BrokerLink

Saturday, August 08, 2015

Leafs, Avs, Flames: When Talking Regression, One of These is Not Like the Others

Severe thunderstorm and tornado warnings have been the storyline of the summer but gloomy forecasts are nothing new for Calgarians. As you'll recall, all winter the call was for dark and cloudy skies for the Flames.

Threats of an upcoming regression began in early November after Calgary shocked the hockey world by getting off to a blistering 8-4-2 start. The cautions continued in early December as the Flames improved to 17-8-2. The same pending bleak outlook was reiterated in January when Calgary's record slipped to 21-18-3 and the alerts intensified in February when Mark Giordano was lost for the year with the Flames record at 33-24-4.

However, that vicious storm, the thunder, the lightning, the funnel cloud, which Flames fans had been hunkered down and bracing for never materialized. No hail. No rain. Just that stiff gust of northern wind in December when the club went 0-7-1.

Instead, Bob Hartley's club kept rolling along under mostly sunny skies, qualified for the Stanley Cup playoffs to temporarily silence the skeptics, then kept it going by eliminating the Vancouver Canucks in six games to advance to the second round for just the second time since the invention of the thermometer.

However, as soon as Calgary was dispatched rather handily by the Anaheim Ducks in five games, the predictions of impending gloom started up once again. Sorry Flames, you may have dodged the storm last season but the dark skies will be back and this time you're going to get drenched.

Or, will they?

That's the million dollar question I'm here to dig into. Are the Flames going to regress in 2015-16 as many contend or has the team, bolstered by personnel moves made by general manager Brad Treliving, put itself in a position to ward off those threats?


The Case Against Calgary

At the heart of the argument against the Flames being able to repeat the season they just had is a pair of advanced stats commonly referred to as Corsi and PDO.

The gist being when the opponent consistently has possession of the puck more often than you and is shooting the puck more often than you, it's difficult to win games consistently. To do so, you require an element of good luck, which anyone that's ever visited a casino before can attest is fleeting and is the furthest thing away from being something that is sustainable.

The two recent examples most frequently cited are the Toronto Maple Leafs and the Colorado Avalanche.
  • In the lockout-shortened 2012-13 season, the Maple Leafs overcame very poor analytics to make the playoffs for the first time in eight years. Then, in round 1, they took Boston to game 7 before spectacularly blowing a 4-1 lead with less than 11 minutes to go and losing 5-4 in overtime.
  • In 2013-14, the Avalanche -- saddled with well below average analytics -- not only bucked convention but kicked it into submission by finishing third overall. They also went the distance in round 1 of the post-season before blowing four leads in game 7 and losing to Minnesota by an identical 5-4 score, also in overtime.

In both instances, each club followed up their advanced stats-defying success by crashing back to earth the next season and missing the playoffs.

But First, What Exactly is Corsi and PDO?

When it comes to advanced stats, there is a myriad of ways to slice and dice the data gleaned from NHL game stats sheets. For the purpose of this article, I'm sticking to two of the more common metrics known as Corsi and PDO. Both of these are tracked officially by the NHL now as part of the league's enhanced stats and are known by the names SAT% and SvSP%.
  • SAT% (aka Corsi) - This measures how many more shot attempts a team makes compared to shot attempts surrendered. Measured at even-strength, attempted shots includes goals, saves, missed shots and blocked shots. Since you need to have the puck in order to shoot the puck, this stat is generally referred to as possession. A number above 50 is good while below 50 is bad. While there are always the rare exceptions, SAT% has a good track record of demonstrating who are the NHL's best overall teams. 
  • SPSv% (aka PDO) - This adds together 5-on-5 shooting percentage plus 5-on-5 save percentage. The general principles at play here is both of these stats over large enough sample sizes will end up being very similar across the league. This stat is frequently referred to as luck. e.g. If you win 2-1 despite being outshot 40-20, that reflects good 'puck luck'. A number above 1000 indicates good luck while below 1000 infers bad luck. The theory being each number will work its way back towards 1000 as one's luck evens out.


With these definitions in mind, here is how the Leafs, Avs and Flames compared in their so-called unsustainable season. You can see why Calgary has been lumped in with the other two.

