Conflict in the Crease

[roboturner 562]

7 minutes ago, Echolalia said:

That was also my point when I made my original comment 

Really? Cuz you specifically asked what was wrong with cherry picking Howard's stats, then continued on to push aside the struggles Howard has had in his career.

[kickazz 5,459]

1 minute ago, roboturner said:

Really? Cuz you specifically asked what was wrong with cherry picking Howard's stats, then continued on to push aside the struggles Howard has had in his career.

Probably  because the other posters cherry pick Mrazek's stats so he followed suit. An eye for an eye! 

[Buppy 1,720]

1 hour ago, Echolalia said:

Except we don't know that what Mrazek did for the first half of his career is his norm.  You label him as an average starter based on two halfs of a season playing well and two halfs of a season playing terribly, which is incredibly misleading.  As you alluded to, the sample size with Mrazek is tiny, and half his career has been one extreme and the most recent half has been the other extreme. The confidence that you can say Mrazek's actual game is one vs the other is very low, thus like I've beat to death 100 times already Mrazek is an unproven commodity who's played like a fringe backup for almost the past year now, and giving him the net without having someone proven ready to clean his potential mess is idiotic.  And let's be clear about the recent stats tipping in Mrazek'a favor.  Unless you're cherry picking the second half of the 2014-2015 season and the first half of 2015-2016 season, the recent stats have been brutal for Mrazek.

But I'm not looking at just the first half of his career. I'm looking at all of it. His entire career, including playoffs...all the good, all the bad, all the ugly. 125 games, 111 starts, 6828 minutes, 3271 shots against. Over that period he has a .917 sv%, .577 QS%, 2.39 GAA. Respectable numbers. The sample may be small, but it's everything there is.

In regards to the recent stats, Howards numbers from 2014 through today: .912 sv%, .52 QS%, 2.52 GAA. Even going back to 2012: .914 sv%, .583 QS%, 2.47 GAA. 

You're basically saying there is a high chance that a 100+ game sample might be anomalous, while also saying there's a high chance that a 17 or 21 game sample isn't. You can't have it both ways.

[ChristopherReevesLegs 7,022]

11 hours ago, Buppy said:

The only "random sampling" going on here is from you.

Looking at the entirety of the careers for both players, the stats are nearly identical. In the most recent years (which as you say is more relevant, plus also removes possible bias from the team in front of them) the stats tip in Mrazek's favor. This year it's tipped to Howard.

But also looking at this year, both are far from their career numbers. Given the small sample size, we should conclude that what we've seen this year is an anomaly, and should not be expected to continue. Most likely going forward both will trend back toward their norm.

Again, it has nothing to do with Mrazek being the future or a potential star. That doesn't matter. Over the course of 7 seasons Howard has shown he is just an average starter. He's not going to be more than that. Mrazek has shown the same thing. So why do we need both?

If your answer is because of this year, you are doing the exact thing you've been mocking others for doing.

It's a stat from hockey-reference.com. A start with a sv% > the league average for that season, or at least .885 if facing 20 or fewer shots.

For players with NTCs, I stopped counting when I got to 20-ish. That was going back to the summer before last season. 

So let me clarify. Your saying 20+ players with NTCs were traded over the last 3 seasons including this one?

[Echolalia 2,961]

9 hours ago, roboturner said:

Really? Cuz you specifically asked what was wrong with cherry picking Howard's stats, then continued on to push aside the struggles Howard has had in his career.

As kickazz mentioned.  I was being cheeky and went off on a little sarcastic tangent to illustrate a point.

 

8 hours ago, Buppy said:

But I'm not looking at just the first half of his career. I'm looking at all of it. His entire career, including playoffs...all the good, all the bad, all the ugly. 125 games, 111 starts, 6828 minutes, 3271 shots against. Over that period he has a .917 sv%, .577 QS%, 2.39 GAA. Respectable numbers. The sample may be small, but it's everything there is.

In regards to the recent stats, Howards numbers from 2014 through today: .912 sv%, .52 QS%, 2.52 GAA. Even going back to 2012: .914 sv%, .583 QS%, 2.47 GAA. 

