How did the Kings miss the playoffs?

[Dr. Doom]

Advanced stats would suggest that this shouldn't have happened, that these guys held the puck with the best of them:

The thing is, it’s not like the Kings collapsed. They were the league’s top possession club (http://www.fenwick-stats.com/?s=2014-10-01&e=2015-04-30&f=0&ld=1&l=82) (getting 54.9% of score-adjusted shot attempts) and still have the 13th-best goal differential (+0.22 per game), which is better than three playoff teams (Anaheim, Detroit and Vancouver).

http://www.tsn.ca/what-happened-to-the-kings-1.253079

[nik-]

but ... the CORSI

http://cdn.meme.am/images/290101.jpg

[troutman]

Can't score goals.

[GranteedEV]

They only got four of ten potential points against the Calgary Flames, while the Calgary Flames got eight of ten potential points against the LA Kings.

[nik-]

Eventually advanced stats nerds will accept that not all shots are created equal.

[Buster]

So in a 30 team league we have at LEAST two outliers. (The Flames and the Kings).

Methinks Corsi needs a methodology adjustment.

[normtwofinger]

They only got four of ten potential points against the Calgary Flames, while the Calgary Flames got eight of ten potential points against the LA Kings.

This.

[Jbo]

Would love to hear Quincy Egg explain it actually

[locsofblu]

Because they didnt accumulate enough points to solidify a playoff spot in the western conference. Corsi can suck my balls!

[Buster]

Eventually advanced stats nerds will accept that not all shots are created equal.

You just need to watch the Hudler/Jonny/Mony line to know that.

[Locke]

Possession only counts if you do something with it.

[Phanuthier]

advanced stats aside....

Losing Voynov and Mitchell really hurt them a lot. Tack on bad years from Kopitar and Brown, and the team just didn't "have it" this year.

Playoff race is incredibly close, hell the gap between us and Anahaim is only 10 points.

[sureLoss]

They lost to the Oilers, twice.

[CedarMeter]

Couldn't score. Weren't they like 20th in scoring this year.

[goodyear]

because they got beat by the Flames?

[spuzzum]

They have played a lot of hockey the last few years. They have what every team needs, a world class goalie, superstar defense man and established number one center. Supplemental scoring and just the massive amount of tough games takes its' toll.

[Winsor_Pilates]

They only got four of ten potential points against the Calgary Flames, while the Calgary Flames got eight of ten potential points against the LA Kings.
This completely sums it up.
We beat them out of their spot this year; 15 other playoff teams can thank Calgary that they won't have any chance of facing the Kings.

LA's terrible OT/Shootout record was a big factor too.

[Gaudfather]

Because they weren't playing for each other enough of the time!

[sureLoss]

advanced stats aside....

Losing Voynov and Mitchell really hurt them a lot. Tack on bad years from Kopitar and Brown, and the team just didn't "have it" this year.

Playoff race is incredibly close, hell the gap between us and Anahaim is only 10 points.

You also have to add the decision not to compliance buyout Richards when they had the chance. That was essentially $5.75 million dead cap space that they could have used to get some more scoring or fill the hole Mitchell left.

[_Q_]

Because corsi does not equal possession. Even the NHL recognizes that on their website.

[Ice_Weasel]

Perhaps someone more adept at data mining can confirm, but if memory serves the shot counting in LA is very skewed in their favor. I couldn't find shot attempts anywhere broken out between home and away games to see if there is an arena bias padding LA's corsi.

I get that Corsi is pure calculation, and as the purists say "stats are stats, they can't be wrong". But what if the raw data (stat) is wrong because its completely subjective (what is a shot or attempt in one building may not be in another).

[memphusk]

Calgary Flames. That's why.

[getbak]

They left three points on the table against the Oilers and one against the Coyotes. Give them those four points, and they're tied with the Flames right now.

If they could have held their two third period leads at home against the Flames, Calgary would have 4 fewer points, and LA would have 2 more.

[dino7c]

for some strange reason perimeter shots at the goalies chest aren't as useful as point blank shots in the slot

[dino7c]

They left three points on the table against the Oilers and one against the Coyotes. Give them those four points, and they're tied with the Flames right now.

If they could have held their two third period leads at home against the Flames, Calgary would have 4 fewer points, and LA would have 2 more.

ifs and buts

if Calgary didn't lose 8 straight they would be giving the Ducks a run

[Locke]

for some strange reason perimeter shots at the goalies chest aren't as useful as point blank shots in the slot

So Sayeth Jiri Hudler, The Kingslayer.

[Imported_Aussie]

Yet still the advanced stats devotees rank LA high:
http://www.nhl.com/ice/news.htm?id=762471&navid=nhl:topheads

We need an advanced stat for shot quality ASAP - treating all shots the same is ludicrous, the NBA doesn't do it, which is why more advanced metrics there make more sense.

[Jason14h]

ifs and buts

if Calgary didn't lose 8 straight they would be giving the Ducks a run

IF we had missed the playoffs that 8 game stretch would be the reason why. People are asking WHY they missed. These are the reasons

[Hemi-Cuda]

Watching the Kings in Edmonton and Calgary, they just looked like an extremely tired group. Stanley Cup in 2012, conference finals in 2013, another cup run in 2014 with 3 series going 7 games. And they had a trip to Europe sprinkled in there somewhere. No one has played as much hockey as the Kings have the last 3 years, they just looked done

With a long off-season to heal and recharge I think they look like a much different team in October

[sharkov]

Corsi numbers are so overrated

[Resolute 14]

Eventually advanced stats nerds will accept that not all shots are created equal.

No, but when you are looking at 1600-1900 shots (at ES) over the course of the season, that usually does not matter, no matter how much Flames fans want to pretend otherwise.

[Fransken]

Key players like Doughty and Kopitar playing about 350 games the last three calendar years with both the WC and Olympics. That equals 3 games a week, which is about regular season pace, for three straight years in (on?) average. This along with Kings psychical style of play might result in a really fatigued team and a reason for them missing the playoffs.

[flylock shox]

Didn't the Kings only win about 3 of 18 games they took to overtime?

If that's right, they got 6 points less from those games than average, enough to cost them a spot.

But I think the real reason they missed is that they're just tired. 3 long seasons in a row, knowing how hard it is to get back to the finals, I don't think they had the same heart for it this year.

