Not to worry; this won't last says Myrtle

[Enoch Root]

How about a quick simple example to illustrate the point?:

Wrong (blogger) way to do it...

Item: possession stats are fairly strongly correlated with winning
Item: Flames have a good record but weak possession stats

Conclusion: Flames record will regress to match their possession stats

Proper way to do it...

Item:
Item:

Questions: those are potentially incongruous, what else is going on here? what else do we know? how significant is this data? confidence levels? do other stats support the incongruity or refute it?

etc

[Resolute 14]

Not conclusion, so much as prediction, based on history and probabilities.

And man, you guys are really butthurt that some people don't think we can sustain this, aren't you?

[Enoch Root]

Originally Posted by Tinordi

Baffling statement

Stats guys just run a bunch of r squared values and call it a day?

All statistical work whether at NASA or the NHL is conducted to draw conclusions. Otherwise it's a meaningless activity.

And that was hilarious because no one at NASA has ever drawn a conclusion from a single statistic.

[Enoch Root]

Originally Posted by Resolute 14

Not conclusion, so much as prediction.

And man, you guys are really butthurt that some people don't think we can sustain this, aren't you?

Not at all. I expect that they won't sustain their record once the youthful push runs its course.

But as someone who has been analyzing stats and overseeing others do it (usually poorly) for decades, I have little tolerance for people who are too confident with incomplete information.

Especially those who call people stupid or ignorant for not accepting their conclusions.

(Edit: not a reference to you, btw)

[driveway]

It's worth noting that our fancy-stats are improving.

In our last 6 games, our 5v5 Fenwick Close is 49.53%, the actual +/- of events is only a -2. We've been playing teams even in posession during the important times of the game for the last 6 games.

Over the previous 6 games, in the same situation, our Fenwick was 47.74% with an actual +/- of -13.

Over our first 10 games of the season, our 5v5 Fenwick Close was 43.48% with an actual +/- of -45.

While a cumulative look at our posession stats would probably indicate a lack of sustainability, a more granular look shows overall improvement in the numbers, which gives me hope that we can continue to have the kind of success we've been having. Though, very simply, if we keep spotting 2-goal leads, we're going to lose more than we win.

[Street Pharmacist]

Originally Posted by Enoch Root

And that was hilarious because no one at NASA has ever drawn a conclusion from a single statistic.

And that's hilarious because you're purposefully mischaracterizing what he said.

[Zevo]

What's hilarious is that any fans would even think about fenwick ,corsi , or any other random advanced stat. Only stat I need? Am I enjoying watching my team play? If yes then that's all I need.

Coaches and management, sure they probably are a good tool, but as a fan? Who gives a #### if your team is winning or at least heading in the right direction.

Advanced stats are for fans of no good teams that want to feel better about their no good team.

[Where ru Chris O'Sullivan]

Originally Posted by Zevo

What's hilarious is that any fans would even think about fenwick ,corsi , or any other random advanced stat. Only stat I need? Am I enjoying watching my team play? If yes then that's all I need.

Coaches and management, sure they probably are a good tool, but as a fan? Who gives a #### if your team is winning or at least heading in the right direction.

Advanced stats are for fans of no good teams that want to feel better about their no good team.

The other thing is confidence, I'm waiting for Mike Corsi or Jim Fenwick to expand their data collection to this next, should be a winner for sure.

The confidence on this team right now is unbelievable, shaky/rusty goalie, gio + brodie having a bad game? No problem, we will still score 2 goals late in the game to send it to OT and a shootout win.

What if these young guys gain more confidence, then what? Right now they seem to like to feel the opposing team out for 2 periods and tend to make a statement in the 3rd. What if they start playing the way they do in the 3rd for the entire game?

[driveway]

Originally Posted by Zevo

What's hilarious is that any fans would even think about fenwick ,corsi , or any other random advanced stat. Only stat I need? Am I enjoying watching my team play? If yes then that's all I need.

Coaches and management, sure they probably are a good tool, but as a fan? Who gives a #### if your team is winning or at least heading in the right direction.