Note: Below I reference SAT% Close. How that varies from regular SAT% is 'Close' means it only includes data from games when teams are within one goal in the first and second period and tied in the third period. It should make it a more pure reflection of possession as it excludes the uncharacteristic more aggressive or more cautious playing style that sets in when a team leads or trails by a couple goals.


2012-13 Toronto Maple Leafs
> 44-29-9, 97 pts* (*48-game record of 26-17-5 extrapolated over 82 games)
  • Possession - 29th (43.45 SAT% Close)
  • Luck - 1st (1030 SPSv%)
> Following season: Regressed 13 pts (38-36-8, 84 pts), missed playoffs


2013-14 Colorado Avalanche
> 52-22-8, 112 pts
  • Possession - 27th (46.23 SAT% Close)
  • Luck - 3rd (1017 SPSv%)
Following season: Regressed 22 pts (39-31-12, 90 pts), missed playoffs


2014-15 Calgary Flames
> 45-30-7, 97 pts
  • Possession - 28th (45.06 SAT% Close)
  • Luck - 6th (1011 SPSv%)
Following season: TBD

Worth noting, last year only three teams made the playoffs with a SAT% Close below 50 -- Ottawa (49.39), NY Rangers (49.26) and Calgary (45.06).

As far as 'Luck' goes, that number is driven by the fact the Flames ranked 28th in shots on goal but were sixth in goals. Leading contributors to that unusually high team shooting percentage were Jiri Hudler (2nd, 19.6%), Sean Monahan (13th, 16.2%) and Lance Bouma (22nd, 15.4%), who all ranked in the top 25.


Countering Regression

When you've overachieved as the Flames did in 2014-15, one way to mitigate any regression the next season is to improve your hockey club.

There are several factors to take into consideration when assessing if a team is (or should be) improved or not. I've narrowed it down to six primary factors.
  1. Personnel Changes - Depending on who has departed and who has arrived -- via trade or free agency, teams can improve considerably, go backwards or tread water in the off-season.
  2. Getting Healthier - Is a team expecting full seasons from players that missed significant time the previous year.
  3. Coming Off a Career Year - Who had a year that was entirely unexpected and could likely fall-off just as easily.
  4. Developing Youth - How many young players are naturally going to be better the next season because they're in that growing and improving stage of their career. (I define as 25 and under).
  5. Rookies Inserted - Are there highly touted prospects ready to step in, become regulars and make an impact.
  6. State of Goaltending - The goaltender is the player that never leaves the ice. Good goaltending can mask other holes while poor goaltending can undermine strengths in other areas.

With these variables in mind, I took a closer look at both Toronto and Colorado after they defied the analytics in their 'unsustainable' season to see how they measured up going into that next year, in which both failed to avoid the regression that was predicted.

Then, for a comparison, I look at how the Flames are shaping up this off-season with opening night less than two months away.


Breaking Down the 2013-14 Toronto Maple Leafs

All stats are from 2012-13 unless otherwise noted.

1. Personnel Changes

Out

LW Clarke MacArthur (40 gm 8-12-20)
C Mikhail Grabovski (48 gm, 9-7-16)
C Leo Komarov (42 gm, 4-5-9)
RW Matt Frattin (25 gm, 7-6-13)
D Michael Kostka (35 gm, 0-8-8)
G Ben Scrivens (20 gm, 7-9-0, .915 SV%)

In (listed are 2013-14 stats)

LW Mason Raymond (82 gm, 19-26-45)
RW David Clarkson (60 gm, 5-6-11)
C Dave Bolland (23 gm, 8-4-12)
D Paul Ranger (53 gm, 6-8-14)
D Tim Gleason (39 gm, 1-4-5)
RW Troy Brodie (47 gm, 3-7-10)
G Jonathan Bernier (55 gm, 26-19-7, .923 SV%)


2. Getting Healthier

LW Joffrey Lupul (69 gm, 22-22-44), after 16 gm, 11-7-18


3. Coming off a Career Year

C Jay McClement (81 gm, 4-6-10), after 48 gm, 8-9-17


4. Developing Youth

C Nazem Kadri (22)
D Jake Gardiner (23)
LW James Van Riemsdyk (24)
RW Phil Kessel (25)
G Jonathan Bernier (25)
G James Reimer (25)


5. Rookies Inserted

D Morgan Rielly (73 gm, 2-25-27)
C Peter Holland (39 gm, 5-5-10)


6. State of Goaltending

James Reimer's save percentage dropped from .924 to .911.
Jonathan Bernier was .923 in place of Ben Scrivens' .915.