[dino7c]

I suspect the Kings will be a bubble team again next year. Their players aren't getting any younger, they have cap issues, Richards is done in this league...Brown is nowhere near the player he once was.

heck they were a bubble team the first time they won the cup, if the Flames could have won a shootout to save their lives that year they would have missed

[SuperMatt18]

Couldn't score goals.

Which led to an unusual amount of 1 goal and OT games.

That they happened to lose the majority of.

[Lanny'sDaMan]

You wanna see funny? Go check out the super 16 on nhl.com. The have the Kings in 10th right now. RIGHT NOW. AFTER last night's games.

I was crying I was laughing so hard reading that tripe. Some advance stats guys just refuse to look beyond their excel sheets.

[Gaudfather]

So Sayeth Jiri Hudler, The Kingslayer.

His 2014/15 shooting % now sits at 19.6% - the Czech Assassin!

[Resolute 14]

Not just that they couldn't score goals, but that they were a hot mess on the road in the first half of the season. They turned it around for a while, but going 0-4-1 in their last five road games sealed it. That points to the common argument that they just didn't have that extra gear this year. The Kings couldn't get pumped on their own - they needed their crowd to bring that energy for them.

[undercoverbrother]

Will Corsihockeyleague clear up this question?

[Incogneto]

Kings missed the Playoffs? Of Corsi they did!

[Reggie Dunlop]

Because they're no good.

[Red]

Wow, this guy sure brings a lot of corsi in to this. And unlucky in OT/Shootouts and one goal games? Sutter coaches for 1 goal games. Remember Kipper specials? 2:1 wins. That was not luck.
I think the team was tired mentally and physically. Just not enough drive to win in regulation and by the time they made OT they were out of breath and hardly interested. Most of these guys already accomplished everything they dreamed off, hard to motivate them any more. Nothing to do with luck. And corsi being down are a reflection of lost drive.

[Finger Cookin]

Only pulling out 3 wins in 18 OT/SO games definitely had something to do with it. 9 of the 15 losses were in games where the Kings even scored first.

[GGG]

Losing voynov and Mitchell made them a bubble team which left them subject to randomness. There overtime and shoutout record was bad enough to have them miss out.

When people hold up this as an incitement of Corsi it's kind of laughable as this kings team has won two cups playing a high possession style. Corsi shows them to be a good team and they are a good team. The difference between Vancouver, Calgary, kings, Jets and wild is within the margin of error of any prediction. So re-role this season and a different one of those teams would miss the playoffs.

[Enoch Root]

No, but when you are looking at 1600-1900 shots (at ES) over the course of the season, that usually does not matter, no matter how much Flames fans want to pretend otherwise.

Pretend? Yes shot quality will typically average out for a typical team.

But Flames fans aren't pretending anything . You watched the games too and you saw how they played - both offensively and defensively. When it comes to the Flames, there is no question that not all shots are created equal.

[Cali Panthers Fan]

I will say this:

Seeing what Mitchell brought to the Panthers this year in terms of leadership, grit, heart, and very dependable play in his own zone, it's no wonder that the Kings seemed to be missing those elements at times this year. I think his loss matters a lot more than anyone would have considered, even if purely from a chemistry standpoint. The Florida Panthers say thank you for letting Mitchell walk.

[Geeoff]

Corsi is a good stat that some people use to draw bad conclusions from.

In my opinion, the Kings missed the playoffs because they have lost more games than they won.

[Pointman]

If you go by goals difference, Kings would be in playoffs too. And goals are pretty conventional stat. Or is it garbage now too, because, just like corsi, goals difference doesn't always tell you who will make playoffs?

[AcGold]

They lost because today's game wears players down to the point that you can't cover up errors with cheap shots and hooking. The Kings have played hockey into the middle of summer every year and they are just dog tired. Don't be surprised if they win the cup next year once they've got a rest.

[Mattman]

By being worse than No Good

[FlamesAddiction]

A few things to consider:



The Kings finished 6th, 5th, 8th, 7th and 6th in the conference the last 5 years. Despite being a good playoff team, they are not an elite regular season team. It really shouldn't be a shock that they fell short finally.
They faced some adversity this year with the Voynov situation. That, combined with bad cap management did not give them a lot of flexibility.
Darryl Sutter is an awesome coach in the short term, but as a long term coach, there are some questions.

[killer_carlson]

A few things to consider:



The Kings finished 6th, 5th and 8th in the conference the last 3 years. Despite being a good playoff team, they are not an alite regular season team.
They faced some adversity this year with the Voynov situation. That, combined with bad cap management did not give them a lot fo flexibility.
Darryl Sutter is an awesome coach in the short term, but as a long term coach, there are some questions.



Right. 2 cups in 3 years and there are questions.

:rolleyes:

[FlamesAddiction]

Right. 2 cups in 3 years and there are questions.

:rolleyes:

Most coaches have a shelf life. Sutter even admitted that himself when he stepped down as the Flames coach.

Some coaches get diminishing returns sooner than others. Two cups in 3 years after joining the team doesn't suggest otherwise.

[Northendzone]

last night on the drive home - wills and lubo mentioned that the kings were something like 3-15 in bonus time - turn a couple of those into W's and they are in the mix again.

saw a few writers mentioning over the last week or two that if the kings could get in the playoffs they were favourites to win

[Resolute 14]

Pretend? Yes shot quality will typically average out for a typical team.

But Flames fans aren't pretending anything . You watched the games too and you saw how they played - both offensively and defensively. When it comes to the Flames, there is no question that not all shots are created equal.

Yes, on a single-game scale, it can make a huge difference. So too can a goalie standing on his head or fumbling the puck like an idiot.

But over 82 games, those largely smooth out.

[Locke]

Right. 2 cups in 3 years and there are questions.

:rolleyes:

There always have been. Coaches like Sutter cant last forever. They're motivational coaches that are extremely demanding, but you cant keep that up indefinitely, eventually it wears off or gets tuned out.

[White Out 403]

The corsi hate here is akin to watching someone on Fox news walk out into a snow storm and claim there's no global warming.

It's a complete misrepresentation of what corsi is, and it's becoming really difficult to read people pile on the Kings as if some how that makes advanced stats useless.

It's a predictive tool that's no where near perfect and has never been claimed to be.

[EldrickOnIce]

^ I guess I need to read info on this.
Because to me it sounds like saying over the course of a season all players should have the same shooting % - that it all evens out.