You're basically saying there is a high chance that a 100+ game sample might be anomalous, while also saying there's a high chance that a 17 or 21 game sample isn't. You can't have it both ways.

I'm not asking to have it both ways.  Again, I have no horse in the Howard race and I don't necessarily see him as a staple of the Wings' future (as I've already mentioned, yet for some reason you still keep going back to him as if I'm arguing that he is the answer for the Wings going forward).  What I'm arguing is that Mrazek is not necessarily the god that everyone made him out to be early in his career and continue to do so despite evidence to the contrary.  He's been unable to establish himself as a number one goalie for more than half a season at any point in his NHL career, he's wildly sporadic, his last year of hockey has been among the worst goaltending I've personally observed from one goalie, and the periods of his career where he played well are limited by sample size to be able to conclude whether that's his norm or if he was playing above his head.  There's a lot of red flags, hence my original point that if Howard is traded away, Holland better find another goalie to work with Mrazek where the above isn't a concern.

[ChristopherReevesLegs 7,022]

The islanders just waived Jaroslav Halak with 2 years and 4.5 aav on the books. Tell me again Howard-haters how you're going to trade Howie? Its amazing how many times i ask this with no response. Mrazek is the odd man out.

[Jacksoni 418]

My money is on Coreau landing the gig as the starter in the outdoor game in Toronto. Lifts this goalie discussion to a threesome. Hopefully it's lighting a fire under Mrazek's behind. He doesn't just need to pick it up, he needs to lift up another level from his previous good performances if he is to be the clear starter.

I do think he has the talent required, it'll be interesting to see how this unfolds.

[frankgrimes 1,836]

Nr.1 goalie is out, Coreau played well enough to earn the opportunity so he deserves this chance

[Buppy 1,720]

13 hours ago, ChristopherReevesLegs said:

So let me clarify. Your saying 20+ players with NTCs were traded over the last 3 seasons including this one?...

The islanders just waived Jaroslav Halak with 2 years and 4.5 aav on the books. Tell me again Howard-haters how you're going to trade Howie? Its amazing how many times i ask this with no response. Mrazek is the odd man out.

Actually, I'm saying 20+ players with NTCs were traded in the last TWO seasons, including this one, plus the summer prior to last season. If I were to go back to the 2014 trade deadline, it'd probably be 30+. Dave Bolland, Derick Brassard, Bryan Bickell, Niklas Backstrom, John-Michael Liles, Alex Tanguay, Eric Staal, Rob Scuderi (twice), Andrew Ladd, Raffi Torres, Dion Phanuef, Colin Greening, Milan Michalek, Vincent Lecavalier, Luke Schenn, Patrick Sharp, Trevor Daley, Troy Brouwer, Phil Kessel, Kevin Bieksa. Might have missed a couple.

Obviously Howard is unlikely to be traded right now. But if he comes back and finishes strong, there'd be a chance during the summer. By your logic, Mrazek would be equally impossible to trade.

 

13 hours ago, Echolalia said:

I'm not asking to have it both ways.  Again, I have no horse in the Howard race and I don't necessarily see him as a staple of the Wings' future (as I've already mentioned, yet for some reason you still keep going back to him as if I'm arguing that he is the answer for the Wings going forward).  What I'm arguing is that Mrazek is not necessarily the god that everyone made him out to be early in his career and continue to do so despite evidence to the contrary.  He's been unable to establish himself as a number one goalie for more than half a season at any point in his NHL career, he's wildly sporadic, his last year of hockey has been among the worst goaltending I've personally observed from one goalie, and the periods of his career where he played well are limited by sample size to be able to conclude whether that's his norm or if he was playing above his head.  There's a lot of red flags, hence my original point that if Howard is traded away, Holland better find another goalie to work with Mrazek where the above isn't a concern.

What you're arguing is a strawman. I'm not saying Mrazek is going to be elite. I don't think anyone is. That is not the reason I think we should try to move Howard.

