More on Mikael Backlund



I know, I know I’ve beat this into the ground recently. However, with our own Vintage Flame getting into a twitter fight with Dean Molberg and Andrew Walker about Backlund this morning, I feel obliged to offer one last, brief article on this topic.

The Fan960 morning crew named Backlund one of the Flames biggest disappointments this year. And to be fair his offense just hasn’t been there. He’s also a minus player by traditional plus/minus.I don’t necessarily begrudge their perceptions of his struggles – it’s hard to escape.

We’ve already disussed how Backlund has really good possession and scoring chance numbers this year. As in, tops on the team. It’s also been mentioned that his struggles are therefore almost entirely to do with "the percentages" meaning save percentage and shooting percentage while he’s on the ice.

This is an entirely counter-intuitive, broad brush stat that probably doesn’t mean anything to most. Gabe Desjardins of and Artic Ice Hockey has a couple of graphs that might help solidify things for some:

Graph 1:

He adds – This plot shows how much future PDO regresses to the mean relative to PDO though a certain number of shots (both for and against.) For example, PDO through the first 1000 shots regresses 87% to the mean over the rest of the season – a team with a 1030 PDO through 1000 shots expects a PDO of 1004 for the rest of the season.

Graph 2:

The point is – the percentages tend to regress/converge to the league mean over time. This is true of teams and individuals. This is why we spend so much time talking about possession/volume of chances and shots here rather than frequency of goals (ie; percentages) – because the latter is almost always going to settle somewhere in the middle over time.

This year, Backlund’s possession metrics are excellent. His PDO (on-ice SV%+SH%) is a near team low 95.0. The former indicates he is playing well – the puck moves the right direction when he is on the ice. The latter suggests he’s been unlucky. Once the percentages regress as they do inevitably over the long-term (see graphs above!) the points will come. 

  • xis10ce

    Kent, I have googled and looked at Gabe’s and Puck Prospectus for the same type of graph for either corsi rates or chance rates and have come up with nothing. Have you ever seen the same charts showing persistence over time for possession rates?

    I think it would assist in putting the scale of regression in perspective.

      • ChinookArchYYC

        “Standard regression to mean for Fenwick is 33%, so a 48.4% team would regress to 48.9% or so. But a team that’s at 48.4% is probably going to miss the playoffs, so it will probably sell its best players and not get quite as close to the mean.


        So, for those following at home, (Kent will straighten me out if I get it wrong) while on-ice shooting+on-ice save percentage tend to devolve 87% back to the average, a Fenwick %age will only devolve 33% of the way back to 50% over the season. Assuming Fenwick and Corsi are close enough for this purpose, Backlund is likely to end the season with something like a Corsi percentage a little under 52% (52.9 from the last Black box, regressing 33% or declining .9 or so).

        To play a bit of Devil’s advocate though, his career shooting percentage is 5.3% over 262 shots. While he hasn’t had a lot of powerplay time, that is still getting to be a good sample size and a big drop from league average shooting.

        Perhaps his ceiling is strong defensive centre with limited offensive upside?

        • He had a low finish rate in the AHL as well.

          That said, Andrew Cogliano had a SH% around 17% through his first 214 shots and then it dived to about 8% for his next 268 shots. He’s settled at 12.1% over 500+ shots so far in his career. You probably need 500+ shots on net to really get a good feel for talent level.

          Also, even if Backlund is a low % finisher I’d be stunned if he was actually a 5% shooter. I mean, that’s defeseman territory. 9% is my rule of thumb for inaccurate forwards who aren’t goons.

          • Mitch2

            Ok, Backlund has 262 shots so far. He’ll probably add ~80 on this season giving him 342. By the end of next season we can pass judgement on him because he will be over your 500 total shot count.

            I can accept that time line.

        • SmellOfVictory

          That’s specifically for teams rather than players. Individual players can repeatably have higher/lower-than-average fenwick/corsi/scoring chance differentials and they will have no regression.