[Resolute 14]

^ I guess I need to read info on this.
Because to me it sounds like saying over the course of a season all players should have the same shooting % - that it all evens out.

Not quite. It would even out somewhat for the same reason why PDO evens out - regression to the mean. But highly skilled shooters will tend to score at higher rates and poorly skilled less.

But over 82 games and almost 2000 shots, every single team is going to generate a great deal of high quality chances, and a great deal of low quality chances. And in general, a team that is generating more chances overall, is going to have more high quality chances. Likewise, a team that gives up fewer chances will give up fewer high quality chances.

As has been the case all season long, the simple fact is that exception teams exist and are inevitable. Both Calgary and Los Angeles are such exceptions.

[Regorium]

The corsi hate here is akin to watching someone on Fox news walk out into a snow storm and claim there's no global warming.

It's a complete misrepresentation of what corsi is, and it's becoming really difficult to read people pile on the Kings as if some how that makes advanced stats useless.

It's a predictive tool that's no where near perfect and has never been claimed to be.

I think there's problems in its predictive capability because of the underlying assumptions.

Basically, when people say that Tampa Bay's PDO is well above 100. Then the responders will be like "of course, they have Stamkos, Johnson, Kucherov, Bishop etc."

But that just means that PDO is being used to reinforce a preconceived notion that Stamkos/Johnson/Kucherov/Bishop are good players. PDO is not predicting anything in this case.

Corsi is very similar. "Oh the Ducks have awful CF%, but they have Getzlaf and Perry!" - again, the outliers are explained away through just naming players that people know are good. All without looking at WHY the Ducks can continually buck the trend.

I personally believe that Calgary is going to consistently buck the trend. Johnny Gaudreau himself is the epitome of an anti-corsi player. He makes one move on Doughty, then fires a pass through his legs to an open Hudler for an insane scoring chance. They lose the puck. This happens consistently. Game after game, shift after shift. Yet Corsi continues to predict that JG is an awful player.

[GGG]

^ I guess I need to read info on this.
Because to me it sounds like saying over the course of a season all players should have the same shooting % - that it all evens out.

It's more like saying all teams regress to the same shooting percentage.

There are a few players with the ability to maintain higher shooting percentages year over year. Most average out overtime. And when you take 22 players over multiple seasons it starts to average out as well.

Corsi also assumes that shot selection is equal among all teams and thus far any analysis to prove otherwise has fallen short. But there are some neat concepts like the royal road that suggest otherwise.

I think Corsi works because the league in general adopted Mike Babcocks possession based style of play. So if all teams play a similar style the measuring Corsi works as the base assumption that over time all teams take the same quality of shots plays out to be true.

[dino7c]

The corsi hate here is akin to watching someone on Fox news walk out into a snow storm and claim there's no global warming.

It's a complete misrepresentation of what corsi is, and it's becoming really difficult to read people pile on the Kings as if some how that makes advanced stats useless.

It's a predictive tool that's no where near perfect and has never been claimed to be.

except when clowns in the media rank the Flames low all year because of it no matter what the results on the ice are. People around here are tired of it and I don't blame them.

unsustainable? Flames have been doing it for a year and a half

had the Flames missed the playoffs with 96 points or whatever the "I told ya so" guys would be coming out of the woodwork

[DuffMan]

When I was watching Regher giving Hudler some cross checks in front of the net, I was thinking that in the old days, LA would've just punked us into submission. I blame the new rules for LA's demise. They are big and tough but can't score, and they were worried about penalties., That's another thing, the Flames haven't taken a penalty in 3 games or something like that, Having a PK that goes 100% in the 3 most important games helps.

[Resolute 14]

I personally believe that Calgary is going to consistently buck the trend. Johnny Gaudreau himself is the epitome of an anti-corsi player. He makes one move on Doughty, then fires a pass through his legs to an open Hudler for an insane scoring chance. They lose the puck. This happens consistently. Game after game, shift after shift. Yet Corsi continues to predict that JG is an awful player.

No it doesn't. It, in fact, argues the complete opposite. Also, don't cherry pick one stat in the very same post where you essentially attack those you disagree with for cherry picking one stat.

The Flames as a team give up far more SATs than they generate, so every player is likely to be lower. That is not an argument that "player x is bad". In the case of Gaudreau specifically, his Corsi For relative to his own teammates was +5.04 and Corsi Against relative was -2.95. What these are saying is that Gaudreau is one of the best players on the team at driving possession.

Bluntly, the Flames will not "consistently buck the trend", because the Flames will not consistently get career seasons from half the team. What will happen - hopefully - is that we will continue to develop and get better on both sides of the puck. Over time, we should be creating more chances and surrendering fewer. We likely won't be a Kings-esque Corsi darling so long as our gameplan is based around counterattacking, but we should be able to get to 50%+ rather than the current 46.

[automaton 3]

Fatigue might be one factor, but overall the Kings looked like a slow team last night.

It will be interesting to see what changes happen there in the off season.

[Rick M.]

The Kings are getting old and slow. Last year Toffoli and Pearson gave them an injection of youth and speed which papered over their weaknesses. Toffoli has been unable to score lately and Pearson is out with injury. The defenceman they got at the trade deadline (Sekara) to make up for the loss of Mitchell and Voynov is out with injury.

[burnitdown]

The Johnny hat trick game answers the question. If he doesn't score one of those goals, we'd be tied at 95 points right now.

[Cleveland Steam Whistle]

The Kings missed because they were a worse than average 4 on 4 / shootout team and lost more than their share of extra points.

The Flames made it because they are a better than average 4 on 4 team, and won more than their share of extra points.

[burn_this_city]

Too much time at the beach and gallivanting around with celebrity babes.

[Regorium]

No it doesn't. It, in fact, argues the complete opposite. Also, don't cherry pick one stat in the very same post where you essentially attack those you disagree with for cherry picking one stat.

Sorry which stat am I cherry picking? The fact that I don't believe PDO is predictive (and that it only reinforces what people believe about the teams themselves) or the fact that I don't think Corsi is adaptable? (based on the fact that Anaheim is bucked the trend for 4+ seasons?)

The Flames as a team give up far more SATs than they generate, so every player is likely to be lower. That is not an argument that "player x is bad". In the case of Gaudreau specifically, his Corsi For relative to his own teammates was +5.04 and Corsi Against relative was -2.95. What these are saying is that Gaudreau is one of the best players on the team at driving possession.

