May 13, 2013; Boston, MA USA; The Toronto Maple Leafs and the Boston Bruins shake hands after the Bruins defeated the Leafs in overtime in game seven of the first round of the 2013 Stanley Cup Playoffs at TD Garden. Mandatory Credit: Bob DeChiara-USA TODAY Sports

Regression; Meet The Toronto Maple Leafs

Regression, according to the dictionary, is defined as:

  • Relapse to a less perfect or developed state
  • Statistics The relationship between the mean value of a random variable and the corresponding values of one or more independent variables

What this has to do with the Toronto Maple Buds is many things, but consider the following. Here are the statistics of goalies who played at least 15 games in a season not named James Reimer since the 2007-08 season:


Vesa Toskala (.904 SV%, 2.74 GAA, 66 GP)

Andrew Raycroft (.876 SV%, 3.92 GAA, 19 GP)


Vesa Toskala (.891 SV%, 3.26 GAA, 53 GP)

Curtis Joseph (.905 SV%, 3.57 GAA, 21GP)


Jonas Gustavsson (.902 SV%, 2.87 GAA, 42 GP)

Vesa Toskala (.874 SV%, 3.66 GAA, 26 GP)

Jean-Sebastien Giguere (.916 SV%, 2.49 GAA, 15 GP)


JS Giguere (.900 SV%, 2.87 GAA, 33 GP)

Jonas Gustavsson (.890 SV%, 3.29 GAA, 23 GP)


Jonas Gustavsson (.902 SV%, 2.92 GAA, 42 GP)


Ben Scrivens (.915 SV%, 2.69 GAA, 20 GP)

I’m not a mathematician, nor do I pretend to play one on the internet, but I can tell you this: The Leafs had crap goaltending from the end of the 2003-04 season until the arrival of James Reimer (If you need a refresher as to what ended that year, watch this responsibly: The Leafs have predominantly been in the bottom five in save percentage over the years. It’s important to keep this in mind while looking at the Leafs’ PDO. PDO is not a fancy acronym, but it’s generally referred to as puck luck. From the hockey analytics website Behind the Net:

PDO is the sum of “On-Ice Shooting Percentage” and “On-Ice Save Percentage” while a player was on the ice. It regresses very heavily to the mean in the long-run: a team or player well above 1000 has generally played in good luck and should expect to drop going forward and vice-versa.

As everyone in their right mind knows, last year was the first time since 2004 where the Leafs made the playoffs. The thing mainstream media guys (*cough* Sportsnet, *cough* Toronto newspapers) latched on about this iteration was how everyone on the Leafs “bought in” to Randy Carlyle‘s system. The Leafs, according to some, played a more “winning” brand of hockey that involved toughness and better communication from coach to players. To some people, those are things are actually real and legitimate. To others, including the guy who’s writing this piece, it’s bullshit.

The Leafs were tied with Pittsburgh at the top of the league in PDO. The Leafs also lead the league in team shooting percentage. The Leafs made the playoffs in major part because they finally got above-average goaltending, the likes they haven’t seen since Ed Belfour was sober in 2003-04. The problem with relying on shooting percentages and goaltending, or in other words “puck luck”, is that the volatility of PDO is real, and unless you’re of the calibre of Pittsburgh or Boston, PDO varies heavily from year to year. The following is the top 10 teams in PDO since the 2007-08 season.

  1. Vancouver Canucks
  2. Boston Bruins
  3. Pittsburgh Penguins
  4. Washington Capitals
  5. Anaheim Ducks
  6. Nashville Predators
  7. Montreal Canadiens
  8. Buffalo Sabres
  9. Phoenix Coyotes
  10. Philadelphia Flyers

Six out of the 10 teams here (5-10) missed the playoffs at least once during that six-year stretch. The Penguins might be the most consistent team out of the bunch yet they lost in the first round against the Flyers. The Capitals and Canucks have been counted out by most in the media more times than a professional wrestler. For reference, here are the top 10 teams in PDO from 2011-12:

  1. Boston Bruins
  2. Detroit Red Wings
  3. Vancouver Canucks
  4. Phoenix Coyotes
  5. Nashville Predators
  6. New York Rangers
  7. St. Louis Blues
  8. Buffalo Sabres
  9. Washington Capitals
  10. Philadelphia Flyers

And 2012-13:

  1. Toronto Maple Leafs
  2. Pittsburgh Penguins
  3. Chicago Blackhawks
  4. Columbus Blue Jackets
  5. Anaheim Ducks
  6. Washington Capitals
  7. Dallas Stars
  8. Tampa Bay Lightning
  9. Montreal Canadiens
  10. New York Rangers

The only team that made it twice on these lists were Washington and the Rangers. Now PDO shouldn’t be solely referred as a death sentence for a team, because as the old adage goes, some teams “create their own luck”. From the 2011-12 list, six out of the 10 teams were in the top half in Corsi percentage and five out of the 10 teams were also in the top half in Fenwick percentage. Concerning the 2012-13 list, the “creating your own luck” adage is tested more with only three of the 10 teams being in the top half in Corsi, and four of the 10 teams ranking in the top half in Fenwick.

What should concern Leafs fans with the current iteration of the team is the lack of possession players they have on the squad. They lost Matt Frattin, Mikhail Grabovski, and Clarke MacArthur; all guys who had varying degrees of success as possession players throughout their time in Toronto. The buyout of Grabovski exposed the split between bloggers/shot metrics supporters and the old guard in the hockey media. The old guard mocked at the value of Grabbo as a result of his declining point production, his heavy salary, and who Randy Carlyle usually played him with on the third line. The “Corsi Nerds” screamed at the old guard saying he’s a terrific two-way player who’s a great possession asset to have, but hasn’t had the opportunities that say a Tyler Bozak has had. Many observers point to his playoff performance as an indicator of what he can do when given an opportunity to succeed. Hell, he’s even produced in the traditional metric that the old guard favours, ranking 17th among centremen in goal-scoring over the last three seasons.

Then there’s the matter of the Leafs having a ridiculous high shooting percentage, the likes of which the NHL hasn’t seen in a long time. According to, no team has ever lead the league in shooting percentage for two years in a row. One comparable would be the Capitals who lead the league in 2010 with a 10.39 per cent mark, a similar percentage to what Toronto produced during the 2013 season. The following season, under a similar “stylistic” change that the Leafs went through, the Caps ranked 21st in shooting percentage and 16th in PDO.

There’s every reason to believe the Leafs will regress heavily this year, an event that would bring satisfaction to any fan/writer who’s been mocked for pointing out the team’s PDO, Corsi/Fenwick and shooting percentage to the old guard. Despite what’s been written here, as a sad/depressed Leafs fan, I hope they continue to buck every stat imaginable and make the playoffs, an event that’s made even harder with the NHL’s new conference alignment bringing the improved Red Wings to the Eastern Conference. Leafs fans will enjoy the rough and tumble play that the Leafs will embody, highlighted by their free agent acquisitions of Dave Bolland and David Clarkson. Sadly, that might be the only “enjoyable” quality the Leafs will embody in the upcoming season.

Tags: NHL Regression Toronto Maple Leafs

20 Comments on Regression; Meet The Toronto Maple Leafs

  1. Stan Smith says:

    While Reimer was one of the main reasons the leafs made the playoffs and scared the crap out of Boston but I think you have it backwards. It was Wilson’s lack of a good defensive system, ignore the man in front of the net, everyone get in the shooting lanes, get your sticks in front of the shot, forget about taking the body, etc that created hell for the goalie. They never stood a chance because they could never see the puck. Take a look at videos from the Wilson era. As soon as the point man had the puck you could see all six leafs going into a goaltending stance.

    Carlyle’s defensive system of taking the man out of the play, clearing the shooting lanes of bodies, the closest player to the shooter being the only person to attempt to block the shot, and moving forward as he was doing so, instead of laying back, all meant the goalie was now able to see the shot to stop it. I contend that it was Carlyle’s defensive system that helped Reimer be a better goalie.

    • Tim Bayer says:

      I agree Ron Wilson employed terrible defensive and penalty kill systems, but I wouldn’t call Carlyle’s system great either. No team allowed more shots than the Leafs last year, yet Reimer and Scrivens still played phenomenally. Their strong play may have had something to do with Carlyle’s “system”, but the numbers don’t back up that line of thinking at all. Reimer played great under Ron Wilson in 2010-11 as well. The bottom line is teams that don’t control the puck as much as their opposition allow more shots on net, and team’s that allow more shots on net tend to not make the playoffs. Some teams get unsustainably high shooting and save percentages and make the playoffs. That was the case with the Leafs last season.