I'm saying it doesn't matter. If Mrazek has red flags and isn't reliable and isn't the answer, and Howard has red flags and isn't reliable and isn't the answer...WHY do we need someone who is in order to get rid of Howard? (Or Mrazek for that matter.) How do we get worse by swapping one unreliable non-answer for a different, cheaper, unreliable non-answer?

Do you just think Mrazek sucks? That what we've seen from him this year is likely to continue?

Because if you actually look at his career, that is just as absurd as thinking he's likely to become elite, and far, far, far more absurd than thinking he could solid, mid-range starter.

 

[Echolalia 2,961]

1 hour ago, Buppy said:

Actually, I'm saying 20+ players with NTCs were traded in the last TWO seasons, including this one, plus the summer prior to last season. If I were to go back to the 2014 trade deadline, it'd probably be 30+. Dave Bolland, Derick Brassard, Bryan Bickell, Niklas Backstrom, John-Michael Liles, Alex Tanguay, Eric Staal, Rob Scuderi (twice), Andrew Ladd, Raffi Torres, Dion Phanuef, Colin Greening, Milan Michalek, Vincent Lecavalier, Luke Schenn, Patrick Sharp, Trevor Daley, Troy Brouwer, Phil Kessel, Kevin Bieksa. Might have missed a couple.

Obviously Howard is unlikely to be traded right now. But if he comes back and finishes strong, there'd be a chance during the summer. By your logic, Mrazek would be equally impossible to trade.

 

What you're arguing is a strawman. I'm not saying Mrazek is going to be elite. I don't think anyone is. That is not the reason I think we should try to move Howard.

I'm saying it doesn't matter. If Mrazek has red flags and isn't reliable and isn't the answer, and Howard has red flags and isn't reliable and isn't the answer...WHY do we need someone who is in order to get rid of Howard? (Or Mrazek for that matter.) How do we get worse by swapping one unreliable non-answer for a different, cheaper, unreliable non-answer?

Do you just think Mrazek sucks? That what we've seen from him this year is likely to continue?

Because if you actually look at his career, that is just as absurd as thinking he's likely to become elite, and far, far, far more absurd than thinking he could solid, mid-range starter.

 

Bro you're the one putting up strawmen.  I never claimed you said you believed Mrazek to be elite, I never said I thought his god-awful hockey would nessesarily continue in subsequent seasons, I never claimed to believe Howard was going to be elite, in fact my discussion with the other posters wasn't even about Howard until you jumped into the conversation out of the blue and started repeatedly comparing him to Mrazek as if it was relevant to what I was saying. I'm honestly not even sure what you're arguing at this point, and based on everything you keep bringing up it seems you're either intentionally bringing in strawmen to argue against or you're legitimately clueless about what my point has been thus far despite me laying it out.  But I would like you to clarify one point.  You just mentioned you don't think it matters that Mrazek has all these red flags and we don't get another goalie who can compensate for those issues on the event Mrazek doesn't pan out?  Is this what you're suggesting?  So, like a Mrazek/Coreau combo for next year?  Or am I mistaken?  

Edited December 31, 2016 by Echolalia

[Buppy 1,720]

54 minutes ago, Echolalia said:

Bro you're the one putting up strawmen.  I never claimed you said you believed Mrazek to be elite, I never said I thought his god-awful hockey would nessesarily continue in subsequent seasons, I never claimed to believe Howard was going to be elite, in fact my discussion with the other posters wasn't even about Howard until you jumped into the conversation out of the blue and started repeatedly comparing him to Mrazek as if it was relevant to what I was saying. I'm honestly not even sure what you're arguing at this point, and based on everything you keep bringing up it seems you're either intentionally bringing in strawmen to argue against or you're legitimately clueless about what my point has been thus far despite me laying it out.  But I would like you to clarify one point.  You just mentioned you don't think it matters that Mrazek has all these red flags and we don't get another goalie who can compensate for those issues on the event Mrazek doesn't pan out?  Is this what you're suggesting?  So, like a Mrazek/Coreau combo for next year?  Or am I mistaken?  