In fact it is. If Gaudreau is an excellent player according to Corsi, he should have a CF% of 55%+. Not relative to his teammates, but relative to the entire league. The fact that he's better than his teammates, in theory would mean that's he's just the best out of an awful bunch. Bringing in team-based comparison means that you also believe that there are systemic differences in team play that lead to these discrepancies.

Bluntly, the Flames will not "consistently buck the trend", because the Flames will not consistently get career seasons from half the team. What will happen - hopefully - is that we will continue to develop and get better on both sides of the puck. Over time, we should be creating more chances and surrendering fewer. We likely won't be a Kings-esque Corsi darling so long as our gameplan is based around counterattacking, but we should be able to get to 50%+ rather than the current 46.

There are many things that we can continue to excel at that are not as luck-based as Corsi would like us to believe:
- 4v4 play can stay impressive. With players that excel with more ice, this can be reproducible.
- Continue to stay disciplined and have a high PP-PK differential. Again, small shifty forwards and a disciplined defence. Corsians assume that all teams should be penalized the same.
- Shot blocking. Having the NHL's best shot blocker in Kris Russell at some point will be seen as a positive rather than a negative.

[CorsiHockeyLeague]

In fact it is. If Gaudreau is an excellent player according to Corsi, he should have a CF% of 55%+. Not relative to his teammates, but relative to the entire league. The fact that he's better than his teammates, in theory would mean that's he's just the best out of an awful bunch. Bringing in team-based comparison means that you also believe that there are systemic differences in team play that lead to these discrepancies.
Yeah, you clearly have absolutely no idea what you're talking about.

There are ten skaters on the ice. One guy cannot fully offset the effect of the other nine. If you put Pavel Datsyuk on the ice with a bunch of twelve year olds against the Chicago Blackhawks, his possession numbers will be terrible in spite of his individual possession ability being well above average.

That should be obvious to anyone who's spent thirty seconds thinking about it.

[PlayfulGenius]

The league has so much parity, there's just not a lot to separate most teams...I suspect the Kings were a tad tired and lacking just enough 'fire' to bring them back to the pack.

My money says they rest up this summer and come back as a dominant, hungry team next season.

[DuffMan]

Because they couldn't beat the oilers.

[Regorium]

Yeah, you clearly have absolutely no idea what you're talking about.

There are ten skaters on the ice. One guy cannot fully offset the effect of the other nine. If you put Pavel Datsyuk on the ice with a bunch of twelve year olds against the Chicago Blackhawks, his possession numbers will be terrible in spite of his individual possession ability being well above average.

That should be obvious to anyone who's spent thirty seconds thinking about it.

I am extremely well versed in Corsi thank you very much. I offered a ridiculous scenario to get your brains moving on possible things that Corsi might be missing. But it's very obvious that you continue to maintain that Corsi is foolproof and that "luck and variance" can explain everything that doesn't fit the overall narrative.

I am well aware that it's ridiculous to use Corsi (relative to his own team or not) to describe a player. I'm just starting to come around to the fact that using Corsi to describe a team is just as bad.

[CorsiHockeyLeague]

But it's very obvious that you continue to maintain that Corsi is foolproof and that "luck and variance" can explain everything that doesn't fit the overall narrative.
If it's so obvious you must be able to quote an instance where I've said that.

I'd honestly like this forum a lot better if you could be banned for straw men. This kind of garbage makes it absolutely impossible to have a reasonable conversation.

[Tiger]

If it's so obvious you must be able to quote an instance where I've said that.

I'd honestly like this forum a lot better if you could be banned for straw men. This kind of garbage makes it absolutely impossible to have a reasonable conversation.




There are ten skaters on the ice. One guy cannot fully offset the effect of the other nine. If you put Pavel Datsyuk on the ice with a bunch of twelve year olds against the Chicago Blackhawks, his possession numbers will be terrible in spite of his individual possession ability being well above average.



Is this an example of a Straw man argument that should be banned?

[opendoor]

The Kings did tend to have a lower number of scoring chances per shot attempt compared to other teams, but even strictly looking at scoring chances they were #6 in the NHL in SCF%.

But really, by any measure they were an elite 5-on-5 team. #1 in Corsi, #1 in shot differential, #4 in goal differential, and #6 in scoring chance differential. All this "all shots aren't created equal" is a red herring because even strictly looking at goals scored, LA was a fantastic 5-on-5 team.

But unfortunately for the Kings, aggregate 5-on-5 stats (whether they be possession numbers or goal numbers) don't necessarily always translate into success in the standings, especially if you lose as many 1 goal games as LA did. LA played 36 one goal games and only won 12 of them. If they had won half of those games, as you'd expect any reasonably average team to do, they'd easily be in the playoffs. But they didn't, so they're out. Still a great team though.

[Oil Stain]

Only pulling out 3 wins in 18 OT/SO games definitely had something to do with it. 9 of the 15 losses were in games where the Kings even scored first.

That's probably the biggest reason right there. OT and SO wins are closer to a coin flip than deciding games in regulation.

If the Kings had even gone .500 in OT and the SO they'd have finished with a 100 points and everyone would be talking about how unbeatable they look going into the playoffs.

[OutOfTheCube]

I'll tell you why.

Right here: http://sabres.nhl.com/club/news.htm?id=743265

We only had 15 regulation wins this year. Happy to help, friends!

[Resolute 14]

I am extremely well versed in Corsi thank you very much. I offered a ridiculous scenario to get your brains moving on possible things that Corsi might be missing. But it's very obvious that you continue to maintain that Corsi is foolproof and that "luck and variance" can explain everything that doesn't fit the overall narrative.

And we are done here. If you lack the wit and intelligence to argue what is being said, and instead feel the need to invent opinions and assign them to others, then you are just wasting your time and mine.

And no, it is quite evident that you are not remotely versed in Corsi. Rather, you are taking a very primitive understanding of it and either through ignorance (which is curable if you so desire) or deliberately (and disingenuously) mis-applying out of context to try and prove a point.

Problem is, the point you are proving is not what you think.

[EldrickOnIce]

It's more like saying all teams regress to the same shooting percentage.

There are a few players with the ability to maintain higher shooting percentages year over year. Most average out overtime. And when you take 22 players over multiple seasons it starts to average out as well.

