Random Thoughts: On Gradual Improvement

random thoughts

It was a rough start to the season for the Flames, and not just because expectations were heightened after last year’s turn as Cinderella or because Brad Treliving improved the roster considerably in the off-season. The Flames were objectively, abjectly bad at the onset. The defense was porous, the offense punchless, the goaltending sievelike. 

It seemed like the club was going to run in place or even take a big step back. Everything that had worked for them previously was gone, but their weaknesses were still as glaring as ever. More disconcerting, the coaching staff had seemingly failed to adjust in the offseason, instead choosing to stick with staid strategies that were in desperate need of updating. 

Auston Matthews seemed like a real possibility after October. And he still might be. But the truth is the team has improved since those first few dark weeks. In fact, they have taken a real step forward over last year.

– Let’s pull back from the trees and take a peek at the forest. This can be difficult in the grip of a playoff battle, but the org should still be taking the long view when it comes to the roster’s evolution. 

Here’s a graph of the Flames possession rate (20 game rolling average) over the last season and half from War on Ice:


In aggregate, the Flames ability to control play has improved compared to last year. They still haven’t cracked the magical 50% mark, of course, so this is more of a baby step, but it looks to be a real one at least. 

– We can see in the improvement by parsing and comparing possession rate stats between last year and this year as well:

2014-15: CF/60 – 49.9, CA/60 – 62.4

2015-16: CF/60 – 52.7, CA/60 – 57.5

(CF/60 = corsi events for per 60 minites, CA/60 = corsi events against per 60 minutes)

This means the Flames have improved their total shot attempts by nearly three shots per hour and their shot attempts against by five shots per hour. That’s about eight shot differential per hour improvement. 

Which sounds kind of modest, but that means Calgary is on pace for a raw corsi differential of -328 this year (currently at -168 through 42 games). Last year, they finished at -838. That’s a net 500+ shot attempt improvement (!!). A not insignificant jump, especially considering what a dumpster fire this team was in October. 

– Due to those early struggles I was fairly harsh on Bob Hartley during the first quarter. There was an over reliance on obviously deficient tactics such as stretch passes as the basis for transition and shot blocking/crease jamming as a defensive scheme. I was hoping the coaching staff would move more aggressively towards a puck control based system in response to the GM’s manoeuvres in the off-season, but it was “more of the same” initially. 

That’s not superficially irrational given the club’s apparent success the season prior, of course. NHL coaches tend to cling to a “what worked last” mentality, even choosing not to change rosters on a game-to-game basis after wins. 

Of course, in reality we need to work from a first principles framework in noisy and chaotic environments like pro hockey, rather than merely being led around by results. An apt metaphor would be hitting a rare flush on the river in a Texas Hold’em poker hand and then deciding to continue to chase flush river cards all the time afterwards. Though you won that one hand, you’d find “chase the river card” to be a poor long term strategy.

So Hartley started the season trying to hit that flush on the river again. The good news is the club’s terrible goaltending and lacklustre SH% early on eventually worked to dissuade him of that notion.

– That’s a long way of saying the team seems to have made some adjustments over the course of the year. In particular the transition game is vastly improved, with much stronger forward support and a greater reliance on short passes and more variety through the neutral zone, rather than three guys hanging around the opposition’s blueline waiting for a defender to thread the needle. 

– Of course, the improvement of certain players has helped move the needle as well. Mark Giordano and Dougie Hamilton started the year looking like they’d swapped jerseys with Like Schenn and Justin Schultz, which made the nine-game absence of T.J. Brodie more keenly felt. They’ve since righted the ship and Brodie continues to prove he’s one of the best defenders in the league. 

Sam Bennett, Michael Frolik and Micheal Ferland all look like legitimate adds to the roster as well. Cumulatively they have likely helped the Flames tilt the ice a bit more in their favour. There’s still a lot of work left to be done with the roster and some of the decisions being made (let’s not see Joe Colborne above the fourth line again, please), but there’s progress.

– Speaking of roster decisions, I was asked a while ago just what kind of effect a player like Kris Russell has on a team’s goal differential in a year. I was also curious, so I calculated Russell’s theoretical impact based on this model created by Garrett Hohl of Hockey Graphs. A brief description:

Now you can describe the value of Corsi in terms of goal differentials over a season. First look at what a player’s 5v5 TOI/60 was to determine which bucket they belong to. Then plug in a player’s Corsi% as a decimal (ex: 50.0% equals 0.500), multiply by slope, and add the intercept.

I plugged Russell’s values into the equation for both first pairing and second pairing defenders. I used a 43% corsi rate since that’s his rough average since he landed in Calgary. According to Hohl’s model, that represents a goal differential between -11 and -13 goals a year given his ice time. 

That’s a lot. It roughly equates to two losses per year.

– This brings me to the future of Kris Russell on the Flames. By all indications, the org and coach like him a lot. He’s a hard worker, valued by his teammates, wears a letter and enjoys a really good reputation around the league and amongst the MSM. 

Those are all fine things, but the fact is Russell gets completely run over at even strength as a second pairing defender. His reputation far outstrips his true effectiveness. His perceived value outweighs his actual value. Now is the time to sell this asset.

People ask me about the Flames trade deadline. There’s a lot still yet to be decided in the interim and trying to read the market this far out is a bit of a fool’s errand. The one thing I can say with certainty is: the club should absolutely trade Kris Russell to the highest bidder, whatever else happens. 

