Devils 2017-18 team discussion (player news and notes) - season begins!

[R8Devs]

For the 2017-2018 devils -- First of all it's only been 18 games so statistics like that can change in a matter of a couple games (and it probably will since Zajac is back). Second the Devils are shooting 11.5% currently, last year the best team finished with 10.5%(Capitals) and the year before that was 10.1%(Stars). Devils are also a good powerplay team which is not included in those stats and it has had a definite impact on the record. And the Devils have good goaltending too.

People misuse it all the time and it shouldn't be an end it all part of a discussion but outright dismissing it just because there are times when it's widely off the mark is pretty much as dumb.

 

[Bleedred]

For the 2017-2018 devils -- First of all it's only been 18 games so statistics like that can change in a matter of a couple games (and it probably will since Zajac is back). Second the Devils are shooting 11.5% currently, last year the best team finished with 10.5%(Capitals) and the year before that was 10.1%(Stars). Devils are also a good powerplay team which is not included in those stats and it has had a definite impact on the record. And the Devils have good goaltending too.

People misuse it all the time and it shouldn't be an end it all part of a discussion but outright dismissing it just because there are times when it's widely off the mark is pretty much as dumb.

You summed this up pretty well, though I think our shooting percentage is down below 11% with tonight's 42 shots and 0 goals. It was 11.7 before this game.

We were driving under the influence of something called PDO, which is not exactly an advanced stat, it's just overall shooting percentage, combined with overall save percentage.

Now if Cory plays like this the entire year, our save percentage is sustainable. Even consider that Kinkaid hasn't been very good, and there's a good chance he improves a bit from where he's at right now. I don't think we were gonna shoot at this rate though. With this kind of goaltending from Cory, we can sustain a higher PDO. Not this high though.

The only teams that seem to consistently defy these PDO stats are Pittsburgh and Washington, and both teams have some of the most higher percentage shooters, that score a ton of goals. Like Ovechkin namely, and Washington has also had some of the best goaltending over the past several years as well.

The Rangers have seemed to sustain a bloated PDO over the last several years, due to an above average shooting percentage and high save percentage. With Lundqvist declining and them having a non-existent backup, I think it will be hard for them to put up a 101 PDO again this year, without an even crazier shooting percentage to compensate for the decline in goaltending.

So to sum things up, we've overachieved a bit, at least in shooting. Goaltending, not so much. Our PDO came down a bit after tonight, should be below 102 by now. Most teams don't get into the playoffs with a PDO over 102 and when they do, they don't sustain from year to year. Just ask the lockout year Leafs and the 13-14 Avalanche. This is another reason why I don't rush to give Hynes and the coaching staff credit for this start. We allow A LOT of shots too. Even more than last year, which might even help boost Cory's save percentage. Cory has a save percentage of .924 (VERY good) with a GAA of 2.64 (BAD) which is a team stat and means we're allowing too many shots. With last year's Cory, we probably would be losing a lot more games, with how many shots are being allowed. That wasn't at all a problem tonight, this was our best possession game of the year, but we were bound to lose one of these, after winning some that we didn't deserve.

 

[devilsblood]

Last year

4 of the top 5 and 7 of the top 10 corsi teams made the playoffs.

The 5 worst, and 8 of the 10 worst corsi teams did not make the playofffs.

It's a good stat.

 

[Emperoreddy]

I certainly don’t buy Corsi telling me that Edmonton doesn’t suck.

I’ve watched them enough and they suck. They aren’t going to go on an insane streak and make the playoffs barring McDavid losing his mind.

 

[Bleedred]

Last year

4 of the top 5 and 7 of the top 10 corsi teams made the playoffs.

The 5 worst, and 8 of the 10 worst corsi teams did not make the playofffs.

It's a good stat.

This is why I’m slow to give Hynes any credit for this start. With that crazy shooting percentage, us making the playoffs would have been like Roy in 13-14. Which you could see was not sustainable.