Advanced stats are for fans of no good teams that want to feel better about their no good team.

Or, on the other hand, people who like hockey want to understand it better. But, you can be a penis about it if you want to.

[Vulcan]

Originally Posted by Where ru Chris O'Sullivan

The other thing is confidence, I'm waiting for Mike Corsi or Jim Fenwick to expand their data collection to this next, should be a winner for sure.

The confidence on this team right now is unbelievable, shaky/rusty goalie, gio + brodie having a bad game? No problem, we will still score 2 goals late in the game to send it to OT and a shootout win.

What if these young guys gain more confidence, then what? Right now they seem to like to feel the opposing team out for 2 periods and tend to make a statement in the 3rd. What if they start playing the way they do in the 3rd for the entire game?

I think a lot of our 3rd period success is due to our hard work rate which wears the other team down and they just aren't in as good condition as we are.

It's like what I've been told happens in a long horse race. The fast finishers aren't really running faster, they're just maintaining their pace while other horses drop off.

However it works it's got to be frustrating for the losing team and we are building a winning attitude.

[TheDebaser]

Originally Posted by Zevo

What's hilarious is that any fans would even think about fenwick ,corsi , or any other random advanced stat. Only stat I need? Am I enjoying watching my team play? If yes then that's all I need.

Coaches and management, sure they probably are a good tool, but as a fan? Who gives a #### if your team is winning or at least heading in the right direction.

Advanced stats are for fans of no good teams that want to feel better about their no good team.

As a fan I can follow my team however I like.

What's really baffling to me is why someone who doesn't want to hear about advanced stats is reading and posting in a thread about advanced stats.

[Street Pharmacist]

Originally Posted by Zevo

What's hilarious is that any fans would even think about fenwick ,corsi , or any other random advanced stat. Only stat I need? Am I enjoying watching my team play? If yes then that's all I need.

Coaches and management, sure they probably are a good tool, but as a fan? Who gives a #### if your team is winning or at least heading in the right direction.

Advanced stats are for fans of no good teams that want to feel better about their no good team.

Taking about contracts, trades, GMs, coaches and players are for fans of no good teams too! Who cares what the lineups are! Just enjoy player A and B winning!

[Lanny_McDonald]

All you need to know about advanced stats is this.

Edmonton Oilers Corsi (for) Rank: 10th - Calgary 19th
Edmonton Oilers Corsi (against) Rank: 19th - Calgary 29th
Edmonton Oilers Fenwick (for) Rank: 10th - Calgary 21st
Edmonton Oilers Fenwick (against) Rank: 22nd - Calgary 26th
Edmonton Oilers Fenwick (for %) Rank: 17th - Calgary 27th

Advanced stats are no good. Seriously. On what planet is Edmonton a better team than Calgary? I mean, watch the hockey games! Advanced stats are all the rage in Shelbyville because they tell a misleading story. Here are the only stats that matter.

Calgary Flames: 13W-7L-2OTL - 28P (8th in NHL)
Edmonton Oilers: 6W-13L-2OTL - 14P (tied for dead ####ing last in the NHL)

Calgary Flames: +10 goal differential (8th in NHL)
Edmonton Oilers: -25 goal differential (29th in the NHL)

Not very advanced or very fancy, but real good indicators of success.

[Zevo]

Wasn't really trying to be a dink. I enjoy looking at stats and seeing how players are doing as much as the next guy, but advanced stats should really mean nothing, as a fan. If you like looking into them, great I guess, each to their own.

If your team is doing well and playing entertaining hockey, what does it matter what a players corsi or fenwick is. It just seems that people, as fans that use these stats when teams are rolling, are looking for a dark cloud on a sunny day. Conversely, if your team sucks and aren't at all enjoyable to watch, do you really care that player A has great advanced stats? In the case of Edmonton, it shows a team can have okay advanced stats if the other team scores quickly and then goes on cruise control.

[Where ru Chris O'Sullivan]

Originally Posted by New Era

All you need to know about advanced stats is this.