      • Stan Smith says:

        Reimer did play great for the last part of the 10-11 season when he was first called up but from watching the leafs under Wilson the previous two seasons I felt that was unsustainable. With Wilson’s system the goalie had to cover as much net as possible and hope the puck hit him. From my limited experience playing hockey I know goalies hate being screened so eventually it was going to get to him, the same as it did every other goalie that died under Wilson. As far as shots goes I think it is a deceiving stat. for example under Wilson with his “get in the shooting lanes” philosophy the opposition would have less shots actually get to the net but a higher percentage go in. Under Carlyle the opposition would get more shots on net but the majority of them the goalie would see and have a better chance of stopping. I just saw a stat where the leafs gave up just over 8 scoring chances a game playing 5 on 5 while getting around 7.5 themselves, which is pretty even. One interesting note was the team that gave up the most scoring chances 5 on 5 was the Chicago Blackhawks.

        • Tim Bayer says:

          Eric T. wrote a post last summer looking at the correlation between shots and scoring chances, and found there is only a small difference and that is most likely due to random variations due to the smaller sample in scoring chance numbers. (

          Cam Charron tracked scoring chances for the Leafs all year, and although he hasn’t posted the full season’s numbers yet, at the 36-game mark the Leafs were down 374-356 at even strength. He wrote that in the final 12 games of the season, the Leafs were absolutely killed in scoring chances, so they ended up pretty far below 50% in the end. (

          Because of the smaller sample and subjective nature of tracking scoring chances, it’s generally a better idea to stick with shot metrics like Corsi and Fenwick. Fenwick Close is widely considered the best predictive team measure we have right now, and the Leafs finished second-last in the league with a 44% mark in the regular season, while the eventual Stanley Cup champion Blackhawks finished second at 55.8%. (

          Lastly, I encourage you to read Pension Plan Puppets’ Randy Carlyle Coaching Myths series. (

          The narrative that Randy Carlyle’s “system” has made the team any better defensively, or is the reason for James Reimer’s success, is quite frankly, total BS.

          • James Gleason says:

            also, Carlyle was a guy that played Colton Orr 10 minutes a night almost. So remember who we are dealing with here

          • Stan Smith says:

            And having Colton Orr dressed and on the bench made a huge difference with the Leafs team this past season. If you don’t believe me ask his teammates.

          • Tim Bayer says:

            So Colton Orr made a huge difference on the Leafs because “ask his teammates”. Right. Good talk.

          • Stan Smith says:

            Wow! I thought you would have at least understood a simple statement but I was wrong. Periods end sentences. There were two sentences. One said Orr made a difference last season. (Please notice the period here. that means that sentence is over) Then I said If you don’t believe me, ask his teammates. (please note the use of another period here)

          • Tim Bayer says:

            Actually, my point is you haven’t backed up your statement that Orr made a huge difference on the Leafs with any proof whatsoever. I’m waiting.

          • Stan Smith says:

            There is nothing that I can say that will convince you that having Orr on the team made a difference. I can tell you that in interviews that Nonis, Carlyle, Dave Poulin, Kadri, Lupul, Kessel and the departed Grabovski have all said having Orr on the bench makes them all a bit tougher and not afraid to be more aggressive because they have him to back them up but, darn, there just isn’t any numbers to back that up and for stats junkies such as yourself, it doesn’t matter what you see on the ice or when the game is being played if there isn’t a number that proves it then it can’t be so. The above names that I have mentioned are the ones I can definitely remember saying they like having Orr on the team. I am sure there are more, thus the “ask the players” comment.

          • Tim Bayer says:

            That’s right, the stats say that Colton Orr is a terrible hockey player. So do my eyes. I don’t give a single crap about what his teammates or coaches or second cousins say about him in the media. He’s an awful hockey player, period.

          • Stan Smith says:

            I never said he was a good hockey player. I said he brought value to the team that can’t be measured by numbers.

          • James Gleason says:

            In what any sense did colton Orr have any value in any game?

          • Stan Smith says:

            This might seem like a copout but I am all right with that. If you do not understand that part of the game of hockey then I can’t possibly explain it to you.

          • Stan Smith says:

            First of all I am not a stats junky. Stats are numbers that are compiled after they have happened. People who create and compile stats have a need to make those stats mean something. Even the way stats are compiled means that two people can come up with totally different numbers for the same stat. You can talk about the Carlyle myth all you want but the only real stat that matters is the .583 winning percentage lifetime in the the regular season and a .581% winning percentage and 1 Stanley Cup in the playoffs. Combine that with the fact that he just happen to be coaching the team that either because of him or despite of him, depending on your point of view, made the playoffs for the first time in a decade makes him an excellent NHL coach. Goals for and against, wins, losses and where you are in the standings are the only stats that count.

