Yes, I agree that whatever Todd was trying to express could've been expressed better. First off, titling the blog "Zajac, Jagr Among NHL's Elite Duos" right off that bat shows me that Todd was trying to hype them up more than they are. That is such a broad and misleading title. To be an elite duo in the league means being elite in multiple categories/stats, not just one. A more appropriate title would've been "Zajac, Jagr Among NHL's Elite Possession Duos." However, even then I'd bet that if he actually charted every duo in the league that there would be a good number with better possession statistics above Zajac and Jagr and a good number of them would likely be 3rd and 4th line duos. Obviously a duo of 2 offensive superstars like Benn and Seguin will have a lower Corsi (which is calculated using shots for vs. shots against) because they are not good defensively. They give up a lot of shots. Look up a shutdown duo and you could find a couple of 3rd lines that have a rate 5% higher or even more. And he also doesn't include Boston's top line duo of Bergeron and Marchand. I'd be interested to see how that duo's numbers compare to Zajac's and Jagr's.

Simply put, Todd used limited stats from duos that purposely aren't as good as Jagr and Zajac in order to hype them up.
- RAGSareDANGERus

And now there are 5 pages of replies...

Zajac = over paid one hit wonder

Jagr = Fossil

Cammalleri = past his prime / on the decline

add it all together and it = no playoffs
- dwoo33

at least we have the perspective of a leafs fan to compare it to.

I love when people call them "fancy stats", just shows the person doesn't know anything about them.
- rmdevil313

Not trying to start poop, but they're not exactly advanced either.

Yes, I agree that whatever Todd was trying to express could've been expressed better. First off, titling the blog "Zajac, Jagr Among NHL's Elite Duos" right off that bat shows me that Todd was trying to hype them up more than they are. That is such a broad and misleading title. To be an elite duo in the league means being elite in multiple categories/stats, not just one. A more appropriate title would've been "Zajac, Jagr Among NHL's Elite Possession Duos." However, even then I'd bet that if he actually charted every duo in the league that there would be a good number with better possession statistics above Zajac and Jagr and a good number of them would likely be 3rd and 4th line duos. Obviously a duo of 2 offensive superstars like Benn and Seguin will have a lower Corsi (which is calculated using shots for vs. shots against) because they are not good defensively. They give up a lot of shots. Look up a shutdown duo and you could find a couple of 3rd lines that have a rate 5% higher or even more. And he also doesn't include Boston's top line duo of Bergeron and Marchand. I'd be interested to see how that duo's numbers compare to Zajac's and Jagr's.

Simply put, Todd used limited stats from duos that purposely aren't as good as Jagr and Zajac in order to hype them up.
- RAGSareDANGERus

oh, i get it then. he should write about how the devils suck to satisfy the rangers fans.

hurr durr

oh, i get it then. he should write about how the devils suck to satisfy the rangers fans.

hurr durr
- MrToast

After I legitimately criticize his blog for very valid points and prove that YOU are the one that truly doesn't understand the statistics and how he's warping them, this is the best you can come up with?

Not trying to start poop, but they're not exactly advanced either.
- kcksvnabty

They are advanced from the standpoint that they are not straight statistical (counter type) values...hits, shots, goals, pims, points, plus/minus etc. They require some additional computation (of straight stats), often within different game situations.

After I legitimately criticize his blog for very valid points and prove that YOU are the one that truly doesn't understand the statistics and how he's warping them, this is the best you can come up with?
- RAGSareDANGERus

why didn't you just lead off with your problems with the blog was grasping at straws and save everybody a lot of time and effort.

After I legitimately criticize his blog for very valid points and prove that YOU are the one that truly doesn't understand the statistics and how he's warping them, this is the best you can come up with?
- RAGSareDANGERus

Statistics can't be warped

They are advanced from the standpoint that they are not straight statistical (counter type) values...hits, shots, goals, pims, points, plus/minus etc. They require some additional computation (of straight stats), often within different game situations.
- icedog97

They are simply analytics. Equations to attempt to find patterns in raw data.

Statistics can't be warped
- willschulme

Of course they can be.

at least we have the perspective of a leafs fan to compare it to.
- MrToast

Your defense is bad... I mean worse than the LEAFS DEFENSE bad. I see forward depth and solid goalie who will burn out because the back up is completely useless.

After I legitimately criticize his blog for very valid points and prove that YOU are the one that truly doesn't understand the statistics and how he's warping them, this is the best you can come up with?
- RAGSareDANGERus

You mean like how he chooses not to provide stats for Bergeron/Marchand, Krejci/Iginla, Kopitar/Williams, Tarasenko/Berglund etc.

Your defense is bad... I mean worse than the LEAFS DEFENSE bad. I see forward depth and solid goalie who will burn out because the back up is completely useless.
- dwoo33

On paper, maybe. But the Devils play an excellent defensive system which goes a much longer way than just having a good defense on paper.

