he can build off it. Yes, he had a horrible playoffs but maybe he can turn it around next year (if they make the playoffs.) Miller is someone you keep and not trade because if he continues to improve, you will be watching him do so on another team!

 

[NickyFotiu]

Miller had a GREAT season. I dont care if he faded off in the end, he broke all his previous totals and took a huge step forward. Lets hope for another step forward next season.

Last offseasons JT hate now looks silly in hindsight so its time to mix it up some.

 

[True Blue]

/60 stats are there to compare players despite not having the same TOI. It's not rocket science

One of the several reasons as to why it is flawed and should not be used as the sole method of comparison. It is not rocket science. But looking at advanced stats in conjunction with what is actually happening is hardly brain surgery.

If you look at goals/60 for the last 5 year period, it would tell you that Rick Nash is better than Malkin, Kane, Ovechkin & Crosby. Is that realistic?

 

[True Blue]

I continue to maintain that the analysis you think people are doing based off of per 60 metrics is not what they are doing.

We will have to agree to disagree there.

 

[True Blue]

This!! I don't want the Rangers to trade Miller. I want to see if he can build off it. Yes, he had a horrible playoffs but maybe he can turn it around next year (if they make the playoffs.) Miller is someone you keep and not trade because if he continues to improve, you will be watching him do so on another team!

Agreed. People were hot to trade Krieder after last year. Let the players develop here and learn here. While everyone is tradable, some assets are better than others.

 

[silverfish]

One of the several reasons as to why it is flawed and should not be used as the sole method of comparison. It is not rocket science. But looking at advanced stats in conjunction with what is actually happening is hardly brain surgery.

If you look at goals/60 for the last 5 year period, it would tell you that Rick Nash is better than Malkin, Kane, Ovechkin & Crosby. Is that realistic?

We will have to agree to disagree there.

No. Look at the bolded.

G/60 doesn't tell you who is better. It tells you who scores more goals per 60 minutes of ice time.

So again. Your beef is with the analysis part, and not the metric. Yet, you are the only one who is out here saying things like: "it would tell you that Rick Nash is better than Malkin...". The hilarious thing here is, is that you hate these per 60 metrics because of the analysis that you yourself think is happening.

It's like a weird dog chasing his tail thing going on.

 

[True Blue]

G/60 doesn't tell you who is better. It tells you who scores more goals per 60 minutes of ice time.

I am fully aware of what it tells you

So again. Your beef is with the analysis part, and not the metric. Yet, you are the only one who is out here saying things like: "it would tell you that Rick Nash is better than Malkin...". The hilarious thing here is, is that you hate these per 60 metrics because of the analysis that you yourself think is happening.

Here we friggin' go again. When the /60 stat was utilized to tell me that Nash performs better in the playoffs than Ovechkin, is my issue the statistic or how it is being presented? When the /60 stat was used to state that Rick Nash is the best ES scorers in the entire NHL since coming to the Rangers, is my issue the analysis or now it is being used?

My issue is not the damn stat. Never has been. My issue was, is, and will be when it is attempted to be used as the sole level of evidence to make a case as to how a player performs.

It's like a weird dog chasing his tail thing going on.

Tell me about it

 

[silverfish]

I am fully aware of what it tells you

Here we friggin' go again. When the /60 stat was utilized to tell me that Nash performs better in the playoffs than Ovechkin, is my issue the statistic or how it is being presented? When the /60 stat was used to state that Rick Nash is the best ES scorers in the entire NHL since coming to the Rangers, is my issue the analysis or now it is being used?

My issue is not the damn stat. Never has been. My issue was, is, and will be when it is attempted to be used as the sole level of evidence to make a case as to how a player performs.

Tell me about it

G
per
60
DOES
NOT
TELL
YOU
WHO
IS
BETTER
OR
WHO
PERFORMS
BETTER
IT
TELLS
YOU
THE
RATE
AT
WHICH
SKATERS
SCORE
GOALS

 

[True Blue]

g
per
60
does
not
tell
you
who
is
better
or
who
performs
better
it
tells
you
the
rate
at
which
skaters
score
goals

i
perfectly
understand
that
however
it
was
being
presented
as
evidence
of
being
better.
Saying
that
he
is
the
best
scorer
at
even
strength
is
an
attempt
to
use
the
statistic
as
evidence.

 

[silverfish]

i
perfectly
understand
that
however
it
was
being
presented
as
evidence
of
being
better.
Saying
that
he
is
the
best
scorer
at
even
strength
is
an
attempt
to
use
the
statistic
as
evidence.

He is the most productive even strength goal scorer. Not the best. There's a difference.