Random Thoughts: Warnings and Good Arrows

random thoughts

The Flames near the 20-game mark in a position nobody could have anticipated heading into the season –  firmly ensconced in the playoff picture near the top of the Western Conference. The Flames have scored a conference high 59 goals in 19 games and are ahead of heavy hitters like Los Angeles, San Jose and Chicago in the standings. 

Will it last? And, if not, should the Flames results so far be completely dismissed as good fortune? No to both, I’d say.

– Those familiar with PDO are no doubt already aware that the Flames are riding a tidal wave of good luck to start the season. They have the third highest combined save percentage and shooting percentage in the league behind Pittsburgh and Nashville (102.8). 

It should go without saying that this won’t continue. 

It’s not a comment on the Flames hard work nor a wet blanket designed to cynically dismiss their abilities. It’s just a statement of fact: teams don’t maintain those sorts of percentages long-term in this league. Especially not teams who don’t boast a Sidney Crosby up front or Henrik Lundqvist in net. Sometimes everything just clicks for a club. Right now is Calgary’s time.

– The only reason to make note of their PDO driven success is to ensure you don’t fall into false narratives about the club’s results. It’s important to stay grounded about the Calgary’s actual talent base and to keep things straight once the percentages regress. 

An object lesson is the Colorado Avalanche. Last season, they rode incredibly good percentages to an unexpectedly high finish, leading their decision makers to conclude their club had turned the corner. As a result, the Avs made all the wrong decisions during the off-season: a lousy possession team, they let two of their better possession players go in PA Parenteau and Paul Stastny and replaced them with lesser two-way players in Jarome Iginla and Danny Briere. They also didn’t do much with their underwhelming, lead-footed blueline. 

Despite their bevy of young talent, Colorado has put up even worse possession rates this season than they did last year. Their goaltending and finishing has come back down to earth and more or less napalmed their playoff hopes in the first quarter of the season. It’s nice to walk the garden path, just make sure you don’t get lost. That thing they say about players becoming overconfident by “reading their press clippings”? That can happen with management too. 

– I’d say the only narrative Flames fans and management should be on guard for (beyond the regular stuff like “the team got complacent” or “the coach has lost the room”) is if the results regress in concert with the return of one or all of Mikael Backlund, Matt Stajan, Joe Colborne and Mason Raymond. If that happens, the two things will likely be unrelated, but it will be tempting to conclude that “the Flames are better without player(s) X,Y, Z”

– All that said, there are lots of reasons to be excited about the Flames start, even if you understand their place in the standings probably aren’t indicative of their true talent. The organization has seen a lot of progression from key assets, with a few pleasant surprises along the way. 

– Let’s start with Sean Monahan, who seems to have taken a very real step forward. Monahan is second amongst regular forwards in terms of possession and he’s playing much tougher circumstances than last year. He’s also continuing to finish at a double digit pace, which, while not definitive, is a good sign that he might be an above average scorer. 

– The best Flames player in terms of driving play this year? The unlikely Paul Byron, who has been a revelation thus far. Re-signed to a one-year, “prove it” deal this summer, Paul Byron doesn’t just look like a legit NHLer so far this season, he looks like a very, very good one. He’s playing in different circumstances in all game states and driving play. For my money, he’s been Calgary’s most consistent forward up front this year.

This isn’t to say Byron is going to become a scoring leader or Selke candidate or some point – just that it looks like he’s going to be a quality depth option for Calgary as they rebuild. As the Edmonton Oilers have ably demonstrated, it takes more than just collecting high draft picks to dig yourself out of the conference basement.  

– Speaking of quality depth guys, Josh Jooris and Lance Bouma seem to be establishing themselves as high utility bottom rotation forwards. Both guys can skate, play the middle and wing and kill penalties. Another step towards establishing functional toughness.

– Of course, there’s also TJ Brodie and Mark Giordano. While their performance last year was obviously stellar, there was a question whether one or both would be able to replicate it again this season. Obviously there’s no question any longer – they have indisputably been one of the best pairings in the league through the early going. The points are nice, but their true value is their ability to control play in the very toughest of circumstances. One doesn’t need advanced stats to see the “Brodano” effect on the ice – things visibly shift for the better whenever they’re on together. 

If the Flames turn this ship around in the next year or two, it will be because of their elite first pairing. 