Well I'm not entirely sold that Coreau is an NHL goalie, but more or less yes. Regardless of the red flags, we do not need "an established #1" as you said earlier in order to justify getting rid of Howard. We only need a solid backup.

I brought up Howard because you are saying we can't get rid of him unless we get a reliable starter. I'm not sure how to make this any more clear. If Mrazek is not the answer for our future goaltending, and Howard is not the answer...then we don't have an answer. How is it worse to also not have an answer without Howard?

I'm not saying getting rid of Howard is the solution to our goaltending future. I'm saying keeping him is not. Whether or not Mrazek is our future is a separate issue. If he's not, then we also need to get rid of him. 

Ok, maybe you're not saying I think Mrazek will be elite, but you mention it in every post as if someone is arguing that. In the previous post you said: " What I'm arguing is that Mrazek is not necessarily the god that everyone made him out to be early in his career and continue to do so despite evidence to the contrary." So if that wasn't directed at me, who was it directed at and why did you include it in a reply to me? 

And I didn't say you thought anything about Mrazek or Howard. I was asking you if that is what you thought. There's a difference. I'll try to be more direct.

Do you think Mrazek is particularly likely to be anything less than what he has been so far in his career; good at times, bad at times, overall pretty average?

Or do you think he has not been average overall in his career?

Do you think Howard is particularly likely to be anything more than what he has been so far in his career; good at times, bad at times, overall pretty average?

[Echolalia 2,961]

24 minutes ago, Buppy said:

Well I'm not entirely sold that Coreau is an NHL goalie, but more or less yes. Regardless of the red flags, we do not need "an established #1" as you said earlier in order to justify getting rid of Howard. We only need a solid backup.

I brought up Howard because you are saying we can't get rid of him unless we get a reliable starter. I'm not sure how to make this any more clear. If Mrazek is not the answer for our future goaltending, and Howard is not the answer...then we don't have an answer. How is it worse to also not have an answer without Howard?

I'm not saying getting rid of Howard is the solution to our goaltending future. I'm saying keeping him is not. Whether or not Mrazek is our future is a separate issue. If he's not, then we also need to get rid of him. 

Ok, maybe you're not saying I think Mrazek will be elite, but you mention it in every post as if someone is arguing that. In the previous post you said: " What I'm arguing is that Mrazek is not necessarily the god that everyone made him out to be early in his career and continue to do so despite evidence to the contrary." So if that wasn't directed at me, who was it directed at and why did you include it in a reply to me? 

And I didn't say you thought anything about Mrazek or Howard. I was asking you if that is what you thought. There's a difference. I'll try to be more direct.

Do you think Mrazek is particularly likely to be anything less than what he has been so far in his career; good at times, bad at times, overall pretty average?

Or do you think he has not been average overall in his career?

Do you think Howard is particularly likely to be anything more than what he has been so far in his career; good at times, bad at times, overall pretty average?

C'mon Buppsie.  You have seen as often as I have the amount of Mrazek dick-sucking that happens on these forums, regardless of how the kid plays.  In fact I think my rant about exactly that was what prompted you to get involved in the discussion, so coming into this you probably should have been aware that I was addressing the topic at hand with the sky-high expectations that many posters have for him in mind.  The conversation didn't evolve into me bringing that up after you joined in.  That was already in play.  And that's why it was in my reply.

With regard to Mrazek, I have no clue what is going to be his future.  Half his short career he's played splendidly, and the other half he's played like garbage, and both of those samples are too small to make any meaningful predictions of.  His career stats are a reflection of two bell curves on the extremes without as much data points around his actual career averages as one would expect to see.  Because of the uncertainty, as well as the direction he's trended in, a Mrazek/Coreau combo is absolutely loony.

And I think similarly about Howard, particularly because he has a knack for getting injured, and then struggling to return to form following injury.  That's a big concern of mine this season when Howard returns from injury, because I have little faith in Mrazek to suddenly turn it on, although if it happens I would be delighted.