Russell is going to want a raise for his next deal, so Treliving and company shouldn’t even be considering a new contract (particularly with a blueline already overloaded with costly deals moving forward). A stagnating salary cap and numerous pending unavoidable raises to key players means the club has to be totally ruthless about pursuing value for dollar deals.   

The Russell situation is a real test for this new regime. Do they fall to the lure of sentimentality and “intangibles”? Or do they leverage those perceptions for a quality return and avoid signing the next ghastly Andrew MacDonald-type boat anchor?

  • MRL

    “Mark Giordano and Dougie Hamilton started the year looking like they’d swapped jerseys with Like Schenn and Justin Schultz.”

    Ouch. A greater insult cold not be uttered in the hockey realm.

  • RedMan

    thanks Kent, I had to hold my breath, expecting to see the dreaded “regression”, but it turns out to be good news.

    but on a serious note, and not to start an argument or be a dink, but we heard many times that last years “fancy stats” suggested that there WOULD be regression this year, right? so does the step forward they have taken suggest that these new stats aren’t as predictive as we were told to believe?

    seriously, not trying to start ab argument, but trying to understand the value of these stats.

    excellent article by the way – thanks!

    • Well their SH% and SV% has definitely regressed a bit relative to last year. The caveat to this topic was everyone beating their fists at the team trying to add pieces to improve.

      Of which, they did, notably seeing returns on scoring chances generated/against, plus you can circle back on Kent’s point about rookies, growth, etc. for a potential reason of this.

      See below on the scoring chances for percentage (all results pulled from War on Ice and 5v5)

      2015-16 (42GP) – 49.3% SCF%, -26 differential which is good for 19th currently.

      2014-15 (42GP) – 43.9% SCF%, -211 differential which was good for 28th.

      I might write about this soon when I have time.

    • supra steve

      The way that the Flames were discussed re:regression last year was that the input (possession) did not match the output (points).

      The Flames are much better! But that doesn’t mean they aren’t still bad. The output this season is much closer to what should be expected with that type of input; and is made to look better than it is by the pathetic division they play in.

    • “we heard many times that last years “fancy stats” suggested that there WOULD be regression this year, right?”

      I think this was based under the assumption the team makes no improvements and players don’t take steps forwards.
      While obviously Monahan is somewhat of a liability defensively. He’s improved in the face offs.
      Wideman and Russell have both regressed.
      Players like Bennett and Granlund have taken steps forwards.
      Engellend has actually been a better player in the 6D spot.
      Johnny is progressing as well.
      There’s also the return of Brodie and restablishment of Brodano success, I would argue they are back playing at the level they were last year (no real regression there).
      On top of that, Treliving made significant moves in the offseason by adding Frolik and Hamilton. Both of these players drive posession forwards. So I think when you look at the overall picture. The Flames improved in key areas which they were deficient in last year. Thus mitigating and even marginally improving their fancy stats reputation. I personally still think they’re a few moves from being a legitmate contender as the start to the year indicates.
      One of those in my opinion is Hartley. I don’t see this team being a contender under Hartley’s system and useage. That being said if they can add a pure sniper and talented winger this team has a lot of the tools to be a 50%+ corsi team and have success.

  • One name that never comes up in trade talk is Stajan. I like him as a player, but at over $3m per he is very expensive for a fourth line center. It seems to me he could fetch a second. We seem to have lots of options on the Heat: Grant, Arnold, Shore. All would save about $2 million AAV next year.

    If we traded Huds, Stajan and Russell at the deadline we might have enough spare picks to do a Hamilton or something close in July. I think we could still be in the playoff hunt this year.

    I also think we’d still have enough veteran leadership: Gio, Wides, Frolik etc.


  • MRL

    Depth wise the Flames have put themselves in a very good position of being sellers (Hudler-Russell-Colborne-Hiller-Jones-Granlund) and still make a strong run at the playoffs…..

    • Kevin R

      Agree, wouldn’t it be something to be that big of a seller & still fight for a playoff spot to the bitter end. The only one in that group I think could setback our playoff aspirations is believe it or not, Hudler. He still has skills, last year was an over achievement but not a stretch. I think if the Ferly experiment with Johnny & Money works, Hudler may be a perfect match with Bennett on that 2nd line. If Huds can set Bennett’s motion the same direction he did with Gaudreau & Monahan, that is worth the 2nd & 3rd round picks we probably will get for him at the TDL & could probably land us with another playoff adventure. The others would not set our playoff aspirations back as much.

  • supra steve

    “The one thing I can say with certainty is: the club should absolutely trade Kris Russell to the highest bidder, whatever else happens.”


    So what might they get? A late 2nd rounder seems not enough, and a late first seems too much to expect? Could they move Russell plus Flames 3rd or 4th and get back a late 1st? Am I over/under valuing Russell? I guess in the end, the market will dictate what he’s worth.

  • ChinookArchYYC

    Kent, this is a good overview that is well supported.

    The team is getting better using eyetest too. The team has largely earned their wins and in those wins have been the better team, in terms of carrying the play and out chancing opponents. Defensively, their much better. Giordano and Brodie are back to playing to their 2014/15 level. The team has 3 bonfire top 6 forwards in Gaudreau, Bennet and Monahan. As well, as decent bottom 6 forwards. For me goaltending is a still a question, but full credit to Ramo for is recent play.

    The team has taken another step forward in their development. All in all, good to see.