          • Tim Bayer says:

            So basically, any stats that don’t support your pro-Carlyle narrative must be ignored and dismissed as hocus-pocus, even if they have been proven to have huge predictive value in hockey. Carlyle’s winning percentage and Stanley Cup only tell me that he stood behind the bench of some very good Anaheim Ducks teams which had excellent goaltending, two Norris Trophy defensemen and, at the time, some of the best young forwards in the game. Goals for, goals against, wins and losses only tell me what happened, not how well the team played, how fortunate they were or whether or not the results are likely to be repeated. I’ve shown you through multiple examples that the Leafs were a terrible possession team last year and benefitted from sky-high shooting and save percentages to make the playoffs. And yes, all this happened in spite of Randy Carlyle and his terrible lineup decisions.

          • Stan Smith says:

            Do you realize you just said whether or not a team scores more goals than the other team and wins games instead of losing them does not tell you how well a team has played? That shows how stats junkies can’t see the forest for the trees. Possession stats mean absolutely nothing. It is what you do with the puck when you have it that counts or what you prevent the other team from doing when they have the puck that counts. If a team has possession of the puck in the opposing team’s zone for a good 2 minutes and has 3 scoring chances but doesn’t score followed by the other team scoring on a 10 second rush, which team is playing better? To also state that Carlyle has been lucky enough to be behind the bench for over 300 wins and 36 playoff wins despite being a terrrible coach, and has not been responsible for those wins is ludicrous and ridiculous. I thought by commenting on your post that I could start a meaningful discussion but you think that leafs sucked but were lucky last year and I thought they played head and shoulders better than they have in over a decade. I guess we just agree to disagree and leave it at that.

          • Tim Bayer says:

            Those stats only tell me results, but nothing about the actual process of winning hockey games. As I said before, I use shot differential statistics or “possession stats” as you call them because they’re the single best predictor of future success we have. Of course there’s more to the game than just possession, which is why I also look at goals, shooting percentages and other statistics when evaluating a player. This isn’t an either/or proposition; it’s about using every tool you have available in making evaluations. I can only comment on Carlyle’s time here as head coach, and from what I’ve seen of his lineup decisions and falling out with Mikhail Grabovski, I don’t personally think he’s a very good coach. You obviously don’t feel the same way, and that’s fine.

          • Stan Smith says:

            It’s too bad we are not privy to the goings on in the dressing room. It would be good to know details about the Grabovski situation. As mention by someone in the blogosphere, I do not recall who, it was interesting that Grabovski and Phaneuf both got married the same weekend and that, from what I understand, most of the leaf players attended Phaneuf’s but not Grabovski’s.

            As far as Carlyle is concerned I liked him as a player (which has nothing to do with him as a coach) but with him being in the West, never really followed him as a coach. I did watch as many games as I could during the 06-07 playoffs and I thought the Sens looked really good going into the final but the Ducks made them look really bad. They used their size and bodies to outmuscle Ottawa in every category. I realize they had great players but to me the type of game they played came from Carlyle.

            Moving to present day the Bruins have been the Leafs biggest nemesis since the Kessel trade and Boston plays a lot like the Ducks did then. Carlyle’s leafs almost beat them at their own game. It was only when the leafs fell into their, what I call, Wilson coached ways and stopped taking the body and clearing the front of the net that they got into trouble. I have watched the last few minutes of game 7 a number of times and you can see them stop hitting, everyone focused on the puck along the boards and assuming the “goalie” stance a la Wilson whenever the puck was at the point. There was no one close to Chara screening Reimer on the tying goal.

            I think that when the leafs can manhandle the Bruins and beat them at their own game they will be ready to take that next step. I also think Carlyle is the right coach to get them there.

            One last note just to stir the pot, I think Gardiner played as well as he did in the playoffs because of Carlyle, not despite of him. He needed to learn how to play defence and become a more complete player. I also think that under Carlyle he could become a Norris candidate if not a winner.

            I also look forward to reading your blog and the other writers for Editor in Leaf in the future even if I don’t agree with them. LOL (Now,on the Nonis extension article)

          • Tim Bayer says:

            And just for the record, I didn’t write this post.

Join The Discussion