Summary

The personnel moves made by GM Dave Nonis amounted to nothing more than re-arranging the deck chairs. If the Leafs had gotten better, it was marginal with Bernier the difference. While a healthier Lupul was welcomed back in the line-up, his production wasn't what it has been earlier in his career and his goals-per-game pace was nowhere near the fast start he had enjoyed the year prior. Speaking of coming back to Earth, McClement scored half as many goals and in nearly twice as many games. There are some nice young pieces up front with Kadri, Van Riemsdyk and Kessel but just four impactful skaters aged 25-and-under limited any boost this improved youth could provide. Rielly was a nice add to the back end and Holland was serviceable in a half-season but that was it when it came to rookies. In net, Reimer's play fell off but Bernier was an improvement on Scrivens so all in all, they weren't worse, but weren't better either.


Breaking Down the 2014-15 Colorado Avalanche

All stats are from 2013-14 unless otherwise noted.

1. Personnel Changes

Out

C Paul Stastny (71 gm, 25-35-60)
RW PA Parenteau (55 gm, 14-19-33)
D Andre Benoit (79 gm, 7-21-28)
D Cory Sarich (54 gm, 1-9-10)
G J-S Giguere (22 gm, 11-6-1, .913 SV%)

In (listed are 2014-15 stats)

RW Jarome Iginla (82 gm, 29-30-59)
C Daniel Briere (57 gm, 8-4-12)
D Brad Stuart (65 gm, 3-10-13)
D Zach Redmond (59 gm, 5-15-20)
G Reto Berra (19 gm, 5-4-1, .918 SV%)


2. Getting Healthier

LW Alex Tanguay (80 gm, 22-33-55), after 16 gm, 4-7-11


3. Coming off a Career Year

C Ryan O'Reilly (82 gm, 17-38-55), after 80 gm, 28-36-64


4. Developing Youth

RW Nathan MacKinnon (19)
LW Gabriel Landeskog (21)
C Matt Duchene (23)
C Ryan O'Reilly (23)
D Tyson Barrie (23)


5. Rookies Inserted

RW Dennis Everberg (55 gm, 3-9-12)
G Calvin Pickard (13 gm, 6-7-3, .932)


6. State of Goaltending

Semyon Varlamov's save percentage dropped from .927 to .921.
Pickard's .932 and Berra's .918 were both better than J-S Giguere's .913.


Summary

Tanguay's return to health and GM Greg Sherman reuniting him with Iginla upfront helped the Avs offset the departure of Stastny and Parenteau. However, I'd call the changes on the blue-line a wash and Berra for Giguere wasn't a game-changer either. O'Reilly's goal production after his break-out year fell off sharply going from 28 to 17. Keep in mind O'Reilly never scored more than 19 goals in major junior. Duchene, Landeskog and MacKinnon are some superb young pieces but all three saw their goal and point totals drop from the previous year. While Pickard had a brief run of success, he only played 13 games. As the only rookie skater, Everberg did not have the type of impact that would move the needle. Between the pipes, the improved numbers from Pickard in a back-up role helped off-set the decline in Varlamov's save percentage.