Edmonton Oilers Corsi (for) Rank: 10th - Calgary 19th
Edmonton Oilers Corsi (against) Rank: 19th - Calgary 29th
Edmonton Oilers Fenwick (for) Rank: 10th - Calgary 21st
Edmonton Oilers Fenwick (against) Rank: 22nd - Calgary 26th
Edmonton Oilers Fenwick (for %) Rank: 17th - Calgary 27th

Advanced stats are no good. Seriously. On what planet is Edmonton a better team than Calgary? I mean, watch the hockey games! Advanced stats are all the rage in Shelbyville because they tell a misleading story. Here are the only stats that matter.

Calgary Flames: 13W-7L-2OTL - 28P (8th in NHL)
Edmonton Oilers: 6W-13L-2OTL - 14P (tied for dead ####ing last in the NHL)

Calgary Flames: +10 goal differential (8th in NHL)
Edmonton Oilers: -25 goal differential (29th in the NHL)

Not very advanced or very fancy, but real good indicators of success.

Thanked for addition in my signature.

[Street Pharmacist]

Originally Posted by Zevo

Wasn't really trying to be a dink. I enjoy looking at stats and seeing how players are doing as much as the next guy, but advanced stats should really mean nothing as a fan. If you like looking into them, great I guess, each to their own.

If your team is doing well and playing entertaining hockey, what does it matter what a players corsi or fenwick is. It just seems that people, as fans that use these stats when teams are rolling, are looking for a dark cloud on a sunny day. Conversely, if your team sucks and aren't at all enjoyable to watch, do you really care that player a has great advanced stats? In the case of Edmonton, it shows a team can have okay advanced stats if the other team scores quickly and then goes on cruise control.

It means the same thing as seeing where Johnny is in the rookie scoring race, or seeing if the flames have the most goals scored, or if the flames are ahead of the Oilers in the standings. Hockey is a sport. Sports are about compelling storylines. Without a storyline, we wouldn't watch. Doffing deeper into why a team is winning it losing is absolutely part of a storyline.

[GioforPM]

I'm no good at looking at numbers, so I don't know if this guy's take is good or not. But here it is:


http://www.matchsticksandgasoline.co...e-to-live-in-a

[Aarongavey]

I am not a fancy stats expert, so this is a question to those that understand these numbers

http://stats.hockeyanalysis.com/team...T&sortdir=DESC

My reading of the fancy stats is that the Oilers are the 17th best team for Fenwick percentage. The Oilers have a better Fenwick than the Rangers, the Habs, Calgary, Ottawa, Philly and a host of others.

The Oilers are the 14th best team for Corsi. They are a better Corsi team than Anaheim, Vancouver, Rangers, Habs, Philly, Colorado, Calgary amongst others.

Am I reading those stats right, are the Oilers a better hockey team than the Habs, Flyers and Rangers based on those two stats? Are the Oilers better than the Ducks based on Corsi?

Do these two predictive stats say that the Oilers eventually will compete with the Ducks this year, assuming that everything regresses to the mean?

[Resolute 14]

Those stats don't say the Oilers are a better team, merely that they have probably possessed the puck better to this point. There are many possible explanations, including score effects. Opponents who are leading tend to play more passive styles. Also, the Oilers have actually had an easy schedule to this point. Their Corsi and Fenwick will likely trend down as they face more difficult opposition.

[FlamesAddiction]

Originally Posted by Resolute 14

Those stats don't say the Oilers are a better team, merely that they have probably possessed the puck better to this point. There are many possible explanations, including score effects. Opponents who are leading tend to play more passive styles. Also, the Oilers have actually had an easy schedule to this point. Their Corsi and Fenwick will likely trend down as they face more difficult opposition.

I think as long as people keep these advanced stats in perspective, they can be interesting to see where teams are doing well and where they need improvement.

My problem with them is that they are just one small piece of the puzzle where the final outcome is measured simply as wins and losses (or goals for and against). A lot of people use advanced stats like they are the bigger picture when in fact, they are a just one tool and not all that significant in the big picture.

I think part of my dislike and mistrust of them is that for the last 5 years at least, we have had to hear Oiler fans use them to prop up their team when the results the whole time have been terrible.