Corsi also assumes that shot selection is equal among all teams and thus far any analysis to prove otherwise has fallen short. But there are some neat concepts like the royal road that suggest otherwise.

I think Corsi works because the league in general adopted Mike Babcocks possession based style of play. So if all teams play a similar style the measuring Corsi works as the base assumption that over time all teams take the same quality of shots plays out to be true.
Thank you. I appreciate the response itself, and the tone of it. Often we who question legitimately get a you are dumb response.
I am very interested in the Royal road work, I expect it will prove the point in question, to some degree.
Also honed in on your 'play the same game' statement. This rings true. The Oilers were a great example. A swarm defence that gives up golden chances plus a coaching strategy to shoot from everywhere (because all chances are equal in the end) created a situation that left you with underlying numbers suggesting improvement that was obviously not going to happen, because they weren't playing 'the right way'. But what we got from most was 'bad luck' - when it was nothing of the sort.
The Flames don't play nice either. It doesn't predict them well. Yes, they need to improve. A lot. But it's clearly more sustainable than the model would have us believe.
That's only my opinion

[Geeoff]

What's there to fool about Corsi as a stat? It's a stat that measures a specific thing: shot attempts. It's about as subjective as any other stat based on shots. The idea that corsi should have a 1:1 relationship with the standings is what is foolhardy.

[Enoch Root]

Yes, on a single-game scale, it can make a huge difference. So too can a goalie standing on his head or fumbling the puck like an idiot.

But over 82 games, those largely smooth out.

I think you mis-read my post. I was referring to the Flames' season not a single game.

The Flames have been doing the same thing for a year and a half. If you break it down into 20-game segments, they have been surprisingly consistent from segment to segment. There is not only a large sample size to observe, the data is all very robust (consistent).

What they are doing, offensively and defensively, they have been doing for more than 100 games. Corsi, shooting percentage, shot blocking, everything.

To continue to argue that it is unsustainable and due for regression is to simply ignore the evidence. Pure cognitive dissonance.

[AcGold]

No it's not cognitive dissonance, it's ignorance. Big difference.

[Regorium]

@ Resolute 14. I'm not arguing effectively because I used to be a huge Corsi supporter (early this year), and that side frustrates the hell out of me. That's fair and I deserve to be called out on it.

I posed 3 questions in my previous post:
- Why does Corsi assume that PP-PK has to be consistent for all teams, and thus remove PP/PK from its analysis? (one of the biggest arguments for the Flames being "lucky" this year despite their bad Corsi numbers, they've been lucky with officiating)
- Why are 4 on 4 results are reduced to luck?
- Does having the NHL's best shot blocker on your team induce enough of a change that shot blocking becomes a net positive? Can shot-blocking itself be a net positive?

There are so many things about the current Flames team that the standard Corsi base of assumptions just does not take into account. However, from the pro-Corsi side, it's purely luck and variance, and regression is inevitable. There's no analysis as to why the Flames have been successful - it's all just luck. That's why it's so frustrating.

The fact that LA had so many one goal games, rather than just dominating their opposition (which they should've been, based on their CF%) shows that perhaps LA wasn't as good of a team as their CF% suggested. Over the last 3 years, this has happened every time (246 game sample), yet we ignore that data set and only focus on the ~65 games of the playoffs. Maybe there's something else at play.

[GGG]

Thank you. I appreciate the response itself, and the tone of it. Often we who question legitimately get a you are dumb response.
I am very interested in the Royal road work, I expect it will prove the point in question, to some degree.
Also honed in on your 'play the same game' statement. This rings true. The Oilers were a great example. A swarm defence that gives up golden chances plus a coaching strategy to shoot from everywhere (because all chances are equal in the end) created a situation that left you with underlying numbers suggesting improvement that was obviously not going to happen, because they weren't playing 'the right way'. But what we got from most was 'bad luck' - when it was nothing of the sort.
The Flames don't play nice either. It doesn't predict them well. Yes, they need to improve. A lot. But it's clearly more sustainable than the model would have us believe.
That's only my opinion

I agree with you that the flames style does not fit what Corsi models based on the eye test. The flames have also maintained a 96 point pace for 110 games so at some point it has to mean something. They weren't a team like Minnisota, Toronto, or Nashville that got off to really hot starts and then faded to the back of the pack. So the question is why? What's funny if next year they finish in 9th you will have people proclaim that they regressed when really it would be the same result.

I think where the discussion around advance stats breaks down is that people try to create a black and white approach. So people who like advance stats characterize the other group as idiots and people who dislike them use any deviation from 100% predictive value to try to discredit the whole thing.

People respond to the worst comments instead of the well thought out discussion you get a mess.

[CorsiHockeyLeague]

Is this an example of a Straw man argument that should be banned?

Look up "straw man".

[Resolute 14]

@ Resolute 14. I'm not arguing effectively because I used to be a huge Corsi supporter (early this year), and that side frustrates the hell out of me. That's fair and I deserve to be called out on it.

Fair enough, and thanks. I can, of course, only answer your questions with both what knowledge I hold, and my own (hopefully logical) considerations:

I posed 3 questions in my previous post:
- Why does Corsi assume that PP-PK has to be consistent for all teams, and thus remove PP/PK from its analysis? (one of the biggest arguments for the Flames being "lucky" this year despite their bad Corsi numbers, they've been lucky with officiating)

I think you are assuming too much here. To the best of my knowledge, special teams aren't typically excluded because of an assumption that they are constant for all teams, but that they make up such a relatively small percentage of the game and the very nature of a man (dis)advantage skews the intent of trying to compare teams in relatively equal situations.

Remember that Corsi is basically an "advanced plus/minus", which likewise excludes PP scoring.

- Why are 4 on 4 results are reduced to luck?

I'm not sure it is reduced to luck, but rather has such a small sample size that it is basically just statistical noise. i.e.: We've played 3822 minutes at 5 on 5, but only 139 at 4 on 4. Also, I presume that the "luck" argument is more likely a belief that 4 on 4 success generally has not been demonstrated as consistently repeatable. Much like OT/SO.

- Does having the NHL's best shot blocker on your team induce enough of a change that shot blocking becomes a net positive? Can shot-blocking itself be a net positive?