Breaking Down the 2015-16 Calgary Flames

1. Personnel Changes

Out

LW Curtis Glencross (53 gm, 9-19-28)
D Raphael Diaz (56 gm, 2-2-4)

In 

RW Michael Frolik
D Dougie Hamilton


2. Getting Healthier

D Mark Giordano (61 gm, 11-37-48)
C Mikael Backlund (52 gm, 10-17-27)


3. Coming off a Career Year

RW Jiri Hudler (78 gm, 31-45-76)
LW Lance Bouma (78 gm, 16-18-34)
D Dennis Wideman (80 gm, 15-41-56)


4. Developing Youth

C Sean Monahan (20)
LW Johnny Gaudreau (22)
C Markus Granlund (22)
C Drew Shore (24)
D TJ Brodie (25)
LW Lance Bouma (25)
RW Joe Colborne (25)
C Josh Jooris (25)


5. Rookies Inserted

C Sam Bennett
LW Micheal Ferland
G Joni Ortio
D Tyler Wotherspoon (possibly)
RW Emile Poirier (possibly)


6. State of Goaltending

Jonas Hiller had a .918 save percentage while Karri Ramo posted a .912


Summary

While in terms of quantity, the Flames have made far less personnel moves this off-season compared to the Leafs and Avs, the quality has been far superior in adding a legitimate top-six winger in Frolik (ranked 19th among all RWs in SAT% last year*) and a top pairing-calibre defenceman in Hamilton (ranked 7th among D in SAT% last year*). With Glencross and Diaz the most significant pieces not returning, that's a very significant upgrade.

Another boost will be having the services of Giordano and Backlund for a full season. Those are two key pieces and both missed significant chunks of last year with arm and abdominal injuries that sidelined them for a combined 51 games.

Now there's certainly a good chance to see drop off from Hudler, Bouma and Wideman after their big years, but a potential Hudler regression is less likely if he remains linemates with Monahan and Gaudreau, who are young and getting better. Hamilton's arrival should offset any regression from Wideman.

Another point of differentiation for Calgary is around youth. There could be as many as eight returning skaters that will be age 25-and-under that could be part of this team come October (and make it nine if you include Hamilton, who already has three seasons under his belt). That group includes Monahan, Gaudreau and Brodie, who are a huge part of the core of this team. Along the same lines, you have another wave of highly-touted prospects on the way with at least three -- Bennett, Ferland (technically not a rookie with 26 NHL games but close enough), and Ortio -- all expected to be regulars this season and make big contributions.

Regardless of what happens in net and if Hiller or Ramo are traded, I don't see Calgary's goaltending being significantly better or worse from last season. If anything, maybe a little bit better.

* Minimum of 40 games.


Conclusion

Calgary barely made the playoffs last season. If the Kings don't blow that 3-0 lead against the Flames in that dramatic final game before Christmas, Los Angeles would have finished with more points and got third spot in the Pacific Division instead. With many teams in the Western Conference improving this summer, it will only take the slightest regression in 2015-16 for Calgary to end up watching the post-season on television.

In fact, if it wasn't for the work Treliving got done during that six-day stretch from June 26 to July 1 -- trading three draft picks for Hamilton and signing Frolik as a free agent -- I would argue it was going to be touch and go all season to see if Calgary could make it back in, needing healthy seasons from all of its key players.

However, Hamilton and Frolik are game-changers. Those additions at the expense of only draft picks puts Calgary in a much better position to start winning games more on merit than on good fortune.

Remove five third period comebacks but replace them with five additional first period leads (the Flames were one of the worst in the NHL for getting first period leads last season but were nearly unbeatable when they got them), that would cancel each other out. Add in a couple more victories overall thanks to a deeper and much improved line-up and instead of regression, we could easily see progression instead.

There's no guarantee that a second straight playoffs lies ahead but what we can say with certainty is Calgary is in a far better position to fend off regression than the Maple Leafs and Avs were.

There will inevitably still be predictions of dark clouds and rain ahead for Calgary -- especially from those perturbed meteorologists in the East, but you know what they say around here. If you don't like the forecast, just wait five minutes... or in this case, make that five games.



By the way, have you liked Flames From 80 Feet on Facebook yet? Go there and do so now. It's just another way to be alerted to new Calgary Flames articles that I've written.