Johnny Gaudreau has already proven that he can play at this level, which was the last, lingering issue surrounding the mighty mite. There’s still some distance for Gaudreau to travel before he’s anywhere near his peak, but he’s already generating 10-bell chances at a relatively high clip. Right now he’s mostly doing that for others through crafty passes from the perimeter. You’ll know Johnny has taken the next step when he starts putting more pucks on net from dangerous areas of the ice himself. 

Sven Baertschi had a bumpy ride back into the big league, but he’s started to look a lot more comfortable over the last week or so. The pass he made to spring Byron for his break-away goal against the Senators, for example, was a rare thing of beauty. 

The underlying numbers are also complimentary of Baertschi’s play – he’s a plus relative possession player (third best on the team) despite some pretty tough zone starts so far (40%). We’re only talking about 6 games here, so caveat emptor, but after a first few underwhelming contests, Beartschi is starting to look legit. 

It goes without saying that it would be a huge boon to the Flames rebuilding efforts if Baertschi can get on back on track as a prospect. Here’s hoping he can stick around and continue build on his recent good performances. 

– Finally, although the Flames team-wide possession is hardly inspiring so far (45.9%, 4th worst in the NHL), keep in mind the club struggled out of the gate (39% fenwick, ahead of only Buffalo) and has actually been slowly improving ever since. This is despite one of the cheapest forward rosters in the league being decimated by injuries. That things haven’t totally collapsed given the 7 rookies and sophomores Calgary is dressing every night is actually a point in the team’s favour. It will be interesting to see how much Calgary can tilt the ice and if they get healthy.

  • MWflames

    Calgary certainly got several lucky wins in the early on. But as the stats say and the eye test would suggest IMO, the flames pdo is restoring itself and they are still winning. This is at least a good sign. Backlund makes this team quite a bit better too when he’s healthy.

    Now I’d argue I like our chances better because we have brodano. Leafs fans would argue they have kessel and jvr. If we are talking about strictly this year, I would take brodano over those two if I wanted my team to have the most success. Defense first IMO.

    • Burnward

      Not sure which impacts wins and loses statistically more, I think I might look at those #’s for teams that make the playoffs and those that don’t. For example do teams that make the playoffs have a greater PDO, if so it would show that they just play better(duh),and are not just lucky(i think sort of???) but I’ll have to check it out.

    • Burnward

      Corsi. PDO is more comparable to a BABIP in baseall aka his average for balls in play. While a very high BABIP and PDO can have something to do with skill, it usually is a good measuring tool for whether or not the play of the team or player is sustainable.

    • playastation

      It would have to be a combination of the two.

      That’s like looking at Ovechkin and wondering if you should look at goals or +/-.

      PDO indicates you’re shooting a high percentage and your goalie is playing out of his mind. But if you’re over 100, you’ll eventually get back to 100. If your corsi is unfavorable and your PDO is high, you can take a look and deduce that your goalie is playing well and you’re taking advantage of (relative to the other team) your opportunities.

      You want to possess the puck more as your PDO is going to regress, just increases the chance of scoring more / winning if your PDO drops below 100.

      OR.

      I don’t know anything and made that all up.

  • RedMan

    I think there’s a good middle ground to be found with how to assess this Flames team. There’s a lot of positives to be taken from this season, the kids are driving the bus and the idea of Sven sticking around with this team to go along with getting to watch Monahan and Gaudreau every night is really exciting. That being said, the Flames are not going to defy all logic for that much longer. I think it’s only natural as fans of a team to overrate your play just a little bit along with the value of your players. I think that metrics for this team are really sobering and remind us of the holes still on this team.

    The things that interest me and I think will be really telling moving forward

    1) Are the Flames willing to part ways with McGrattan or someone else on a 1 way contract when this team gets healthy? It may never happen that this team gets back to full health but I think most fans would be dissapointed if players kept their jobs based on their contracts and it forced some kids back to the minors

    2) Can we find a way to get rid of one of those contracts on the back end? That to me has to be goal #1 with this team.

    3) Do the Flames really have anything of value to trade at the deadline to contenders if they’re not in it? Glencross and maybe Hudler are the only ones that really jump out. Hudler would still be tough because taking on another years contract is tough when trading with a team.