Depends on your view, I think. Personally, I do find shot blocking to be a positive - you can't score on shots that don't reach the net - but since Corsi is only attempting to measure shot attempts, the argument about blocking isn't a flaw in the stat itself.

There are so many things about the current Flames team that the standard Corsi base of assumptions just does not take into account. However, from the pro-Corsi side, it's purely luck and variance, and regression is inevitable. There's no analysis as to why the Flames have been successful - it's all just luck. That's why it's so frustrating.

If you've read my posts in the past, you would find that I have consistently noted the idiocy of relying too much on one stat. Or even one stat in one context - thus my rebuttal of showing Gaudreau's Corsi relative to his teammates. And yes, many analytics proponents tend to lose sight of the forest for the trees. That is not a flaw in the statistic itself, but in the humans creating the predictions.

[Resolute 14]

I think you mis-read my post. I was referring to the Flames' season not a single game.

The Flames have been doing the same thing for a year and a half. If you break it down into 20-game segments, they have been surprisingly consistent from segment to segment. There is not only a large sample size to observe, the data is all very robust (consistent).

What they are doing, offensively and defensively, they have been doing for more than 100 games. Corsi, shooting percentage, shot blocking, everything.

To continue to argue that it is unsustainable and due for regression is to simply ignore the evidence. Pure cognitive dissonance.

Fair enough, I did misread the intent of your post. But while we consider that the Flames have been doing well for about 100 games, lets not forget that for how well we have played, we are still only barely a playoff team. We need to get better still, and that will reflect in a myriad of statistical areas, including Corsi.

[Enoch Root]

Fair enough, I did misread the intent of your post. But while we consider that the Flames have been doing well for about 100 games, lets not forget that for how well we have played, we are still only barely a playoff team. We need to get better still, and that will reflect in a myriad of statistical areas, including Corsi.

no question

[PugnaciousIntern]

Oh look, another discussion about corsi!

[Tiger]

Look up "straw man".

It is when you take an argument. Make it more extreme and stupid but the same premise and then prove your right by saying the extreme case is wrong.

Datsyuk playing with 12yr olds seems like a extreme case when it was used to argue gaudreau with other nhlers.

Please explain if it is wrong

[mikeecho]

We need to get better still, and that will reflect in a myriad of statistical areas, including Corsi.
And the nice part is that the Flames are in year 2 of a rebuild. It's a fair assumption that the team will get better.

[Red]

Here is what I don't get.

You have to score more goals than the opposition to win.
If corsi is correct, then high possession = lots of goals. So all good corsi teams should be high scoring and have wins to show for it.

How do we explain NJ, Cgy, LA etc? LA never scored a lot yet they always had great corsi. Flames are one of the highest scoring teams this year. NJ has good corsi but is the 3rd lowest scoring team. Isn't Dallas also a Corsi darling? Scores a ton, but can't win.
It just doesn't add up.

[EldrickOnIce]

The go to response is PDO (luck).

[sureLoss]

Here is what I don't get.

You have to score more goals than the opposition to win.
If corsi is correct, then high possession = lots of goals. So all good corsi teams should be high scoring and have wins to show for it.

How do we explain NJ, Cgy, LA etc? LA never scored a lot yet they always had great corsi. Flames are one of the highest scoring teams this year. NJ has good corsi but is the 3rd lowest scoring team. Isn't Dallas also a Corsi darling? Scores a ton, but can't win.
It just doesn't add up.

Those who believe in Corsi would tell you that Calgary has an extremely high shooting percentage (i.e. shots that turn into goals) that is unsustainable. In other words, luck. They would argue that eventually the shooting percentage will fall and goals scored would drop.

[Vinny01]

They played 50+ playoff games over the past 3 years this team has to be gassed. Voynov was a huge loss. Richards continuing his decent to pure garbage, prized deadline pickup Sekera getting hurt all big factors.

Also I got the impression is the Kings thought they would get in no matter how bad they were playing. Get in and go far was their mindset but they were more often out than in. They didn't have that extra gear this year and it really showed in their recent 3 game western Canadian swing.

[transplant99]

It's pretty simple why the Kings missed this year.

They are gassed as a collective.

The Kings have played 65 more games the last 3 seasons than the Flames as an example. Those games are also among the hardest played typically both mentally and physically. That's almost an entire season when you tack on Olympics for many of them.

The human body can only take so much before it simply can't respond like it wants. It's like any of us if we worked day after day and week after week of overtime. At some point you just start to burn out and are not as effective as you can be.

It really is as simple as that....Corsi and PDO and all the other fluffy explanations aside.

[EldrickOnIce]

They played 50+ playoff games over the past 3 years this team has to be gassed. Voynov was a huge loss. Richards continuing his decent to pure garbage, prized deadline pickup Sekera getting hurt all big factors.

Also I got the impression is the Kings thought they would get in no matter how bad they were playing. Get in and go far was their mindset but they were more often out than in. They didn't have that extra gear this year and it really showed in their recent 3 game western Canadian swing.
The way they had played on their previous road trip, one might have expected 4-6 points.
They did look gassed, and got 1.

[dieHARDflameZ]

In my opinion they've just played way too much hockey over the past few seasons. I don't think they had enough left in the gas tank.

Make no mistake, this team should be back in the playoffs next season contending for another cup.

[edslunch]

The go to response is PDO (luck).


PDO is the most meaningless of stats. If a team has a PDO if 100 are they:
A) totally average
B) lights out shooters with lousy goaltending
C) offensively inept but saved by an all world goalie
D) somewhere in between

The answer is Any of the above. What a helpful stat

[mikephoen]

No, but when you are looking at 1600-1900 shots (at ES) over the course of the season, that usually does not matter, no matter how much Flames fans want to pretend otherwise.

I'm having trouble buying this argument. If a team plays the same style all year with mostly the same personnel and that style causes them to take a lot of low quality perimeter shots, wouldn't thier 1600-1900 shots be of lower quality than another team that generates high quality shots on a regular basis?

[underGRADFlame]

The league has so much parity, there's just not a lot to separate most teams...I suspect the Kings were a tad tired and lacking just enough 'fire' to bring them back to the pack.

My money says they rest up this summer and come back as a dominant, hungry team next season.

This.

