LGD: The Great Regression


Tonight the Monsters of Regression face-off once again. The 2013-14 regressors, the Toronto Maple Leafs and the 2014-15 regressors, the Colorado Avalanche. The two teams are eerily similar to the extent that both jettisoned talented centers (Grabovski and Stastny) after their playoff season, and both for some reason thought aging, injured defenseman would be the answer for them this season (Robidas and Stuart). 

Of course the biggest similarity was hot goaltending in front of poor defense, and neither of these teams have done much to remedy that defense issue. That should make tonights game fun to watch and if the two teams combine for less than 70 shots, I’ll be shocked.

The Leafs

Yeah, that game in Arizona was pretty bad, but things were going good before that, right? The Avs aren’t exactly the Blackhawks, but they are similar to a broken down version of the Blue Jackets and the Sabres, so maybe the Leafs can play them that way. At the very least if the Leafs can avoid reverting to playing Bozak with Kessel I’d be happy. Kadri has been solid in that role. It would also be nice if Carlyle can find a way to play both Holland and Santorelli over 15 minutes a game.

Projected Lines via Daily Faceoff

Winnik – Kadri – Kessel

van Riemsdyk – Bozak – Clarkson

Komarov – Holland – Santorelli

Panik – Carrick – Ashton

Phaneuf – Franson

Rielly – Polak

Gardiner – Robidas

The Avs

What more can really be said about the Avs that Adrian Dater didn’t already cover in his book? The team has been completely fixed and they are as good as they have ever been. Joe Sakic and Patrick Roy only make sound decisions and bringing in Daniel Briere, Jarome Iginla, and Brad Stuart were unquestionably smart moves. That’s not to say there aren’t nice players on the Avs, and MacKinnon, O’Reilly, Landeskog, Duchene, and Barrie will give this team a lot to cheer about, especially once they add another good young player with their high draft pick this summer. For now, let’s enjoy one of the worst defenses in the league and assume the Leafs will leave Denver with two points tonight.

Projected Lines via Daily Faceoff

Landeskog – Duchene – MacKinnon

McGinn – O’Reilly – Iginla

Tanguay – Mitchell – Everberg

McLeod – Cliche- Talbot

Hejda – Johnson

Barrie –guinan2

Redmond- Holden

Starting Goaltenders




What to Watch For

  • Two teams who thought they were ready to take the next step but aren’t
  • Adrian Dater probably having another twitter meltdown if Leafs fans are doing their job
  • That MacKinnon kid is really good.
  • Hey look it’s John effin’ Mitchell! Guess he’s going to get a goal tonight
  • If the Leafs don’t get a goal every time that Avs 3rd defensive pairing is on the ice, then the Leafs forwards aren’t doing their job.
  • Fans of both teams try to figure out where they are on this chart 


Caption this photo (via @NHLHistoryGirl). Best caption in the comments section wins ten pieces of internet respect.


  • Bertly83

    The data this season suggest that the leaf Corsi has improved and the PDO has fallen. My take is that the leafs analytics has influenced is being felt and is slowly and surely transforming the leafs from a “turtle” high quality, low total scoring chance off the rush strategy into the “jason blake” high shot volume strategy. They are not there yet but that appears to be the trend.

    The funny thing is that with either approach (high or low PDO) that the leafs POINT total is the same year over year. That is, the “output” so far is the same from either approach -dinosaur hockey vs analytic hockey. We already know that Carlyle is not a great coach but in addition, I think this might suggest this team might not be that good either.

    • BubbaLou

      Only if they have both fully changed, and are satisfied with the changes. I would say that besides some new focus from management the overall play on the ice has at best been tweaked since last year. I would bet that, this being the first year of the new management, the team takes a wait and see approach this year and moves are made towards the offseason. That leaves it up to the coaches improve the performance of the team – we’ve seen only minor changes.

      This is a common misconception – “if Corsi is based on shots, then you want to take more of them to improve, so Corsi advocates are just pushing for the high shot volume”. The premise is flawed because it ignores that Corsi is a zero sum game – of the 100% of that period, what % were you on the attack -and the other team was not- is the question, and the general hockey precepts that drive on-ice strategy – you cannot blast low% shots every chance you get the puck and expect to succeed, you have to make them good. In fact, because it’s the NHL, the shots are as genuine as it gets.

      What I’m trying to say is that to go to low% shot hockey is not the goal, but the anaconda-like squeeze of a team in their own zone by outplaying(forecheck/cycle/passes/shots) their defense, and your defense foiling their offence before they can do damage. Corsi% driven by high-event volume from both teams looks like run and gun hockey – what I’m saying you really want is the offense of the other team reduced to a bare minimum.

      Finally, as @iknowfancynumbers alluded to, when viewed at the single game level good Corsi% just reflects which team had more production that night. The only stat that predicts who wins in a single hockey game accurately is which team had more goals – that’s why its exciting to watch. Corsi% over a season tends to predict your standing – the Leafs have improved, but from historically bad to just subpar

  • CMpuck

    Love these games, there isn’t a more smug fanbase than Colorado, love reading what an ‘elite winger’ ROR, how EJ is a true #1D ect, all there players love Colorado and will definitely resign at cost… while dismissing Kessel for being ‘one dimensional’, well I love watching these games and seeing that one dimension = Leafs win.

    • BubbaLou

      Yeah when that one dimension is offense, it’s a pretty damn game-altering dimension to have.

      And who is Colorado, with their defensive implosions, to paint their stars any differently?

  • Benjamin

    For effing once could I just read an article without anti bozak propoganda. I mean no Kadri has actually been pretty ass with kessel and bozak can actually produce very well with him. “but bozak and kessel wowy…” WOWY is useless if a player is shown to actually produce.

  • Benjamin

    For effing once could I just read an article without anti bozak propoganda. I mean no Kadri has actually been pretty ass with kessel and bozak can actually produce very well with him. “but bozak and kessel wowy…” WOWY is useless if a player is shown to actually produce.

  • Benjamin

    I think every one values corsi way too much. Corsi so in my opinion show how hard and efficient a specific player and his line mate works to reach success(ie. Goals, forecheck, momentum swing, evening shots). Poor corsi has more to do with the line mates than the player. This is usually because the line was outmatch, unlucky, mismatched, low work ethics, wasn’t given fair or favorable minutes, plays a style that require or forced shot blocking, plays a style that doesnt require much shot attempts (skillfully line) or doesn’t need to generate a lot of corsi to be successfull due to their high finishing rate (bozak, perry, nash, ryan, burns, vatanen, getzlaf)

  • Benjamin

    So maybe the stats geeks can keep some stats on Bernier such as his preponderance of giving up a soft goal, especially late in a game or in o.t. with ten seconds left as he did against Detroit. Or how about zero saves in the shootout.

    How about a stat on bone headitis such as putting your team shorthanded in o.t. by jumping over the boards way, way, way too soon. Obviously some of these guys don’t belong to Mensa. They did manage to get one point against two bum of the month clubs. But when you think about it the leafs were playing two teams of the same ability of the leafs. So slightly above 500 is where the leafs are at which projected over the year would give them their usual 80 to 85 points good for around a 10th or 11th draft pick.