  • playastation

    Corsi though is already becoming somewhat of an outdated stat as we begin to make more and more gains into hockey analytics. Very much like baseball and OPS. It was a useful stat at the time but then much better measuring tools came along after it. Same thing will happen with Corsi and sort of already has

  • beloch

    It’s interesting to look at the Flames numbers split up by before and after the thirteenth game of the season. That was the game against Montreal in which Backlund, Colborne, and Matt Stajan were injured (Mason Raymond was injured one game before that). Note that I’m not including the thirteenth game in either half, but it would have brought both halves up because it was a dominant game for the Flames (CF%: 59.8, FF%: 64.3).

    Before G#13:
    Average CF%: 43.7
    Average FF%: 45.6

    After G#13:
    Average CF%: 45.8
    Average FF%: 47.7

    So subtract Backlund, Raymond, Stajan, and Colborne and this team gets
    better
    ?! That’s just nuts. This must be a dividend of having what Hockey’s Future rated on Friday as the second best prospect pool in the league!

    I honestly don’t know where the Flames will finish this season. The blue-line is still lacking in depth and, if a key player like Giordano were to be injured, the wheels could really fall off the team. However, if the team gets healthier, it’s only going to get better with the addition of Backlund, Stajan, Raymond, etc.. The Flames’ underlying numbers could continue go up substantially before the season is over.

  • prendrefeu

    Hi, thanks for the responses.

    This may sound like a naïve question, but it is genuine:

    Can “regression” go in a positive direction?

    In other words, I’m wondering this when I read about a team’s CORSI and PDO and the word “regression.” Say there’s team A and team B. Team A on paper, player-by-player, looks phenomenal. They do well. Their PDO is “great”. When they have a bad spell, experts say something along the lines of “things will ‘go back to where they should be over time’ – not sure if they use ‘regression’ per se, or maybe another term with a better connotation. Team B isn’t great on paper, but they’re doing very well for various reasons (let’s not get into those reasons). Suffice to say they are doing better than the paper numbers on their players suggest. Experts use the word “regression” again… and predict that things will come down to where their paper numbers say they should be.

    What if they (Team B) does not fall back to where the predicted paper numbers suggest they will be? What if the team improves, players improve, and over time their numbers improve? Is that considered ‘regression’ to a norm?

    Thanks.

    • Parallex

      Not sure what you’re asking here.

      Yes, regression can go in a positive direction (what you’re regressing to up or down is the ‘mean’). However I don’t think what you’re talking about is regression.

      If Team B legitimately improves they can continue winning while still regressing.

    • Matty Franchise Jr

      I think the word regression applies to moving back to the norm, or the peak of the bell curve, regardless of the direction (up, down, forward, backward), but I could be wrong.

      For the second part, I think a team with “bad” numbers (low Corsi, high PDO) could improve its Corsi before its PDO regresses, but improving the Corsi is probably (I’m guessing) a long process and the PDO regression would happen before a full Corsi adjustment, meaning they would start to lose before they got better.

      So, I think it could happen but is very unlikely.

    • beloch

      Let me help you become more confused.

      Regression towards the mean is usually used on the context of a single data set. Let’s say that we’re measuring X, and X has a true average value that we don’t know. The first few measurements might be way off and give us a very wrong idea of what X’s true average value is. However, as we increase the number of measurements the mean of those measurements will gradually get closer and closer to the true value. This is what we call regression towards the mean. Note that, by random chance, you could make dozens or even hundreds of measurements before getting a good idea of the true value.

      In hockey, a lot of people (I’m guilty of this too) use regression in a slightly inaccurate and confusing manner. We’re using one data set (shots, shot attempts, etc.), which we assume gives us a better idea of a team’s true skill level, to predict the true mean that another data set, which has a smaller sample size and is therefore more prone to being misleading (i.e. goals and wins). If the outcome of this data set don’t match the mean, we say it’s due to regress.

      For example, say that the Oilers started the season out by playing against L.A.. The Kings dominated the whole game and out-shot, out-fenwicked, out-corsi’d etc. the Oilers in every conceivable way by a huge margin (let’s say by a factor of 3 to 2 on average). However, in this fictitious alternate timeline Dubnyk (who the Oilers never should have traded), stood tall and shut the Kings out. The Oilers won the the game 1-0. That’s your sample. Now go predict the rest of the season.