[Buppy 1,720]

2 hours ago, Echolalia said:

...With regard to Mrazek, I have no clue what is going to be his future.  Half his short career he's played splendidly, and the other half he's played like garbage, and both of those samples are too small to make any meaningful predictions of.  His career stats are a reflection of two bell curves on the extremes without as much data points around his actual career averages as one would expect to see.  Because of the uncertainty, as well as the direction he's trended in, a Mrazek/Coreau combo is absolutely loony.

And I think similarly about Howard, particularly because he has a knack for getting injured, and then struggling to return to form following injury.  That's a big concern of mine this season when Howard returns from injury, because I have little faith in Mrazek to suddenly turn it on, although if it happens I would be delighted.

But if you have similar feelings and concerns about both, I don't understand why having both is better. I guess there's situations like this year, where one is playing well while the other is slumping, but you'll also have spells where both are bad or both are good. I'd rather have the extra cap space and a roster spot open for someone else who could potentially become our future, particularly right now during a rebuild when it's less important.

In regards to the bold: If you look into goaltending data, high and low extremes are actually far more common than "average" performances. In any given game, giving up 2 or fewer goals is probably going to be above average, while 3 or more is below. It's pretty rare for a goalie to be completely consistent with an "above" game, then a "below" game (much less above and below by equal amounts). So even when you go out to a 5 or even 10 game stretch, chances are it's going to be either above or below average. In some cases (and I'm guessing probably much less rare than people would think) it can be well above or below. Even when you look at full season stats for a goalie it's fairly common to have some significant variation year to year. But when you get a good sample size, like 100+ games, it should give you a good idea about what a goalie is.

Mrazek may or may not be prone to unusually long "down" periods and conversely, unusually long "up" periods, but it's more likely that it's just a fluke. In the future, we should probably expect a more normal distribution of both. But regardless, based on his career to date, we should expect a more high than low, with all of it adding up to fairly average.

[LeftWinger 4,951]

First post of 2017! Ya!

Coreau is good.

[DickieDunn 2,571]

There's a difference between a goalie with a 2.5 gaa( random number) who consistently allows around that average and one who had a long stretch where he was at 2.0 gaa then an equally long stretch at 3.0 gaa.

Sent from my LGLS676 using Tapatalk

[Buppy 1,720]

1 hour ago, DickieDunn said:

There's a difference between a goalie with a 2.5 gaa( random number) who consistently allows around that average and one who had a long stretch where he was at 2.0 gaa then an equally long stretch at 3.0 gaa.

Sent from my LGLS676 using Tapatalk
 

Debatable, but also irrelevant. If it's a problem for Mrazek, it's a problem for Howard. So why do we need, or want, both?

[Echolalia 2,961]

14 hours ago, Buppy said:

But if you have similar feelings and concerns about both, I don't understand why having both is better. I guess there's situations like this year, where one is playing well while the other is slumping, but you'll also have spells where both are bad or both are good. I'd rather have the extra cap space and a roster spot open for someone else who could potentially become our future, particularly right now during a rebuild when it's less important.

In regards to the bold: If you look into goaltending data, high and low extremes are actually far more common than "average" performances. In any given game, giving up 2 or fewer goals is probably going to be above average, while 3 or more is below. It's pretty rare for a goalie to be completely consistent with an "above" game, then a "below" game (much less above and below by equal amounts). So even when you go out to a 5 or even 10 game stretch, chances are it's going to be either above or below average. In some cases (and I'm guessing probably much less rare than people would think) it can be well above or below. Even when you look at full season stats for a goalie it's fairly common to have some significant variation year to year. But when you get a good sample size, like 100+ games, it should give you a good idea about what a goalie is.

Mrazek may or may not be prone to unusually long "down" periods and conversely, unusually long "up" periods, but it's more likely that it's just a fluke. In the future, we should probably expect a more normal distribution of both. But regardless, based on his career to date, we should expect a more high than low, with all of it adding up to fairly average.

