WCR1: Edmonton Oilers (P3) vs Los Angeles Kings (P2) - Oilers win 4-2

[midniteowl]

Originally Posted by stemit14

This. Kings are outplaying the oilers by a wide margin but all it takes is the oilers getting 4 powerplays in period 1 of game 3 from whining to the league and from diving. Really going the league does not fall for it this time.

You mean like this?

[NegativeSpace]

Originally Posted by Enoch Root

This goal, and about 3 others last night, are a great example of what I mean when I say xG is a terribly flawed stat. I have tried to explain, over and over, how the stat can't measure time and space, and as a result, does not accurately reflect the real danger of actual chances.

In I think 4 of the goals last night, the shooter was in a location that was somewhat dangerous, and the 'expected' outcome would be maybe 20% chance of a goal. However, the stat can't measure time and space, and in each case, the shooter was able to walk in with his head up, look at Skinner, pick a spot, and get a perfect shot off. The ACTUAL likelihood of a goal, in each case was MUCH higher, because of how much time and space they had. But the stat can't measure that.

So the result (according to the stat) will be that the Kings got more goals than expected, and Skinner was terrible. But in reality what happened was that the Oilers defense gave 4 guys ALL DAY to walk in and pick a corner. And when you do that, today's shooters will score almost every time.

Not to get to legalistic, but I think that "terribly flawed" is too harsh. The stat has flaws - as you right point out. It is not a great measure in any individual game because there are deficiencies. I also think that mixing the stat of expected goals with actual goals in one game or another has trouble.


However, it is a useful way to measure longer term trends with teams and players. As Bingo said to me in our discussion, the stat reflects whether more good or bad things happen on the ice with a player there. That remains a useful metric.

[NegativeSpace]

Originally Posted by Paulie Walnuts

Yes. Remember what happened in the Finals.

Just keep coming at them and put evertyhing on Skinner. They are going point blank chances and LA has some deadly shooters. Those guys like Kempe, Fiala, Byfield and Kopitar with time and space will rifle it home all day.

I agree that LA needs to win one in Edmonton. However, many people have been pointing out that this Edmonton team is very different than last year. Last year's team could play decent defence. So far, this Edmonton team has no answer to LA rolling lines.

[Enoch Root]

Originally Posted by NegativeSpace

Not to get to legalistic, but I think that "terribly flawed" is too harsh. The stat has flaws - as you right point out. It is not a great measure in any individual game because there are deficiencies. I also think that mixing the stat of expected goals with actual goals in one game or another has trouble.


However, it is a useful way to measure longer term trends with teams and players. As Bingo said to me in our discussion, the stat reflects whether more good or bad things happen on the ice with a player there. That remains a useful metric.

Yes, the problems are dampened with larger data sets. Unfortunately the stat is rarely used for large data sets, it is almost invariably used for single-game analysis where it is, IMO, worse than using nothing. Better to simply rely on observation and not let a wildly misleading stat cloud that observation.

[Paulie Walnuts]

Originally Posted by NegativeSpace

I agree that LA needs to win one in Edmonton. However, many people have been pointing out that this Edmonton team is very different than last year. Last year's team could play decent defence. So far, this Edmonton team has no answer to LA rolling lines.

Don't want to get too excited just yet great start but as Kobe said job done? jobs not done.

[Coach]

Kempe is ice cold man, love it.

[Rejean31]

Originally Posted by Lewis_D

Is this an actual quote lol? Didn't have Stauffer outing himself as a Karen on my bingo card this series but I'm not surprised.

A better question would be what Grease Pigs media mouthpiece isn't a Karen??

[GioforPM]

Originally Posted by Coach

Kempe is ice cold man, love it.

I was confused, it seemed like you were saying his offence dried up. But I guess you are saying he's like a cold blooded killer.

I'm old so I doubletake at a lot of meaning changes. When did "selling out" start to mean putting your body on the line for a block, instead of betraying someone?

[Bingo]

Originally Posted by Enoch Root

Yes, the problems are dampened with larger data sets. Unfortunately the stat is rarely used for large data sets, it is almost invariably used for single-game analysis where it is, IMO, worse than using nothing. Better to simply rely on observation and not let a wildly misleading stat cloud that observation.

I think huge conclusions on one game are a mistake.

But the stats are the stats.

The system is designed to count things happening both good and bad and then those occurrences are applied to averages for goal results.

Player has more good things happen than bad is always going to have a better game than a player having more bad things than good. A goalie can make a save and change the actual outcome, but that doesn't change the underlying performance.

[Barnet Flame]

Originally Posted by CroFlames

Agreed with most of what has been said about the game.

The media though. It's disgusting how they are just Oiler fans calling a broadcast with no impartiality at all. By contrast, you can tell Ryan Leslie, for example, is great buddies with most/all the Flames players, but you don't see him whining on national TV how bad the Flames are being treated, and that they need to talk to the manager lol. Ryan is all pro all the time. Hell, even Pike who is clearly a fan keeps things professional when he's working.