      What do you base your prediction on? If you go with goals, the Oilers will score 82 goals in the season (not good) to the Kings 0 (utterly disastrous!). Likewise, the Oilers will win 82 games and the Kings 0. However, if you look at the other stats, the Kings outshot, etc. the Oilers by a factor of 3 to 2. Thus you’d expect the Kings to win 3 games to every 2 games the Oilers win. This is much less unreasonable. Although a one game sample size is still dangerously small, using advanced stats gives a much more realistic idea of how the season will play out.

      So what does this mean in the Flames case? The advanced stats say the Flames shouldn’t be doing as well as they’re doing. However, the Flames’ advanced stats are also improving game over game. Thus, we expect the Flames to regress towards a mean that is actually rising. There are solid reasons why the mean is rising. Several of the key difference making players on this team are talented youngsters (e.g. Gaudreau, Monahan, Byron) who are in the process of establishing themselves in the league. There are excellent players in sick bay who will give the team a boost when they return. This makes the true quality of this team a bit of a moving target, as compared to a healthy, veteran heavy team where every player is more or less a known quantity.

      • mk

        3 Reasons FlamesNation is the best site for Flames reading:

        1. Quality articles. Kent and the crew do a great job of writing about the Flames, not just typing numbers and ranting.

        2. Quality commenters. There are some gems written in the comment section that really hit the nail on the head. Some people really seem to think the game well and are articulate in their posts. Beloch & others fit this bill.

        3. Oilers misery. If you ever need something to cheer you up, OilersNation is just a click away. Many of them have even started to sing the Flames’ praises on their own site.

    • Byron Bader

      Regression can definitely go up. As in a team’s underlying numbers are all amazing but they’re losing every game. You would expect that they would “regress” closer to the mean of teams that have performed in a similar way with similar underlying numbers (over a large sample set).

      In your case, I’m not sure that’s regression, I think, generally assumes “all things equal”. This is sort of related to my question. A team that stinks at possession or everything that generally points to long term success but is on a hot streak but then fixes many of the possession problems that’s no longer the “all things equal” equation. A large amount has changed.

      In Colorado’s case they were bound to “regress” after their year last year. But they also lost two of their better possession guys (as Kent notes) which compounds their “regression”. Whereas if they added 2-3 possession beasts would we see that “regression” or might they have fixed part of the problem and kept on keeping on.

  • Parallex

    This team is soo young right now and everyone seems to think the only place to go from here is down. What if this team progresses?? This isn’t a team filled with vets on the wrongs side of 30, this is a young and improving team.

  • Jeremy

    With centres Monahan, Granlund, Stajan and Backlund with Bennett arriving next year and Jooris and Bouma having to be sent to the wings, what is going to give? Is Backlund and his possession driving style or Stajan and his manageable 3,125 mill contract be moved?

    • everton fc

      You forgot Colborne, who could also land on the wing, but who appears to be figuring it out at centre.

      It’s why I said earlier Backlund is in an odd spot. True, Granlund may be moved to the wing as well, but I doubt it, as he’s putting up points in the pivot, and we have too many wings. So both Stajan and Backlund are in tough spots. Not to mention Knight and Reinhart on the farm. We are beginning to look solid down the middle. What if Backlund played 4th line centre, flanked by Bouma and Jooris? That would be a very good 4th line. Would Backlund do this? I wonder…

      If Granlund’s for real, does anyone prefer him over Backlund? I can’t see Granlund on the 4th line. He seems to bring more immediate upside offencively, than Backlund, who I do like. Having Monahan/Bennett/Granlund down the middle is pretty decent. As is Monahan/Bennett/Backlund. Or Monahan/Bennett/Granlund/Backlund. I like the latter. A lot. But Backlund’s not a 4th line centre. But neither is Stajan. Or are they?

      Tough calls coming up for this team.

      As for their success so far, I think they are proving they can compete with anyone, home or away, every game. Again, Hartley deserves massive applause. We continue to posts wins, with four starters in sick-bay. Not trying to drop the gloves w/any Oilers fans here, but you have to wonder why things are so much more difficult up north, with all that talent.

      I bet Hartley could have done the same, for the Oilers. Or close to it. It really proves the value of a good bench-boss when it comes to a team of youngsters, mixed nuts, and also rans.