          PDO regresses for individuals because it involves primarily factors out of their control (the sv% of two goaltenders, the shooting ability of 9 other skaters at any given moment, bounces, etc). Thus even the better offensive/defensive players will only affect PDO on their own merits by a comparatively small amount.

          Fenwick and corsi are less reliant on external factors to determine their outcome on a game-to-game basis*, which means we can’t really expect substantial single-player regression. We can expect team regression simply because there isn’t that great a divide between the best and worst teams in the NHL; there is, however, a massive divide between Sidney Crosby’s individual ability and Jody Shelley’s individual ability.

          *To clarify, it’s easier to isolate and adjust for the external factors that do affect corsi/fenwick (e.g. zonestart, qualcomp, qualteam). Since we can account for most of the external factors, these measures (in context) lose most of their randomness in comparison to other measures.

          • Bob Cobb

            Sure, it was a team level reference, but so were the PDO charts from Gabe that Kent referenced. I am certain there is a difference in individual PDO and possession rates regression to the mean from team level, but we make do with what we have. If you have the individual level regression to the mean rates for PDO, Corsi or Fenwick I am happy to be shown the differences. I couldn’t find them.

            But of course, they will regress to some degree. The issues is over what sample size we are regresssing from and to.

            As a hypothetical, say a player has a one game corsi rate of 75% or so (6 events for, 2 against). It is a virtual certainty that it will regress over the next 81 or 160 or 500 games closer to 50%.

            If, however, a player has a 55% corsi rate over 200 games, I would agree with you that we probably won’t see that corsi rate move substantially back to 50% over another 200 games.

            And it is not completely true that players don’t drive on-ice shooting percentage. See Tom Awad’s analysis starting here ( Finishing ability drives as much of the results as possession – it is just easier to see who will repeatably drive possession than who will repeatably finish. If Mikael Backlund turns out to be Ryan Clowe or Scott Gomez (good at posession, bad at finishing) I can live with that result. I will still hope he doesn’t.

          • RexLibris

            Agreed on the contract and that promotion story is pretty awesome.


            He was 6.9% last season so that is where I put the under 7%. In order to get to 7% on this season, and assuming his same shot rate, he would need about 10 more goals. Wouldn’t that be nice with Glencross out.

  • xis10ce

    For Gods sake why dump on Backlund. Where did we pick him?? If the expectation is perfection from our best youngest player we have, then perhaps we should start when we get a lottery pick. Right now, lets cut the kid some slack, it’s not if he sucks, its how he sucks. I think the injury took away the jump to get out the gate quick this year & there is nothing we can do about it. So lets trade him & watch him breakout with some other team. That’ll make everyone feel great. So those trashing him now, will they scream fire Feaster if he trades him & he scores 70 points & plays great with a new team next year. & how old is he?? I dont think it’s Backlunds fault we are in 12th place right now.

  • SmellOfVictory

    Giordano, Hannan, Bourque. There are the three main disappointments thus far. Iginla qualified as the biggest through the first chunk of the season, but he’s absolving himself of that now.

  • jeremywilhelm

    Of topic, but Josh Rimer, same guy who broke the Crosby done for season last year, is saying Taylor Hall is done for the year. This seems incredibly odd.

  • Derzie

    If you are a forward and you don’t score, you are incomplete. You may be a great back checker which makes you a 4th pairing defenseman, not a forward. Backlund gets points for effort and responsibility but not scoring. 50% of his expectations is not success. I love Tim Jackman and his effort but he does not show up on the scoresheet. Effort is not enough in the bigs.

    • the forgotten man

      Ditto…but the Flames cannot give up on Backlund yet – I still hope he is a critical component to this franchise post-Iginla.

      My worry though is that you can’t teach “how to snipe” – that is God-given…I am hoping Backlund is just big time snakebitten.

    • The point of this article is that his current scoring rate is atypically low due to variance/poor bounces.

      In his first season with the Flames, Backlund’s scoring rate was 2.33 even strength points per 60 minutes of ice. Second best on the team.

      Last year it dipped to 1.58, which isn’t great but still functional.