[EldrickOnIce]

They don't really belong together.
Combining them for 100 is just saying that a shooting % of 9 is about average, as is a save percentage of about 91%
So if you are lucky in one, the other, or both - that number over 100 is representative of 'luck'.
I don't buy it. Of one does buy it, you also have to say that The NY Rangers and Tampa Bay are as 'lucky' as the Flames are this year, too. I don't recall much talk about how lucky they are...

That said, I can not see the Flames maintaining the shooting percentage they are putting up this year

[Street Pharmacist]

Again, it's only accurate in context with obvious exceptions. The trouble is, with so many rookies you can't know if it's luck or exception.

It's been well established that save% is not reproducible except for a very few elite goaltenders. Same goes for shooting%. The trouble for the flames is the shooting %. Well with so many rookies/sophomore playing important roles, you can't say it's not sustainable because you don't know if it's their baseline.

And adding them up is silly

[Monahan23]

3 reasons

-couldn't rely on another defence-man other that Doughty
-They played with 0 sense of urgency
-they couldn't score

[EldrickOnIce]

Again, it's only accurate in context with obvious exceptions. The trouble is, with so many rookies you can't know if it's luck or exception.

It's been well established that save% is not reproducible except for a very few elite goaltenders. Same goes for shooting%. The trouble for the flames is the shooting %. Well with so many rookies/sophomore playing important roles, you can't say it's not sustainable because you don't know if it's their baseline.

And adding them up is silly

Not quite sure I understand what you mean, unless a team like that is the exception?
Tampa has been 1st, 3rd, 8th and 1st in the league in shooting % over the last 4 years. That seems reproducible , with the outlier being 8th (instead of higher).

[Sylvanfan]

I suspect that Boston and Pittsburgh are also teams suffering a bit from a compressed 2 and a half seasons. The 48 game lockout season was pretty compressed, last year with the Olympics was compressed, and these guys are feeling it this year.

I personally thought Jonathan Quick wasn't as good this year as he's been in previous seasons. He should be much fresher next season.

[edslunch]

In my opinion they've just played way too much hockey over the past few seasons. I don't think they had enough left in the gas tank.



Make no mistake, this team should be back in the playoffs next season contending for another cup.


The gas tank is probably as much emotional as physical. Been through the war three years in a row, won it twice. Even subconsciously is your heart still in it to the same extent?

[Resolute 14]

I'm having trouble buying this argument. If a team plays the same style all year with mostly the same personnel and that style causes them to take a lot of low quality perimeter shots, wouldn't thier 1600-1900 shots be of lower quality than another team that generates high quality shots on a regular basis?

No, because of the exact same argument being made to try and dismiss this: not every shot is the same. And over this many shots, everybody is taking massive amounts of both low quality and high quality shots.

The Kings took nearly 20% more shots at 5 on 5 than the Flames did. Obviously that means they took more low quality shots, but it probably also means they took more higher quality shots too. So while one could argue that the Flames skill level is higher than expected, which resulted in a higher shooting percentage than expected, it is hard not to believe the same should have been true of LA. I don't believe for one second that the difference between our offence and their offence is the idea that we eschew low quality chances for high. In fact, at 5 on 5, the Kings actually outscored us 148-145 - our abnormally high shooting percentage ate up that nearly 20% difference in shots, however.

[Enoch Root]

No, because of the exact same argument being made to try and dismiss this: not every shot is the same. And over this many shots, everybody is taking massive amounts of both low quality and high quality shots.

The Kings took nearly 20% more shots at 5 on 5 than the Flames did. Obviously that means they took more low quality shots, but it probably also means they took more higher quality shots too. So while one could argue that the Flames skill level is higher than expected, which resulted in a higher shooting percentage than expected, it is hard not to believe the same should have been true of LA. I don't believe for one second that the difference between our offence and their offence is the idea that we eschew low quality chances for high. In fact, at 5 on 5, the Kings actually outscored us 148-145 - our abnormally high shooting percentage ate up that nearly 20% difference in shots, however.

Well, the two most plausible explanations are:

a) the Flames are taking higher quality shots overall, or

b) the Flames are lucky

Over 82 games (and in fact well over 100 now), choosing 'b' is pretty silly. If 'a' didn't sit well with me, I would be looking for a 'c' before I started arguing for 'b'.

[Resolute 14]

False dichotomies don't strengthen the case for the option you prefer.

[GGG]

PDO is the most meaningless of stats. If a team has a PDO if 100 are they:
A) totally average
B) lights out shooters with lousy goaltending
C) offensively inept but saved by an all world goalie
D) somewhere in between

The answer is Any of the above. What a helpful stat

Pdo is useful in measuring the sustainability of sucess. Pdos of above 1.01 and below .990 are definately not repeatable year over year. So if you are outside of those bounds even with great goal tending or great shooting (assuming those exist) you will likely regress to between those two bands. The bands might be even narrower.

People hate pdo because people keep saying all teams should have a pdo of 1 and it's entirely random. But it isn't quite. Better teams will have marginally higher pdo.

[EldrickOnIce]

This is particularly interesting work

“It’s not that my first thought was Corsi is bogus, it’s just there needs to be a stage 2,” said Valiquette, now an analyst with the New York Rangers and a goalie coach. “The old train of thought was, ‘Pucks to the net, pucks to the net. Good things happen when you get pucks to the net.’ Actually (most of the time) it’s, ‘Pucks to the net, nothing happens.’ ”

The Royal Road
http://www.thehockeynews.com/blog/wp-content/uploads/2015/01/Royal-Road-e1422548236504.png

Valiquette doesn’t think about shots. He thinks about shot sequences, and said red shots are highly overvalued while green shots are hugely undervalued.
A red shot is one where a goalie has more than a half-second of clear sight on either side of the Royal Road. These shots require minimal movement. Valiquette claims NHL goaltenders, on average, stop 97 percent of red shots. Green shots are those where the puck crosses the Royal Road, either by the puck carrier or a passer, then are shot on goal. According to Valiquette’s research, 76 percent of goals so far this season have been on green shots.
http://www.thehockeynews.com/blog/whats-the-next-dimension-of-analytics-former-nhl-goalie-steve-valiquette-knows/

[Itse]

Bad stat reading is bad. The Kings are not a statistical aberration, and neither are the Flames. Anaheim is.