On the other hand, most of the top teams in the standings are really high on PDO right now. Tampa’s is 103 something the last I saw. St Louis and LA were quite high on the PDO scale too. And Arizona, Buffalo, Montreal, all them have a PDO that’s in the toilet right now.

Florida is the only bad team with a PDO that’s not completely in the toilet. It’s 99.3, rather than 97.1 or 96.7. It’s barely under 100, which is the mean. That would a team with an average shooting and save percentage. So that’s why I say, Florida is the worst team in the league and there’s a great case to be made for that.

They’re currently beating the Sharks 1-0 and outshooting then, so I expect they’ll have close to a 100 PDO tonight, while still being 2 games below 500 if they win this.

 

[Emperoreddy]

Last year

4 of the top 5 and 7 of the top 10 corsi teams made the playoffs.

The 5 worst, and 8 of the 10 worst corsi teams did not make the playofffs.

It's a good stat.

16 teams make the playoffs, but only 7 of them were in the top 10? That doesn’t sound as concrete as you are claiming.

If Corsi was such an excellent predictive model of success I would expect 12-13 of the top 16 teams to make the playoffs with only one or two outliers, and for that to happen every year.

 

[devilsblood]

In the 2016 playoffs.

The complete top 10 list of corsi teams made the playoffs.

8 of the the bottom 10 did not.

 

[devilsblood]

16 teams make the playoffs, but only 7 of them were in the top 10? That doesn’t sound as concrete as you are claiming.

If Corsi was such an excellent predictive model of success I would expect 12-13 of the top 16 teams to make the playoffs with only one or two outliers, and for that to happen every year.

I'm going with the more extremes because the difference between the 16th team and 17th team is pretty minimal. But when 85% of the top 10 corsi teams over the last 2 years make the playoffs(and the same on the other end of the spectrum) I think that's proving to be a pretty solid predictor of success.

And no it's not concrete. I'm saying it's a good stat, not the greatest stat ever.

 

[Zippy316]

Okay, one more example before bed, because math is fun.

The 15th worth TEAM CORSI ever recorded in history belongs to your 2017-2018 New Jersey Devils at a horrendous 45.7 CF%.

Your ignorance is flat out ridiculous.

What you're doing is the same as pulling up goals from last year and going Oshie scored 33 goals in 68 games and Ovechkin scored 33 goals in 82 games so Oshie is the better goal scorer. How stupid does that sound? Any person who has any knowledge of hockey would know that's not going to continue on. The difference is CF% is new and not an archived, historical stat like goals.

So yes, over small sample sizes there are times when CF% doesn't align with the narrative, but the same can be said about every other stat. If 8/10 of the top 10 teams in CF% make the playoffs every year, is that not saying something? And again, just like any stats it can be derailed by anomalies. Oshie shot at a ridiculous percentage last year (23.1% versus his previous career high of 14.1%) whereas Ovechkin was ~ 4% below his normal shooting percentage with a shot/game much lower than usual so that's unlikely to happen again. Oshie may be a case of him being a lethal finisher (he's at 24.1% with 9 goals so far), but Ovechkin is shooting at a higher rate than last year with a percentage closer to his norm and has 13 goals so far.

This Devils team has been riding some high shooting percentages -- as R8NJ noted, as a team the Devils are shooting at 11.5%, last year the highest team % was the Capitals at 10.5%. Is that really going to continue? The Devils have also been riding an extremely strong PP and have had Schneider steal a couple games so far this year.

High shooting percentage, goalie standing on his head, and a PP are all of the things that aren't reflected in CF% and will help cover up a poor one. Over the course of a season, these things tend to normalize -- Devils will have games like tonight where they won't score, they will have games where the goaltending is awful, the PP should come back down to earth. The Devils have barely had any of that this year which is why the record is the way it is despite the CF%.

 

[BenedictGomez]

In the 2016 playoffs. The complete top 10 list of corsi teams made the playoffs.

False.

Only 6 of the top 10 did.

And the #2 and #3 best Corsi teams in the entire NHL that season BOTH missed the playoffs (Carolina & Boston).

