• Moderators, please send me a PM if you are unable to access mod permissions. Thanks, Habsy.

Red Wings @ Leafs - 7:00 PM - CBC

leaffan2005 said:
I didn't invent anything. Not sure what you are on about here. Best guess is you don't care to acknowledge a very readily available statistic and it's significance because it doesn't jive with your bias against Anderson.
Click to expand...

you decided to post a stat that you've never used before to attack "stats nerds" and tbeir hypocrisy.
 
Pretty much all the ex-goalies that have become any part of the analyst teams try to shove the narrative of wanting to get into the game with lots of shots down our throat. Can we trust them on that? Maybe not but, if you are creating more high danger shots than your opposition who cares about the outside shots. Theses should be saved by most goalies anyways.
 
leaffan2005 said:
I didn't invent anything. Not sure what you are on about here. Best guess is you don't care to acknowledge a very readily available statistic and it's significance because it doesn't jive with your bias against Anderson.
Click to expand...
The point is that these counting stats are meaningless if there's no demonstrable correlation to success or failure. It's just a number in a vacuum.
 
Neurospasm said:
Pretty much all the ex-goalies that have become any part of the analyst teams try to shove the narrative of wanting to get into the game with lots of shots down our throat. Can we trust them on that? Maybe not but, if you are creating more high danger shots than your opposition who cares about the outside shots. Theses should be saved by most goalies anyways.
Click to expand...

Yep.

Don't get me wrong, allowing that additional opposition puck possession that leads to shots from the outside areas of the ice has some value, but it's minimal. When you look at how high event a team the Leafs are, and how much of the possession against turns into high danger chances, the Leafs are a pretty average defensive group overall, and have gotten better as the year has gone on (with a pretty massive jump to being legitimately very good defensively in March).

What aggravates me is when people who pretty bluntly dismiss better statistics, turn around and use poor statistics to try to make a point. SOG is a shit statistic. It belongs in a bin with +/-, blocked shots, hits, etc
 
nobody ever mentions all the shots the lightning give up.

heck, the uber-defense preds give up plenty too.
 
It is such a strange game, because at the end of the game, I thought Howard was standing on his head to keep his team in the game, we dominated the game except for extended periods in the 2nd period, and Freddie had a so-so game.

So I was VERY surprised to see Andersen faced 41 shots last night, and had a .926SV%.

I guess it's where shots can be a little deceiving. The Wings fired a lot of low % shots whereas we dominated offense all night long, but some chances didn't even result in SOGs.

It's good to keep the winning streaking going, but at the end of the day these wins mean so little. It is still important we start the playoff on a high rather than a low.

I really like what Johnsson brings to the team, the 4th line is SO good, and even Pleks is playing better and better every game imo. I've got a feeling he'll start chipping in offensively soon.

OTOH I think Hainsey is really looking so tired and so slow. He can really use some rest, more than any other player on the team.
 
It seems like every other sport, rest is a legitimate reason to bench a player, but not hockey, when was the last time a top team rested anyone late in the year before the playoffs? I hope the Leafs do it with some of their players but based on hockey culture I doubt it.
 
MindzEye said:
Yep.

Don't get me wrong, allowing that additional opposition puck possession that leads to shots from the outside areas of the ice has some value, but it's minimal. When you look at how high event a team the Leafs are, and how much of the possession against turns into high danger chances, the Leafs are a pretty average defensive group overall, and have gotten better as the year has gone on (with a pretty massive jump to being legitimately very good defensively in March).

What aggravates me is when people who pretty bluntly dismiss better statistics, turn around and use poor statistics to try to make a point. SOG is a shit statistic. It belongs in a bin with +/-, blocked shots, hits, etc
Click to expand...

You dismiss them as superficial, but they’re the absolutes.

You want to peel down a few more layers, fine, but that doesn’t make the hard stats worthless.

If anything, I think the added analysis you do often tends to dismiss the absolutes, and then stretches to conclusions that aren’t necessarily borne out in reality. I still think the possession statistics are pretty ridiculous.
 
why do you think they are ridiculous?

they seem pretty simple and straightforward to me.
 
LeafOfFaith said:
You dismiss them as superficial, but they’re the absolutes.

You want to peel down a few more layers, fine, but that doesn’t make the hard stats worthless.
Click to expand...

Why do the "absolutes" have value? Are they predictive of anything? If they are predictive, are they more predictive than other methods?