Of course variation is a part of what makes an average, but the level of Mrazek's variation is higher than most, and doing a random comparison of some other goalies his variation was higher than all.  I just pulled five other goalie's career game logs from nhl.com and compared the deviation from mean of save percentage to Mrazek's.  Goalies were selected semirandomly, with the only criteria being that I selected a goalie roughly every 10 spots on the current save percentage ranking all the way down to where Mrazek is so I wasn't cherry picking either the top goalies or the scrubs, and I wanted to look at some goalies who had more games played than Mrazek, and others who had less or similar games played.  I also wanted to have at least a couple of the goalies around Mrazek's age too. I copied their career game logs into excel, used average and standard deviation equations (deviation from mean shown below) on save percentage and compared the results.  I also added Howard in there for s***s and giggles.  I know how much you like comparing him and Mrazek so I thought you'd like to see his data too   But to be fair if you looked at the last three years of Howard's data his numbers would be much higher than what his career is.

Mrazek: .084
Holtby: .073
Lehner: .060
Vasilevskiy: .059
Allen: .083
Pickard: .070
Howard: .066

Jake Allen was damn close so at the very least what Mrazek is doing isn't revolutionary, but at least from the goalies I looked at, the data with Mrazek and where he ends up on the spectrum of deviation from the mean is consistent with what I'd expect it to be, which shouldn't be a surprise to anyone who has watched him play the past few years.

 

edit: I actually wanted to see what Howard's deviation from his career average is between now and through the 2013-14 season.  Its .074

Edited January 1, 2017 by Echolalia

[Buppy 1,720]

2 hours ago, Echolalia said:

Of course variation is a part of what makes an average, but the level of Mrazek's variation is higher than most, and doing a random comparison of some other goalies his variation was higher than all.  I just pulled five other goalie's career game logs from nhl.com and compared the deviation from mean of save percentage to Mrazek's.  Goalies were selected semirandomly, with the only criteria being that I selected a goalie roughly every 10 spots on the current save percentage ranking all the way down to where Mrazek is so I wasn't cherry picking either the top goalies or the scrubs, and I wanted to look at some goalies who had more games played than Mrazek, and others who had less or similar games played.  I also wanted to have at least a couple of the goalies around Mrazek's age too. I copied their career game logs into excel, used average and standard deviation equations (deviation from mean shown below) on save percentage and compared the results.  I also added Howard in there for s***s and giggles.  I know how much you like comparing him and Mrazek so I thought you'd like to see his data too   But to be fair if you looked at the last three years of Howard's data his numbers would be much higher than what his career is.

Mrazek: .084
Holtby: .073
Lehner: .060
Vasilevskiy: .059
Allen: .083
Pickard: .070
Howard: .066

Jake Allen was damn close so at the very least what Mrazek is doing isn't revolutionary, but at least from the goalies I looked at, the data with Mrazek and where he ends up on the spectrum of deviation from the mean is consistent with what I'd expect it to be, which shouldn't be a surprise to anyone who has watched him play the past few years.

 

edit: I actually wanted to see what Howard's deviation from his career average is between now and through the 2013-14 season.  Its .074

And none of that address the question of why we need to keep Howard.

Secondly, standard deviation doesn't really mean anything. A large percentage of any goalies stdev is going to come from a small number of games that are well below the norm (below because you can't go much higher). So if Mrazek is in fact prone to 1 or 2 extra complete s*** games in a full season, so what?

>95%	32.00%	38.53%
92-95%	20.80%	18.18%
89-92%	20.00%	14.72%
85-89%	10.40%	15.58%
<85%	16.80%	12.99%

The above is two goalies, 125+ games each, and the percentage of games in each save% range. What really is the difference there?

 

 

[Echolalia 2,961]

37 minutes ago, Buppy said:

And none of that address the question of why we need to keep Howard.

Secondly, standard deviation doesn't really mean anything. A large percentage of any goalies stdev is going to come from a small number of games that are well below the norm (below because you can't go much higher). So if Mrazek is in fact prone to 1 or 2 extra complete s*** games in a full season, so what?

>95%	32.00%	38.53%
92-95%	20.80%	18.18%
89-92%	20.00%	14.72%
85-89%	10.40%	15.58%
<85%	16.80%	12.99%

The above is two goalies, 125+ games each, and the percentage of games in each save% range. What really is the difference there?