You mean like how he chooses not to provide stats for Bergeron/Marchand, Krejci/Iginla, Kopitar/Williams, Tarasenko/Berglund etc.
- Feds91Stammer

Yup. Basically, let's compare our duo which is made up of 1 guy who is a great two-way forward and the other a savvy veteran who play in a defensive system and compare their shots for/shots against rate to run-and-gun duos who cheat on defense. It's comparing apples to oranges.

Yup. Basically, let's compare our duo which is made up of 1 guy who is a great two-way forward and the other a savvy veteran who play in a defensive system and compare them to run-and-gun duos who cheat on defense. It's comparing apples to oranges.
- RAGSareDANGERus

This sums up the usage of advanced stats.

On paper, maybe. But the Devils play an excellent defensive system which goes a much longer way than just having a good defense on paper.
- RAGSareDANGERus

I wish the Leafs would play a good defensive system its painful

This sums up the usage of advanced stats.
- Feds91Stammer

EXACTLY. They have their uses but NOBODY uses them correctly or objectively. This blog, written by one of the biggest advanced stats promoters on this website, is the epitome of using the stats to only favor one side of the argument on purpose. And that's literally how every advanced statistic is used in debates and blogs.

Why should anyone trust this crap when one of the biggest supporters of it around here uses it grossly incorrectly ON PURPOSE and is well aware of it?

Wow, this comment thread was a flash bang full of concentrated stupid. I should have looked away.

what a cluster(frank) this turned into.
- MannySilvers

Yes, yes it did. The advanced stats v. traditional stats debate really came to the fore a couple of years ago with the Trout/Cabrera debate, and now they have invaded hockey talk as well. Personally, I see advanced stats as a useful tool, but the amount of discourse about them over the last couple of months has become mindnumbing. Again, I fully realize their utility, and the limits thereto, but I could certaintly use a vacation from Corsi, Fenwick, et.

Dallas-NJ Final

We need revenge for 2000

Off season blogs comments.



Who needs advanced statistics anyway, they are all numbery and stuff. We can just guess and hope and things.

I find it interesting that people bashing the advanced stats don't see a direct correlation between the top teams and there fenwick and Corsi number.

If you look at the below page you can see there is a combination of TWO needs.
Possession and skill to bury the biscuit.

So at the end of the season if you looked at the numbers you would likely be right picking any of the top possession teams, OBVIOUSLY!

http://www.extraskater.co.../teams/on-ice?season=2013

There are TWO unmeasurables:

CLUTCH........NY had that characteristic in the playoffs especially in goal.
LUCK.......damn you MTL.

If we were guessing before the playoffs last year, using observation, normal GA and GF, etc, most of us would have picked LA, Chi, and Bos.

Using advanced stats would support those observations wouldn't it.

NY was a surprise to a degree, but now that I see them so high in the FF and Corsi, I can't say I am surprised.

A bit of luck and a hot goalie and they made it to the SCF. But there luck and hot streak ran out because of the consistency that LA has had in possession and skill to score.

And great job not putting either Bergeron and Marchand or Bergeron and Smith on that list.

Just looked and 5 v 5 they were a top pair either way in Corsi, Fenwick, and GF to GA.

But if you ignore a bunch of the duos ahead of your guys I guess that would more them up the list though........

And great job not putting either Bergeron and Marchand or Bergeron and Smith on that list.

Just looked and 5 v 5 they were a top pair either way in Corsi, Fenwick, and GF to GA.

But if you ignore a bunch of the duos ahead of your guys I guess that would more them up the list though........
- JIwasinskiJr

This was my exact point. This chart does not prove anything about them being elite because it does not compare them against the actual elite duos in that category. It simply proves they have the best possession stats as a duo compared to 8 other duos he hand picked that all played on bad defensive teams.

It's the misuse of advanced stats like this that gives it such a bad reputation among the "haters".

The blog, like a graph when writing statistical papers or doing experimental design & analysis, needs to be named correctly. To avoid ambiguity and bumhole generalizations by retards like those present in this comments section. Terms like "duo" or "elite" should be avoided because they will need to be defined further.

Unfortunately, there are only so many characters.

The blog should've effectively been named, for instance, something similar to:

"Jagr & Zajac among best in NHL in Goals For % & Shots Attempt For % with minimum of 800 mins played together in 5v5 situations, excluding pairs with 5v5 Goals For % < 40%"

I'm sure you could even do better than that, but you get the idea. Good luck fitting all that into the title.

You could’ve gone with “Statistical Analysis of Jagr & Zajac” and left out “elite”. This may have spared you some of the absurd comments.

People who try to de-value advanced analytics tend to be retarded or lack the context to apply them in, which as I mentioned is seldom a "catch-all" analysis.

They're relevant only to the parameters expressed.

EDIT: I don't know of anyone in the advanced statistics community that is so naive to suggest they're the complete picture, but they're very useful when used properly and when the analysis is done with specificity. The audience also needs to be knowledgeable or made aware of the context, but good luck getting an intelligent crowd on any public message board.