Wild Card Race (Western Conference)

[Dan02]

Originally Posted by EldrickOnIce

Problem is guys like Lambert, CHL, etc. will tell you it is the only thing that matters, to the point where wins and losses are wrong and don't reflect which team's are better. Which is complete nonsense.

This indeed, good teams will usually have a good corsi, this is obviously intuitive, however, that does not mean having a good corsi makes you a good team. One does not prove the other.

[dash_pinched]

Go Ducks and Go Penguins tonight!

[edslunch]

Originally Posted by Street Pharmacist

It hasn't been shown to be garbage. It's very useful and does have predictive value. It just doesn't describe everything.

This kind of absolutism is just as wrong and frustrating as the "corsi explains everything" assertions

I find Corsi and other such stats interesting when used to compare teammates and line partners. Same system, same games, same scores, same goaltender.....now let's see who helped tilt the ice in which direction. Between teams most of those variables are different

[CorsiHockeyLeague]

Originally Posted by getbak

That's 87.5 - 93.75% accuracy, so why bother trying to find some other solution, when the most basic formula already exists and is at least as accurate?

Maybe I should just put that chart from Steve Burtch in my signature since it seems to answer another question roughly every ten posts or so.



Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team. That's absurd. The 2011-2012 Edmonton Oilers went 7-2-2 in October and were near the top of the league, were they at that time an elite club? Funny how they ended up with the first overall pick. I honestly can't believe people think this.

[Dan02]

Originally Posted by CorsiHockeyLeague

Maybe I should just put that chart from Steve Burtch in my signature since it seems to answer another question roughly every ten posts or so.



Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team. That's absurd. The 2011-2012 Edmonton Oilers went 7-2-2 in October and were near the top of the league, were they at that time an elite club? Funny how they ended up with the first overall pick. I honestly can't believe people think this.

okay, where is total GF-GA/G listed on that table? Excluded because it doesn't fit the narrative?

This is only a list of which fancy stat is best, how about the naked ones?

[EldrickOnIce]

Originally Posted by CorsiHockeyLeague

Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team. That's absurd. The 2011-2012 Edmonton Oilers went 7-2-2 in October and were near the top of the league, were they at that time an elite club? Funny how they ended up with the first overall pick. I honestly can't believe people think this.

A team that has more points at the end of a season is better than a team who has less points.
That's pretty simple math, no?
So If Calgary makes the playoffs and finishes the season with more points than the Kings, who miss the playoffs, we can correctly conclude that Calgary was the better 2014-15 team, no?

[Street Pharmacist]

Originally Posted by Dan02

okay, where is total GF-GA listed on that table?

This is only a list of which fancy stat is best, how about the naked ones?

GF % is really the percentage of total goals that are goals for. Pretty close to goal differential

[Dan02]

Originally Posted by CorsiHockeyLeague

Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team.

And this is everything wrong with advanced stats in one sentence. Do you propose we stop giving the stanley cup to the team who wins all 4 rounds?

Or do you actually believe that Edmonton is better then Calgary and almost as good as Montreal?

[Dan02]

Originally Posted by Street Pharmacist

GF % is really the percentage of total goals that are goals for. Pretty close to goal differential

That's not on the chart though as far as i can tell. I see 5v5 GF%

[Textcritic]

Originally Posted by CorsiHockeyLeague

This discussion is pretty similar to the one that came up earlier in the thread when that graph was posted, so no point going over old ground. All I'll say is it's based on a methodology with a record of making correct predictions about future team success. That doesn't mean it's always right and it doesn't mean the Flames won't make the playoffs in spite of the methodology. It's based on this, subject to a bunch of adjustments made by Micah McCurdy.

http://www.silversevensens.com/2014/...tions-with-pip

I asked the question some time ago, and haven't yet received a response:

Does this model factor in each team's remaining schedule to factor into consideration each team's remaining games against one another?

If not, then that is a problem.

[Street Pharmacist]

Originally Posted by dissentowner

And yet Montreal and Calgary are poor Corsi teams but one is the best team in the East and the other is in the thick of the playoff race. It has predictive power but so does a magic 8 ball.

Seriously? I respect you as a poster, but that might be the dumbest argument yet. If a predictive probability is 80% yet you get a negative answer it's the same as 50%?

I hope you don't gamble

[Street Pharmacist]

Originally Posted by CorsiHockeyLeague

Maybe I should just put that chart from Steve Burtch in my signature since it seems to answer another question roughly every ten posts or so.



Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team. That's absurd. The 2011-2012 Edmonton Oilers went 7-2-2 in October and were near the top of the league, were they at that time an elite club? Funny how they ended up with the first overall pick. I honestly can't believe people think this.

Sample size, come on. At 41 games, if we're are looking at playoffs as the end point, it's pretty successful. Stronger than corsi in fact (though less useful I'd agree)

Incidentally, I think we are often arguing about subjective descriptors while using objective data. It's both true to say LA is a better team than Montreal and to say that Montreal is a better team than LA. It's an intellectually fruitless argument

[CorsiHockeyLeague]

^Of course sample size, but I'm dealing with the view that if you want to know what hockey team is better, just look at the standings and whoever is higher is better.

Originally Posted by Dan02

And this is everything wrong with advanced stats in one sentence. Do you propose we stop giving the stanley cup to the team who wins all 4 rounds?