2016 Stanley Cup playoffs - Wikipedia


EDIT: Actually, eyeballing that further, 4 of the worst 10 Corsi teams in the NHL made the playoffs that season, which depending on perspective is even more striking.

 

[NJDevs26]

So yes, over small sample sizes there are times when CF% doesn't align with the narrative, but the same can be said about every other stat. If 8/10 of the top 10 teams in CF% make the playoffs every year, is that not saying something? And again, just like any stats it can be derailed by anomalies. Oshie shot at a ridiculous percentage last year (23.1% versus his previous career high of 14.1%) whereas Ovechkin was ~ 4% below his normal shooting percentage so that's unlikely to happen again.

Honestly I think it's a bit confirmation bias. Of course most of the good teams are going to have high Corsi but do teams really win because of possession and shot totals or because they have enough offensive talent to take advantage of said possession and shot totals, which generally lead to good Corsi anyway? My problem with the whole Corsi argument is people acting like it's only shot difference that's a determinant. I've seen plenty of Devils teams be anti-Corsi arguments to trust the stat

For two years after the last lockout it was how unlucky they were and how historically low their shooting percentage was instead of well they just didn't have enough actual talent to score goals despite their possession. Maybe if we had a different team I wouldn't roll my eyes as much at the stat but since the Devils were the biggest anti-Corsi argument for a long time as it was becoming in vogue it's a different story.

 

[BenedictGomez]

Your ignorance is flat out ridiculous.

What you're doing is the same as pulling up goals from last year and going Oshie scored 33 goals in 68 games and Ovechkin scored 33 goals in 82 games so Oshie is the better goal scorer. How stupid does that sound? Any person who has any knowledge of hockey would know that's not going to continue on. The difference is CF% is new and not an archived, historical stat like goals.

I just read this entire post twice, because the first time I read it it didnt make any sense. Read it again, doesnt make any sense. It's not even remotely similar.

And I'd say 331 data points each comprised of over 20 separate data points, which comprises thousands of data points, is fairly decent representative sampling.

 

[Emperoreddy]

And at what sample size is a team’s Corsi valid to be used as a prediction for future success or failure?

10 games? 20? 50? Or do we need all 82?

To me it just seems like one of many stats that tells one part of the picture, but does not predict with any certainty how a team is actually performing.

Yes you want to have a better Corsi and most likely you will get better results, but it isn’t anywhere near guaranteed and plenty of other factors play into overall team success over an 82 game season.

I think far too many new journalists on hockey blogs, and on websites simply look at their spreadsheets to make their sweeping claims without ever actually watching the actual product on ice night in and night out. So they only get part of the picture and not enough context to put it together properly.

Never mind that plenty of people also introduce their own bias because they have a point they want to make so they pick the stars that back up their point and dismiss what doesn’t go with them as outliers or unimportant.

I don’t think the vast majority of fans and journalists use statistics or analytics correctly in regards to hockey.

 

[BenedictGomez]

If Corsi was such an excellent predictive model of success I would expect 12-13 of the top 16 teams to make the playoffs with only one or two outliers, and for that to happen every year.

Another poster has become Woke.

 

[Emperoreddy]

Honestly I think it's a bit confirmation bias. Of course most of the good teams are going to have high Corsi but do teams really win because of possession and shot totals or because they have enough offensive talent to take advantage of said possession and shot totals, which generally lead to good Corsi anyway? My problem with the whole Corsi argument is people acting like it's only shot difference that's a determinant. I've seen plenty of Devils teams be anti-Corsi arguments to trust the stat

For two years after the last lockout it was how unlucky they were and how historically low their shooting percentage was instead of well they just didn't have enough actual talent to score goals despite their possession. Maybe if we had a different team I wouldn't roll my eyes as much at the stat but since the Devils were the biggest anti-Corsi argument for a long time as it was becoming in vogue it's a different story.

I think confirmation bias is the biggest sin committed in almost any analytics discussion, or article.

It makes sense. It is human nature to want to do that.