      This year, it’s just 1.00, despite the fact he is objectively a better player on the ice than he was the previous two seasons.

      I agree forwards need to score to some degree in order to be useful. It’s good bet that Backs will, however, once the percentages regress to the mean.

  • RexLibris

    Great article, Kent. My favourite table was the one with the four pretty colours. The blue one was nice, too.

    But picking on a 22-year old centre who is struggling offensively despite putting in a solid effort and obvious skills? That’s so “Edmonton”!

    If the Flames do trade him, my hope is that it is to an Eastern conference (or whatever new conference comes around) team. I’m not saying he’s going to become Patrick Sharp all of a sudden, but I think he could be, with the right development and wingers, in the range of expectations as Kyle Turris. Minus the attitude. He creates, he protects, and he can more or less break even in the FA circle. If he is a major disappointment to these guys, they need to learn a little patience.

    • Flames fans are so used to their forward prospects failing to do anything in the NHL that with Backlund’s struggles this year a lot of us are willing to write him off as just another wasted first rounder who couldn’t hack it in the bigs.

      • RexLibris

        Oh, the failure of Flames first rounders (Flame-outs, if you will) is well documented so I totally understand the cynicism. It’s just that usually, with modern media and analysis, most fans are willing to give a prospect longer than it seems Backlund has been given. Especially considering he has been able to operate largely in the shadow of other players like Iginla, Jokinen, and such.

        Is there any chance that the criticism Backlund faces today would have been avoided or at least reduced if Tanguay had been healthier? Even if Tanguay struggles then it could give fans and media like those mentioned above someone else to focus on. In fact, was Tanguay even mentioned as a disappointment for Flames fans given the amount of time missed this season due to injury?

  • SmellOfVictory

    Backlund’s production has been disappointing, but not his overall play. Like Kent said, his possession numbers are solid, but his PDO is 95.0, when the league average is around 100. That’s terrible luck!! If Backlund simply ‘stays the course’ the points should start to come.

  • xis10ce

    you can only miss so many empty nets before you start hitting some, he’ll come around. I just hope that he does before mgmt tries to give him away for almost nothing in return.

    • ChinookArchYYC

      This is exactly what I’m afraid of.

      The boys from the Fan were at it again this morning, but in fairness to them the NHL is an end-results. You can hardly blame them for criticizing a top center on the team with a paltry 3 gosls. The problem I have with the prevailing attitude is that their are many instances where the wrong players are rewarded for goals, while being a complete liability in every other aspect of the game – enter Kaberle, Babchuck, Ehrhoff and many others.

  • RexLibris

    I see the same player that your stats show. He’s evolving into a VG two way player. He’s just missing is that killer net presence he had as junior.

    As an aside…. It doesn’t strike me that it would take much to get into an argument with Dean ‘Apathy’ Molberg.

  • Bob Cobb

    Congrats!! The Flames have found their version of Shawn Horcoff in Backlund. A guy with some offensive potential who is unlucky around the net, reliable in his own end and destined to be a second or third line centre on a good team but has to play top line minutes on a team that has no real top line centre. Just another indication that the Flames have to rebuild and not try to do it on the fly by trading draft picks and prospects for over the hill vets.

  • Bob Cobb

    This list of forwards with back to back seasons of sub-7% shooting seasons in their 20-23 year old seasons and 40 or more GP would also appear to be relevant.

    Upside names?

    Shane Doan (Career 10.1%)
    Olli Jokinen (Career 9.7%)
    Brooks Laich (Career 9.9%)

    Not so much?

    Mathieu Dandenault (Career 6.9%)
    PJ Axelsson (Career 7.6%)
    Greg Campbell (Career 7.3%)
    Patrick O’Sullivan (Career 6.9%)

  • Bob Cobb

    The silver lining in Backlund not having that breakout year we all were hoping for is his contract expires this year. So maybe Feaster could get him locked up to a 3-4 year deal at Glencross type money & hope he breaks out next year. If he had nailed it, he would be looking at a 4-5 year deal at Stajan type of money or Tanguay. This could actually be a good thing.:)