Long version:
If you look at what the goal differential suggest for each team in the west:
- STL, CHI and NSH are in a league of their own when it comes to goal differential, and this is pretty well reflected in the standings.
- CGY, LAK, VAN and WIN are relatively consistently close to each other throughout the season, both in points and in goal differential. Nothing to see here either.
- SJS hangs with the group above in points for the first half of the season, but for a large part their goal differential lags a bit (but not a lot) behind. In the second half of the season first the goal differential and then points start to drop clearly out of playoff competition numbers.
- For the first half, MIN goal differential and points put them clearly out of the playoff picture. Then something happens, and their record in the last 39 games is +46 goals and 59 points. "Something" in the case of Minnesota is not a mystery however.

Short version:
At the end of the season the top 7 teams in goal differential (STL, CHI, NSH, CGY, VAN, MIN, WIN) are all in the playoffs. The 8th in goal differential are LAK, who are 9th in standings.

The aberration here is Anaheim. Most of the season their goal differential suggests they should be in the same group with CGY, LAK, WIN and VAN, but their points keep having them in the same group with STL, CHI and NSH. From January until the end of the season they actually have a negative goal differential, but this did not stop them from ending up at the top of the conference.


So, the real question is not why did LA make it (they're a playoff bubble team, some of them don't make it), nor is it why did CGY make it (they're a playoff bubble team, some of them make it, and CGY was the better of those teams for most of the season).

The question is, how come Anaheim has 107 points? Are they Corsi monsters?

[Buster]

No, because of the exact same argument being made to try and dismiss this: not every shot is the same. And over this many shots, everybody is taking massive amounts of both low quality and high quality shots.

The Kings took nearly 20% more shots at 5 on 5 than the Flames did. Obviously that means they took more low quality shots, but it probably also means they took more higher quality shots too. So while one could argue that the Flames skill level is higher than expected, which resulted in a higher shooting percentage than expected, it is hard not to believe the same should have been true of LA. I don't believe for one second that the difference between our offence and their offence is the idea that we eschew low quality chances for high. In fact, at 5 on 5, the Kings actually outscored us 148-145 - our abnormally high shooting percentage ate up that nearly 20% difference in shots, however.

This is an assumption. It also seems to be an assumption that is not proving valid as we draw more scrutiny to it.

Any statistical model must also take into consideration the sensitivity of an input. If the Flames are playing a style that results in a higher shot quality, it is possible that even a very minor difference in this regard can result in a significant difference over the course of a season with respect to points.

[d_phaneuf]

Slap Shots has been told by two sources that the Kings locked the door to their locker room following a defeat on the road within the last two weeks so that Sutter could not get in and deliver what the players apparently expected to be another in a series of lectures/tirades.
As the tale was told, after Sutter finally tracked down an arena operative to unlock the door, he was greeted by three heavy waste receptacles lined up as barricade to what had become an empty room.

http://nypost.com/2015/04/11/calendar-coach-exhaust-defending-champion-kings/

[jayswin]

I don't believe for one second that the difference between our offence and their offence is the idea that we eschew low quality chances for high. In fact, at 5 on 5, the Kings actually outscored us 148-145 - our abnormally high shooting percentage ate up that nearly 20% difference in shots, however.

From watching them play this season this seems to be exactly what the Flames do. Not a lot of chances, but the ones they do take are of high quality.

They seem more geared towards a "keep the pressure on until that high scoring area chance comes up and make it count" style rather than coming in and just looking to overload the other team with shots like most teams do.

[CorsiHockeyLeague]

It is when you take an argument. Make it more extreme and stupid but the same premise and then prove your right by saying the extreme case is wrong.

Datsyuk playing with 12yr olds seems like a extreme case when it was used to argue gaudreau with other nhlers.

Please explain if it is wrong

It is. A straw man refers to the practice of inventing an argument, attributing it to the person you're arguing against and then defeating it. Often easily recognized on the internet by posts beginning with the phrase "so what you're saying is..." Or some variant thereof. It's particularly obnoxious to have someone try to tell you what you think.

An extreme example used to disprove a general principle is not a straw man, it's a reductio ad absurdum which is a logically sound form of argument. Datsyuk with 12 year olds demonstrates how someone's corsi rel can be a more useful expression of their possession ability by adjusting somewhat for the other 9 skaters on the ice.

[Moneyhands23]

Lol, tsn " the reporters " are about to tell us!

[jg13]

Milbury is no good

http://i.imgur.com/CDpeRF8.jpg

[jayswin]

Wait, that didn't actually happen did it?

[mikephoen]

This is particularly interesting work



The Royal Road
http://www.thehockeynews.com/blog/wp-content/uploads/2015/01/Royal-Road-e1422548236504.png


http://www.thehockeynews.com/blog/whats-the-next-dimension-of-analytics-former-nhl-goalie-steve-valiquette-knows/

This article is awesome and really puts into words my own thinking on shot quality needing to be factored in.

[Resolute 14]

From watching them play this season this seems to be exactly what the Flames do. Not a lot of chances, but the ones they do take are of high quality.

They seem more geared towards a "keep the pressure on until that high scoring area chance comes up and make it count" style rather than coming in and just looking to overload the other team with shots like most teams do.

Actually, I would say the Flames style was more rope-a-dope, or a form of the Stampeders' philosophy of a 'bend don't break' defensive strategy. We spend a majority of our time chasing the puck, but have done well to keep the opponents to the outside, then using our speed to create a counterpunch offence.

Becoming a "keep the pressure on until that high scoring area chance comes up and make it count" would be the next logical progression for our play style, and it would carry with it an improvement in both traditional stats as well as those like Corsi and zone starts (only Buffalo started in their own defensive zone more often than we did).

[GioforPM]

The reason they missed the POs is Wideman to Raymond to Hudler, to the back of the net. :w00t:

Plus no Voynov and cap issues (caused by the Richards contract) to getting any decent replacements.

Plus a season series loss to the team that beat them out.

What I like is that this makes John Shannon wrong, when he said out of Calgary and LA, LA would win out, at a time when Calgary was ahead (he based it on the Sutter factor).

[EldrickOnIce]

^ Are zone starts strongly correlated with corsi?
I expect they are.

Excellent article here regarding writing off possession metrics failures simply as 'luck'.
http://hockeyanalysis.com/2015/04/12/kings-flames-avalanche-and-possession-analytics/

If you follow me I use possession metrics all the time, I just think that there is too much consideration for when possession metrics succeed in predicting outcomes and too little consideration of when it fails and when other metrics succeed.