 

 

Who said anything about keeping Howard?  Jesus Buppy I don't know what the deal is but every single post you make brings up a point irrelevant to the discussion or assumes that I take a stance on something that I never said.  Also, standard deviation means everything when talking about consistency.  By definition it is the measurement of consistency to the mean, and thus is the ideal tool to use when asking whether a goalie's play is consistent relative to how his peers play.  And while your point about deviation being more heavily influenced by low values than high because there are more possible values below the average than above it, it doesn't matter when comparing to other players because everyone's numbers are affected similarly because the same equation is being used across the board, and that's ultimately what I'm interested in.  Lumping the data sets as you did into broader categories is a less precise and way more arbitrary way of looking at the data.  You're consolidating the entire spectrum of save percentage into 5 points, whose margins could have been made a ton of different ways (and thus different results when you plug in the same raw data).  You're losing lots of information doing that.

[Buppy 1,720]

10 hours ago, Echolalia said:

Who said anything about keeping Howard?  Jesus Buppy I don't know what the deal is but every single post you make brings up a point irrelevant to the discussion or assumes that I take a stance on something that I never said.  Also, standard deviation means everything when talking about consistency.  By definition it is the measurement of consistency to the mean, and thus is the ideal tool to use when asking whether a goalie's play is consistent relative to how his peers play.  And while your point about deviation being more heavily influenced by low values than high because there are more possible values below the average than above it, it doesn't matter when comparing to other players because everyone's numbers are affected similarly because the same equation is being used across the board, and that's ultimately what I'm interested in.  Lumping the data sets as you did into broader categories is a less precise and way more arbitrary way of looking at the data.  You're consolidating the entire spectrum of save percentage into 5 points, whose margins could have been made a ton of different ways (and thus different results when you plug in the same raw data).  You're losing lots of information doing that.

Seriously? Ok, keep Howard or replace him with an established #1. Feel better? Do I really need to type out the qualifier every time? Because that has been the argument since the beginning.

" you better trade for an established number 1 goaltender to fill his spot "

" But if you're gung-ho about getting rid of Howard, you best trade for an established number 1 NHL calibre goalie "

"  if Howard is traded away, Holland better find another goalie to work with Mrazek "

I've said all along I don't care about Mrazek's consistency, because Howard isn't any better. Replacing Howard with an average backup doesn't make us any worse. You keep wanting to make it about Mrazek's consistency, despite me saying numerous times it's a different argument.

In regards to that argument:

Let's say you have two goalies, both played exactly 231 games. 4 years as a regular starter basically. One has a stdev of .07227, the other is .05752. What does that difference mean in real terms? How does that help you decide which is better? What kind of different performance would you expect from one to the next? Can you glean anything remotely useful from that?

The reason I lumped the save%s into ranges is because it gives useful information on what you might expect.

To post a save%of .950 or higher means allowing one goal or less most of the time. Two if you face 40+ shots. In either case, your team should have an exceptional chance to win the game. .920 but below .950 is mostly going to be two goals, and again a good chance to win. Excellent chance if you're keeping the shots low, but getting risky if allowing a lot. .89-.92 you need to either keep the shots against low, or do a lot of scoring of your own. Most likely a game that could go either way for most teams. .85-89 you really need the rest of the team to bail you out. Usually going to be 3 or 4 goals, fair chance for an OT point if the shots are lowish and the offense is good. Below that is either giving up 4+ goals or getting pulled early, either way probably a loss. 

If you want more data:

94.00%	2.4%	6.1%
93.00%	8.0%	6.9%
92.00%	10.4%	5.2%
91.00%	5.6%	5.2%
90.00%	8.8%	6.1%
89.00%	5.6%	3.5%
88.00%	2.4%	6.1%
87.00%	3.2%	4.8%
86.00%	1.6%	2.6%
85.00%	3.2%	2.2%

Edited January 2, 2017 by Buppy

[DickieDunn 2,571]

Keep both of them and hope one of them is always on a hot streak.