 

[NJDevs26]

Okay, one more example before bed, because math is fun.

The 15th worth TEAM CORSI ever recorded in history belongs to your 2017-2018 New Jersey Devils at a horrendous 45.7 CF%.

That #317 of 331 Corsi means 96% (literally) of all NHL teams over the last decade of existing data have better Corsi than the current Devils.

The 2017-2018 New Jersey Devils currently have the 3rd best record in the entire NHL.

Considering our goalscoring numbers before tonight this is hilarous if true. Maybe this year will make up for the post-lockout Pete years with our abnormally, historically low shooting percentage.

 

[Zippy316]

I just read this entire post twice, because the first time I read it it didnt make any sense. Read it again, doesnt make any sense. It's not even remotely similar.

And I'd say 331 data points each comprised of over 20 separate data points, which comprises thousands of data points, is fairly decent representative sampling.

I don't think I could've put it any simpler.

CF% is indicative of team success. There is anomalies like every other stat. The two main things that can cover up a poor CF% are high PDO (shooting % and save %) and special teams. Devils have benefitted from both so far this year.

To liken it to goals, just cause one player outscores the other doesn't make him better. Ovechkin should rebound to his normal numbers which he's historically produced over his career. Devils should fall down to earth with their PDO and special teams if their CF% doesn't improve. It's not hard to grasp.

 

[devilsblood]

False.

Only 6 of the top 10 did.

And the #2 and #3 best Corsi teams in the entire NHL that season BOTH missed the playoffs (Carolina & Boston).

2016 Stanley Cup playoffs - Wikipedia


EDIT: Actually, eyeballing that further, 4 of the worst 10 Corsi teams in the NHL made the playoffs that season, which depending on perspective is even more striking.

Are we talking 5v5? What stat site are you using?
https://www.naturalstattrick.com/te...=all&loc=B&gpf=82&fd=2014-10-08&td=2015-04-11

 

[Zippy316]

I think confirmation bias is the biggest sin committed in almost any analytics discussion, or article.

It makes sense. It is human nature to want to do that.

But it's just as ridiculous to pull out the 2012-13 Devils or a Kings team and use that as confirmation bias against CF%.

Like I keep saying, that's the equivalent of saying Oshie scored at a better rate than Ovechkin last year so he's a better goal scorer. There's years of prior data from those two players that shows that likely won't be the case again and if you further analyze the stats and look at shooting percentages and shot rates, you'll see why it happened last year.

You can do the same for CF%. You compare that to shooting percentage, save percentage, and special teams and you can have an overall picture of what happened, why it happened, and if it will continue. Devils in 2012-13 had awful special teams if I recall correctly, had Hedberg do his best Swiss cheese impression for 10 games, and had a forward group devoid of talent that shot at low percentages. That tells the story of a Devils team that dominated the game, but lost the game because they couldn't finish, couldn't make a save, and had terrible special teams.

 

[Emperoreddy]

I mean f*** it isn’t even the 100% agreed stat for possession. There is an excellent argument that Fenwick does a better job of measuring possession over the course of a game.

How can you try and claim that this one stat can predict with high accuracy how future results are going to go over the course of a season? If you want to use it as one part in a while to discuss a team’s current performance? Sure go ahead, but to predict future results? No way.

I’ve watched every game this year and I feel confident in saying we are not a lotto team no batter what our Corsi is. I’ve watched a number of Edmonton games and I feel confident in saying they are not a good team and their good Corsi does not mean they are unlucky. They are just a bad team.

 

[Jets012]

I've personally always felt Corsi is more important when evaluating defenseman than forwards. Obviously just looking at Corsi without any context isn't going to tell you anything. You need to see who each player is playing with and what type of competition the player is typically facing (I do not think zone-starts have much of an effect at all, it's also statistically shown they rarely have an effect over a season).

If a forward produces on the 5 on 5 points wise per minute, I'm not going to care much about their Corsi. I don't care that Ovechkin might not have the best Corsi #s consistently if he's still outscoring almost everyone on the 5 on 5. But yea, I'm going to pay attention to Corsi with defenseman when I think there's a strong correlation that the defenseman that have the puck out of their own zone the most are going to be the ones that are on the ice for the least amount of goals and scoring chances.