If anything, I think the added analysis you do often tends to dismiss the absolutes, and then stretches to conclusions that aren’t necessarily borne out in reality. I still think the possession statistics are pretty ridiculous.
Click to expand...

The issue then is your lack of knowledge about the predictive value of the better analytics. You think that possession statistics (I'm using that term as a broad umbrella for the various shot based and goal based metrics we commonly use here) are ridiculous, but they're highly predictive of team success. Far more predictive than any other publicly available method of analysis.

So if we're given only a few different data points about a hockey game that was played, shots for/against, corsi for/against, and expected goals for/against...which is the best predictive tool for determining who was most likely to win that game? The answer is that their predictive ability is in the reverse order that I listed them, with expected goals being a very good predictive tool for guessing who won the game, corsi for/against being an okay predictive tool, and shots for/against providing absolutely minimal predictive value.

So if a certain statistic has a poor predictive value in determining who is winning hockey games, what gives it value? It just becomes a stand alone number that actually gives you the wrong idea about who the better team was on a given night. Shots, hits, blocks absolutely belong in that category. If you used any of those to guess on who won a hockey game, you would be worse than average at guessing the winner of a hockey game. If you use Corsi, or better yet expected goals for/against you would be better than average at guessing the winner of a hockey game. There's been a huge pile of work done to prove this.

The most powerful argument for those who don't get, or don't want to get the math involved though is that the people who have developed these methods that you find "pretty ridiculous" are all working for NHL teams in their analytics departments today where they were just random dudes (most with some sort of academic statistics or mathematics based backgrounds) with a website 4-5 years ago. NHL teams aren't hiring these guys because their ideas are "pretty ridiculous".
 
zeke said:
why do you think they are ridiculous?

they seem pretty simple and straightforward to me.
Click to expand...

Yeah, it's basically: Team A took this many shots from these specific areas on the ice. Those areas one the ice have expected SV% of "X" (with the SV%'s calculated using the real world location of every shot taken in every NHL game since they've started tracking them like this) and thus team A should scored "Y" amount of goals. Do the same for the opposition and you have, with a fairly high degree of accuracy a predictive model for figuring out who should have won that game.

You can't look at raw SOG numbers and even be on the same planet for predictive accuracy. There are a number of circumstances where raw SOG would actually have an inverse affect on predictive value (score effects)
 
Alright, Anderson being on pace to face a historic number of shots this year means nothing, and shouldn't be included in any kind of honest evaluation of his value and play this year.

Gotcha.
 
zeke said:
you decided to post a stat that you've never used before to attack "stats nerds" and tbeir hypocrisy.
Click to expand...

I used a stat to show how unusual a workload that Anderson has faced this year which makes his numbers and level of play that much more impressive.

But it doesn't fit a zeke narrative and it is thus invalid.
 
enh, i think looking at SOG last night is pretty misleading. the leafs had a tough time breaking the neutral zone in the 2nd, but i can think of far more flubbed chances by the leafs, and difficult stops by howard than vice versa.

e.g., the dermott post, the marner:matthews feed (matthews shot wide on an open net), kapanen x 2 (missed the net on an open net in the second, flubbed that backhand on the three on one in the third), the nylander breakaway... then the ridiculous stops by howard on kadri on the pp, and on johnsson.

on the other hand, andersson made a good stop on the penalty shot, and made one or two post to posts. most of the wings shots were from a stand-still positions, and didn't force andersson to go post to post much at all (most of their shots were, oddly, aimed to beat freddy in the top corners, where he is usually tough to beat). i can recall one play in the third where the wings had a good chance and missed the net on a tip in following a cross crease pass.
 
They both got vezina talk when up around .930sv%.

and they both didn't get vezina talk when around .920sv%.
 
zeke said:
They both got vezina talk when up around .930sv%.

and they both didn't get vezina talk when around .920sv%.
Click to expand...

Those two haven't come close to facing as much rubber as Anderson.

And they have continued to get Vezina talk all season, in spite of slumps. While here..

Also to address your comment above, the Bruins have the same amount of points over the last 32 games as us.
 
Vas was getting getting vezina talk when he was leading the league with a ~.930sv% the first half. He's not anymore now that he's down closer to .920.

Rinne wasn't gettijg any vezina talk when he was around .920 in the first half. he is now getting vezina talk as he's leading the league at around .930.

and good we're agreed that the leafs have been as good as the 2 best and hottest teams in the league since dermott joined the lineup.
 
You must log in or register to reply here.
Back
Top