-----------------------------------------------

Recent Flames Reading
  • Smiling Brandon Hickey: Genuinely Enjoys Playing the Game - Get to know the highly-touted Flames defence prospect, who is making people stand up and pay attention to him after a terrific rookie season at Boston University, which helped earn him an invite to Canada's summer World Junior camp. (July 31, 2015)
  • Fork in the Road: Comparing Kylington's Options of AHL vs. WHL - I take a look at the pros and cons of Oliver Kylington going to Stockton or going to Brandon. Spoiler alert: A good argument can be made for both paths. This is going to be a good debate. (July 30, 2015) 
  • StickTAP: Your Questions Answered - In my first mailbag feature of the summer, I tackled 24 of the over 60 questions submitted by readers by answering in the form of a Thought, Analysis, or Prediction. Lots of interesting conversation topics on the minds of Flames fans these days. (July 27, 2015)
  • Breaking Down the Flames 2015-16 Roster - I've updated this roster breakdown to include the recent RFA signings of Josh Jooris and Paul Byron. How close are the Flames to the cap? Well, come and find out. (July 26, 2015)
  • Eight From 80 Feet: Eight Perspectives on Bouma Contract - It was a deal that was celebrated by many while also being soundly criticized by others. To help evaluate it, I dig into the contract and viewed it from a number of different perspectives. (July 23, 2015)

6 comments:

  1. Nice job outlining the regression discussion. Lots went right for Calgary last year, but goaltending was not one of the them. Compared to Toronto and Colorado goaltending, Flames goaltending was merely average. Also, getting healthy TOI from their existing best +CORSI players (Giordano and Backlund) and adding two equally great +CORSI players (Hamilton and Frolik) are very positive steps. The Flames are still a bubble team by my eye, but they are far closer to the playoffs, compared to this time last year.

    ReplyDelete
    Replies
    1. I can't disagree with what you said although I thought the goaltending was alright. Not great, but OK. Each had their hot streaks and Hartley rode those hot streaks fairly effectively I thought. But the goalie of the future may not be any of the three currently on the roster -- so good thing they're building a wall on the blue-line. To make the playoffs this season, that path will come through top three in the Pacific. Central looks tough so that division likely gobbles up both wild card spots.

      Delete
  2. Excellent article Darren, really agree with alot of it. Like you, I thought the Flames were going to be in tough to make the playoffs again without adding some significant pieces, especially because teams like LA, DAL, EDM, and COL would all likely see some improvement. The additions of Hamilton and Frolik, as well as the insertion of Sam Bennett into the lineup, have me alot more positive regarding the Flames making the playoffs in consecutive years.

    ReplyDelete
    Replies
    1. Making the playoffs will always be tough. This isn't the old days of 16 teams making out of 21. It could be 16 of 32 in a few years. Helpful for Calgary is the Pacific is not the same division as the Central, which looks awfully difficult. Attaining a top three spot in the Pacific looks more likely than one of the two wild card spots. I'm interested to see how the Pacific fares vs Central in head-to-head. How Calgary does against divisional teams was a huge (HUGE) factor last year. That record will be near impossible to repeat but they'll still need to be very good in that area to assure themselves of a post-season spot. Thanks for the comments.

      Delete
  3. Nice article. it's always good to measure the stats with the game play elements. My opinion is that while Corsi is important; it is also a just a correlated proxy of goal differential (which leads to winning). In the long term, I think Corsi will prove to be something that coaches learn to work around, much like the trap and the butterfly. And to that end the Flames addressed this in upgrading personnel.

    Its PDO that is a little more difficult to overcome. There is always mean regression. I would expect their shooting percentage to drop overall this year. ...But based on above, they should be getting more shots, so that's a wash.

    You're right about the division. It's hard to think the central doesn't send 5 teams to the playoffs. But thanks to the work Benning is doing dismantling the Canucks, I can see the Flames finishing 2nd/3rd in the division again.

    It would be nice to see them win a game in the Honda Center this year though...

    ReplyDelete
    Replies
    1. Good points and I've talked before about the angle you mentioned that while shooting percentage may come down, if shots go up, that will mitigate that. The Canucks are a mystery for sure. It sure looks like they don't know what they're doing but surely there's a plan behind the changes? They know their personnel better than I. Very curious to see how things evolve there with their big financial commitment to the Sedins and how that plays out not just this year but beyond. With Bieksa in Anaheim now, maybe that will be the tonic for the Flames finally winning a game there. Anaheim has made a few changes of note this summer and how will that play out? I'm not so sure they'll run away with things.

      Delete