And obviously there are going to be exceptions to the rule when evaluating a team by Corsi. Not surprising that a team like the Devils in 2012-2013 that had a great Corsi didn't make the playoffs when they didn't get good goaltending and had next to no one that could score goals.

But yea like someone said, pretty f***ing hard to find teams that win it all that aren't at the very least in the top half of the league in corsi or even better. Look at recent champions and runner ups:

2017: Penguins - #16 (#1 in xGF though) / Predators - #5
2016: Penguins - #2 (and they were first in the league after Sullivan took over) / Sharks - #9
2015: Blackhawks - #2 / Lightning - #4
2014: Kings - #1 / Rangers - #7
2013: Blackhawks - #4 / Bruins - #3
2012: Kings - #1 / Devils - T#12
2011: Bruins - #14 / Canucks #5
2010: Blackhawks - #1 / Flyers #14

Oh yea and the year the Wings won, they were #1 in Corsi in the regular season too. So yea as you can see, pretty freaking strong correlation in Corsi with the teams that end up going the furthest. Especially in Power Plays. Unless you get absurd goaltending, great puck luck, or a wicked hot stretch, it's quite obvious the teams that play the best on the 5 on 5 are the teams that advance each round.

 

[BenedictGomez]

Are we talking 5v5?

Hell NO I'm not only using it only 5v5! I'm using it for all situations.

If there's any predictive team value to this metric to begin with, why the hell would I eliminate all powerplay and penalty kill minutes, which is further weakening my data set?!?!?!?! By a lot.

That makes absolutely zero sense.

 

[Zippy316]

I mean **** it isn’t even the 100% agreed stat for possession. There is an excellent argument that Fenwick does a better job of measuring possession over the course of a game.

How can you try and claim that this one stat can predict with high accuracy how future results are going to go over the course of a season? If you want to use it as one part in a while to discuss a team’s current performance? Sure go ahead, but to predict future results? No way.

I’ve watched every game this year and I feel confident in saying we are not a lotto team no batter what our Corsi is. I’ve watched a number of Edmonton games and I feel confident in saying they are not a good team and their good Corsi does not mean they are unlucky. They are just a bad team.

Fenwick is just Corsi without considering shots that missed the net. I'd be curious to see if anyones compared those to playoff teams, but there usually isn't much of a difference between the two.

Why can't it be used to predict future success? Ovechkin shot at a much lower rate last year and a lower shooting percentage (10% versus his 14% in his 50 goal seasons). If Ovechkin starts shooting more and scores closer to his percentage, he should likely have better numbers. I think that's a fair assumption to make and so far this year, that's exactly what's happening.

Similarly, if CF% has been shown to be an indicator of team success and it has, the Devils showing a poor CF% should normalize the same way Ovechkin's stats would. I think part of the problem is people don't try to understand what the stat actually means and just dismiss it because way too many people use it improperly.

 

[Zippy316]

Hell NO I'm not only using it only 5v5! I'm using it for all situations.

If there's any predictive team value to this metric to begin with, why the hell would I eliminate all powerplay and penalty kill minutes, which is further weakening my data set?!?!?!?! By a lot.

That makes absolutely zero sense.

The game is primarily played at even strength 5-on-5. If you win that battle, more often than not you win the game.

Over a full season, if you get dominated 5-on-5, you likely won't be winning many games and vice versa.

 

[BenedictGomez]

I mean **** it isn’t even the 100% agreed stat for possession. There is an excellent argument that Fenwick does a better job of measuring possession over the course of a game.

That's not even an "argument", it's a fact.

If you're going to buy into this crap, Fenwick without a doubt is better than Corsi, because it at least corrects a glaring Corsi error.

Which is why I say that the "Corsi people" often just want to sound smart, because if they really understood what they were talking about they'd be using Fenwick instead, yet almost nobody does.