Principe, Stauffer, DeBrusk, all of em, trash.

We are incredibly lucky to have a class act like Ryan Leslie. In fact, despite all the whinging we’ve done about Sportsnet, we have been blessed with excellent hosting, commentary and colour over the past few years. Ball was a gem and I’m really warming to Jon Abbot. Though I will really miss Greg Millen.

[GioforPM]

Was at a friend's place, sound off, and Principe appeared on the screen. The wives both gasped "who the hell is that? Is that his real hair"?

Pretty sure I've pointed him out before but they forgot.

[PaperBagger'14]

Originally Posted by bc-chris

IMG_8027 by bc-chris, on Flickr

I just wanted to see it again

[Erick Estrada]

Originally Posted by GioforPM

Was at a friend's place, sound off, and Principe appeared on the screen. The wives both gasped "who the hell is that? Is that his real hair"?

Pretty sure I've pointed him out before but they forgot.

I can only surmise that his wife likes the look of frizzled out hair and leather skin. I hope that guy gets himself checked out frequently at skin cancer clinics as it looks like he lives under a UV lamp.

[Rollin22x]

Originally Posted by Coach

Kempe is ice cold man, love it.

The Raz death stare but Kempe is aimed at the Oiler players. "I dare you too look me in the eyes!"

[Enoch Root]

Originally Posted by Bingo

I think huge conclusions on one game are a mistake.

But the stats are the stats.

The system is designed to count things happening both good and bad and then those occurrences are applied to averages for goal results.

Player has more good things happen than bad is always going to have a better game than a player having more bad things than good. A goalie can make a save and change the actual outcome, but that doesn't change the underlying performance.

Again, I think you give it too much credit. There are only a few events that generate stats, so saying more good than bad is not accurate, because most of what they do during the game is not being counted. Instead, only a few things are counted, and those things can be very different than their numbers, due to time and space - not all shots from the same location are created equal.

Very few data points + inaccurate awarding of points = very skewed and misleading results.

We see it almost every night where the xGF and xGA are wildly different than what is expected from observation.

[Enoch Root]

Here's the thing about observation vs stats...

In small sample sizes, observation is extremely accurate, but as sample sizes get larger, memory gets skewed, and biases have greater and greater impact.

Whereas data is the opposite: small sample sizes are effectively useless and skewed, but as you get more and more data, it becomes more and more accurate and useful.

[Paulie Walnuts]

the xGSA has Skinner as last and Kuemper as 2nd last.

[Bingo]

Originally Posted by Enoch Root

Again, I think you give it too much credit. There are only a few events that generate stats, so saying more good than bad is not accurate, because most of what they do during the game is not being counted. Instead, only a few things are counted, and those things can be very different than their numbers, due to time and space - not all shots from the same location are created equal.

Very few data points + inaccurate awarding of points = very skewed and misleading results.

We see it almost every night where the xGF and xGA are wildly different than what is expected from observation.

Everything inside the offensive zone is counted. Not a few things.

An outside shot is low danger (a corsi event that doesn't register as a scoring chance or high danger), and then shot attempts inside the home plate as a scoring chance, ... same chance elevates to a high danger if it comes on a tip, rebound or direct pass.

They're all counted.

Over the course of time each type of shot is graded by chance of going in and then builds expected goals.

So in a game a player that is on the ice when the opposition is able to shoot more from in close, and on dangerous measures (tips, rebounds, wide open passes) then his xGA goes up and it generally matches the eye test.

What shouldn't be done is focusing on xGF% in an individual game because players are deployed differently.

If Kevin Rooney has the highest xGA60 on the team in a game he has a problem because his job is to make sure nothing bad happens. But he may generate little offensively (not his job), and be 1st in xGA60 but still at the bottom in xGF% because he doesn't generate much.

I'm all for the evaluation to provide more shades of gray (more disparity in events so a tighter expected goals answer), but for now it's the best way of assessing a player as eye test is often correct, but almost always subjective.

[Bingo]

Originally Posted by Enoch Root

Here's the thing about observation vs stats...

In small sample sizes, observation is extremely accurate, but as sample sizes get larger, memory gets skewed, and biases have greater and greater impact.

Whereas data is the opposite: small sample sizes are effectively useless and skewed, but as you get more and more data, it becomes more and more accurate and useful.

In assessing player X overall ... agreed.

But in assessing player X in game 18 ... disagree.

[Lewis_D]

Originally Posted by GioforPM

Was at a friend's place, sound off, and Principe appeared on the screen. The wives both gasped "who the hell is that? Is that his real hair"?

Pretty sure I've pointed him out before but they forgot.

Gene Principe has poodle hair. I always think of that Tim Robinson skit when I see him.