No, but I also don't think that the team that's objectively better is necessarily determined in a 7 game series. Boston, for example, probably should've beaten the Habs last year. To win a cup, you, at the very least, need avoid a bad run of injury luck.

That being said, you've created a straw man. All I'm saying is that standings position does not mean a team is better at hockey than another. This is not a "problem with analytics", it should be obvious to anyone whether they've ever HEARD of corsi or not. When the Kings were in 10th, they were not somehow a worse hockey team than the Avalanche, and suddenly despite no significant personnel changes in the interim became better than those teams over a span of a few weeks. They always were better. Similarly, in 2012-2013 when the Blackhawks had a mediocre start and, IIRC, sat outside the playoffs, didn't mean they were worse than eight other Western Conference teams and suddenly got better despite having the same players and went on to win the Cup.

If, next year, the Sabres somehow win their first four games while the Ducks and Penguins lose theirs, that will not mean Buffalo is at that point better than Pittsburgh and Anaheim at hockey. That's just silly. I'm amazed I even have to talk about this.

Or do you actually believe that Edmonton is better then Calgary and almost as good as Montreal?

I don't recall saying so. Why do people keep trying to put words in my mouth? I hope everyone can understand how unbelievably annoying this is.

Edmonton has been an improved possession team this year, and their lack of success is attributable to a number of things, the most obvious being atrocious goaltending. If you want to see what atrocious goaltending can do to an otherwise good team (which I am not saying the Oilers are), look at the difference between Minnesota pre and post God-mode Dubnyk.

[Street Pharmacist]

Originally Posted by Dan02

That's not on the chart though as far as i can tell. I see 5v5 GF%

It's not, but lots of work has shown what happens 5v5 is a better predictor

[dissentowner]

Originally Posted by Street Pharmacist

Seriously? I respect you as a poster, but that might be the dumbest argument yet. If a predictive probability is 80% yet you get a negative answer it's the same as 50%?

I hope you don't gamble

Actually I don't gamble, I have terrible luck. I am sure Corsi does have value but I find it annoying. Anyhow we have de-railed this thread badly.

[Dan02]

Originally Posted by Street Pharmacist

Seriously? I respect you as a poster, but that might be the dumbest argument yet. If a predictive probability is 80% yet you get a negative answer it's the same as 50%?

I hope you don't gamble

So if you had at your disposal a tool with a predictive probability of 90%, you'd rather use a predictive tool with a probability of 80%??

I hope you don't gamble.



Graphic/analysis courtesy of thehockeywriters.com not me.

As clearly can be seen Goal differential in the regular season is more predictive then corsi at detemining who makes the playoffs, who makes the second round, who makes the third round, who makes the finals AND who wins the stanley cup.

So through all their fancy math these stat guys have managed to concoct a new method of predicting things, which sadly isn't even as good as a old one.

[CorsiHockeyLeague]

^That chart does not give you reliability, which the one I posted before does.

Originally Posted by Textcritic

I asked the question some time ago, and haven't yet received a response: Does this model factor in each team's remaining schedule to factor into consideration each team's remaining games against one another? If not, then that is a problem.

I actually do not know if that is one of the adjustments made. I don't know if it factors into sportsclubstats's model either.

Originally Posted by EldrickOnIce

So If Calgary makes the playoffs and finishes the season with more points than the Kings, who miss the playoffs, we can correctly conclude that Calgary was the better 2014-15 team, no?

No. After 82 games, it may be a better indicator, but it's not an absolute. Take the New Jersey Devils. They finished behind the Coyotes in points. They went 0-13 in the shootout. That's amazingly unlucky. I would suggest that they were better than the Coyotes. This is one of many examples. Sometimes nothing goes right for a team during a season; sometimes everything goes right for them. Similarly, the Avalanche finished 3rd in league standings last year. The Kings and Blackhawks finished 7th and 10th respectively, then played the biggest battle of titans series in recent memory. Those teams were both better at hockey than the Avalanche and it wasn't close.

Let's take another spin on it, shall we: by your argument, the Vancouver Canucks were the best team in the National Hockey League for two years running in 2011-2013.

[Textcritic]

Originally Posted by Street Pharmacist

Incidentally, I think we are often arguing about subjective descriptors while using objective data. It's both true to say LA is a better team than Montreal and to say that Montreal is a better team than LA. It's an intellectually fruitless argument

Yes!

I see that these data and their interpretations can be useful for explaining why certain teams are or are not successful, but they clearly do not work universally to such an extent that they clearly demonstrate which teams are better than others, whatever that means.

[Enoch Root]

Originally Posted by CorsiHockeyLeague

Incidentally to the above I find it baffling that anyone can take the view that a team being higher in the standings means they're a better hockey team. That's absurd. The 2011-2012 Edmonton Oilers went 7-2-2 in October and were near the top of the league, were they at that time an elite club? Funny how they ended up with the first overall pick. I honestly can't believe people think this.

It is far from absurd. A team higher in the standings has, by definition, performed better. To draw the conclusion from that that they are better is not at all absurd. Premature, as with any other stat, sure. Absurd? No. The fact that you think it is, shows again your overconfidence in your opinion.

As for the Oilers, you might want to take a bit of your own advice and consider sample size. Any team, over a short number of games, can win the vast majority of them. Sample size applies on that front too.

[CorsiHockeyLeague]

Eh, but subjective descriptors can still have at least some truth value to them. If someone tells you they think the Oilers are a better hockey team than the Blackhawks, it won't strike you as being equally true as the inverse.