[kliq 3,755]

35 minutes ago, DickieDunn said:

Keep both of them and hope one of them is always on a hot streak.

There is something to this. Putting all your eggs in one basket can really hurt you, just look at Montreal last season when Price went down, or a few years ago in the semi's when Price went down.

[Buppy 1,720]

3 hours ago, DickieDunn said:

Keep both of them and hope one of them is always on a hot streak.

And that's a lot of cap going toward your goalies, and no spot for someone who might be even better than either. All for the sake of maybe a few more points during a rebuild? 

[Echolalia 2,961]

13 hours ago, Buppy said:

Seriously? Ok, keep Howard or replace him with an established #1. Feel better? Do I really need to type out the qualifier every time? Because that has been the argument since the beginning.

" you better trade for an established number 1 goaltender to fill his spot "

" But if you're gung-ho about getting rid of Howard, you best trade for an established number 1 NHL calibre goalie "

"  if Howard is traded away, Holland better find another goalie to work with Mrazek "

I've said all along I don't care about Mrazek's consistency, because Howard isn't any better. Replacing Howard with an average backup doesn't make us any worse. You keep wanting to make it about Mrazek's consistency, despite me saying numerous times it's a different argument.

In regards to that argument:

Let's say you have two goalies, both played exactly 231 games. 4 years as a regular starter basically. One has a stdev of .07227, the other is .05752. What does that difference mean in real terms? How does that help you decide which is better? What kind of different performance would you expect from one to the next? Can you glean anything remotely useful from that?

The reason I lumped the save%s into ranges is because it gives useful information on what you might expect.

To post a save%of .950 or higher means allowing one goal or less most of the time. Two if you face 40+ shots. In either case, your team should have an exceptional chance to win the game. .920 but below .950 is mostly going to be two goals, and again a good chance to win. Excellent chance if you're keeping the shots low, but getting risky if allowing a lot. .89-.92 you need to either keep the shots against low, or do a lot of scoring of your own. Most likely a game that could go either way for most teams. .85-89 you really need the rest of the team to bail you out. Usually going to be 3 or 4 goals, fair chance for an OT point if the shots are lowish and the offense is good. Below that is either giving up 4+ goals or getting pulled early, either way probably a loss. 

If you want more data:

94.00%	2.4%	6.1%
93.00%	8.0%	6.9%
92.00%	10.4%	5.2%
91.00%	5.6%	5.2%
90.00%	8.8%	6.1%
89.00%	5.6%	3.5%
88.00%	2.4%	6.1%
87.00%	3.2%	4.8%
86.00%	1.6%	2.6%
85.00%	3.2%	2.2%

Because a Mrazek/Coreau combo would have made this season, and quite possibly  subsequent seasons brutally unwatchable which defeats the entire purpose of NHL hockey games, although I know you mentioned that's more or less the type of hockey you would prefer to see.  And again your table is still dropping data, and when measuring for variance it's altogether inferior to sd.  On a side note, it's also misleading to say that the data below .85 or whatever arbitrary point doesn't matter.  Your claim that a save percentage less than .85 can be written off as a loss is exaggerating.  Yeah the chances go down as your percentage drops, but just looking at Mrazek's sub .850 games this year, 29% of them were wins.  They account for almost a quarter of all the wins Mrazek has had this year. Someone could get pulled after allowing two goals on three shots for a save percentage of .333 and it's not remotely farfetched that the team could overcome a two-goal deficit and win the game.  It's not data worth writing off, and considering the past few seasons where one single win was the difference between playoffs and not, I wouldn't be content with tossing that data away and labeling it the same.  And when it comes to simply measuring variation from the mean to calculate consistency as was my original purpose, your method is overall subpar.

Edited January 2, 2017 by Echolalia

[DickieDunn 2,571]

10 hours ago, Buppy said:

And that's a lot of cap going toward your goalies, and no spot for someone who might be even better than either. All for the sake of maybe a few more points during a rebuild? 

According to KFH they're not rebuilding