      • MWflames

        I think people are getting a little ahead of themselves with granlund. Hes done great so far, and has pretty much convinced everyone that he’ll be an NHL regular at some point. But has he done enough to displaced backlund!? No not even close is my answer. I think people are being a little short sighted with backlund. He wasn’t great to start the season, but he obviously fighting an injury. Backlund in the last half of last year was a respectable 2nd line equivalent centre. Grandlund isnt there yet, and could still learn and grow on the farm.

        • piscera.infada

          I would say most people were probably assuming Backlund gets second line shut-down, or even first line minutes with Granlund playing 3rd line centre. That seems the most reasonable to me anyway.

  • Section205

    Great articles like this keep me coming back to this site.

    Flames PDO at “5-0n-5 close” is 1.001 (stats.hockeyanalysis.com), which should be sustainable. They have decent goaltenders this year for 82 games.

    When the score situation is “not close”, the Flames have been scoring at very high rate (unsustainable). They are getting extra bounces and padding their scoring stats as a bonus. They are currently 5th in goals per game.

    When that regresses, the Flames will score fewer goals. But if the game is “not close” when the bounces regress, then it may not make a difference in the outcome of the games. Sometimes a good team is going to come from behind and beat you, but how often?

    However… the schedule gets way tougher (especially Feb/Mar) so we really need to see more games against stronger opponents before we can know whether this team can battle for a playoff spot in the spring.

  • MWflames

    Alright, but even stajan or colborne? I’m still not convinced that granlunds play warrants waiving a proven nhler not named setoguchi. Just my opinion. I like him, and most of the call ups, but for most of them this isn’t their year to make the jump. There’s just too many bodies with one way contracts.

    • piscera.infada

      Fair enough, I’m just spit-balling. But no one is saying play “Granlund over Backlund”. My assumption is Stajan’s out for a prolonged period – the safer you can be with knee injuries, the better. Backlund slots in at 2C, Granlund plays at 3C, Colborne plays the wing with Backlund and Gaudreau. Leaving you a great third line in Sven-Granlund-Byron.

      Reinhart will go down, Jooris moves to line 4, leaving you with Jooris-Bouma-Bollig/McGrattan.

      Setoguchi gets waived as the team gets healthier – it looks like just a formality at this point. Hartley will put just about anyone in the lineup over him at this point.

      Then, when Stajan comes back, you have to make one more decision, whether that’s trade someone, waive someone (McGrattan), or send Granlund down. But that’s largely dependant on how individuals are playing and how the team’s playing as a whole.

      Edit: Forgot Hudler exists, so there goes my whole plan. Frig.

  • RKD

    This is an indication of how successful our development program is. Who would have guess so many AHL guys were NHL ready to not only step in but step up and contribute in big ways. The Flames are way ahead of where we thought they would be. The goaltending has been solid, Hiller looks very strong when he starts. Gaudreau and Barts started slow but they are just finding their way and looking pretty good now. I think Mony is our #1C, he is looking better and better every game. Hopefully Bennett can slot in as a #2 guy down the road.

  • ChinookArchYYC

    If you want to see just how well Monahan is doing statistically speaking, have a look at mikeFail’s recap of the Flames/Sens game over at Matchsticks & Gasoline. He’s done some solid work and included a player profile on Monahan this season. Monahan passes the eye test, and as Wilson has already brought up his numbers are looking very good, especially for a sophomore.

  • Grayone

    Morning all!

    Starting with Kent’s excellent piece, this thread has been one of the very best presented on FN. Some very, very sober and thoughtful replies. Well done.

    Pre … yes one can regress while improving, but generally it is short-term. I always thought of such situations as a saw-blade tipped upward. The teeth obviously are micro ups ‘n downs but in the meso or macro context, improvement continues. SJose might have their saw-blade tipped downward … time will certainly tell.

    Kent: one of your best insights was to guard against the potential false correlation of a regression as the vets get back onto the ice. It would take a discerning management to realize that.

    Finally, did not the Legend, Peter Maher, also counsel to not get too, too excited about early season results until the first quarter of the season is done? This would seem to argue that the Flames will be in the playoff hunt for sure … the big “but” for me is an injury to one of Brodano. That would be a real negative catalyst.