FN Mailbag – March 20, 2015

Mailbag

It’s Friday, March 20th which means it’s the first day or spring and it’s time for another mailbag session. I’ll apologize up front if you don’t see your question answered here. Unforunately I tend to receive a lot of requests, so I have to pare things down a bit.

This week we try to explain the Flames unexpected scoring prowess, the value of hits and the organization’s depth down the middle. 

While that is a plausible hypothesis give Calgary’s above average SH%, there isn’t any evidence that the Flames take more shots from scoring areas than average. Just the opposite in fact. From War on Ice:

CGYShotsRel

These graphics display the relative shot rates from various areas of the offensive zone for (left) and against Calgary (right) this year. Blue = less relative shots. Red = more relative shots. 

As you can see, Calgary is below average at generating shots from areas of the offensive zone, be it from the slot or otherwise. On the other hand, they tend to give up more shots from scoring areas than average. Which makes their performance this season all the more amazing. 

Somewhat related…

Likely because the Flames are grossly overachieving relative to pre-season expectations, whereas the Habs were considered to be playoff contenders at the very least. In addition, the fact that Carey Price has been excellent for a long time makes his propping up his club less noteworthy.

That said, most knowledgable Habs fans that I follow acknowledge that the club is more or less riding elite goaltending and is in deep trouble if Price regresses or gets hurt.

 

Backlund is a very tough player to accurately value and predict. His outstanding defensive capabilities are harder to measure and put a price on and his frequent injuries the last few years means a smaller body of work (and therefore more perceived risk for the decision makers). 

That said, Bob Hartley has come to lean heavily on Backlund since he warmed to him last year. No centre on the Flames sees tougher circumstances and it’s clear Hartley rates Backs as his most dependable two-way forward. If Backs can finish the season strong (and healthy) my guess is the club won’t take long to offer him an extension in the off-season. 

The good thing about having a lot of centres is it is relatively easy to convert them to wingers. Right now I would say Paul Byron and Joe Colborne aren’t even considered pivots by the team. Josh Jooris will probably be moved permanently to the wing as well. I would image Markus Granlund will be shifted to the flank at some point as well.

That leaves us with Sean Monahan, Mikael Backlund, Matt Stajan, Sam Bennett, Drew Shore and maybe Bill Arnold in the mix next season. Stajan will likely stay as a placeholder on the 4th line until the day a kid convinces the Flames to bump him for good. That could come as early as next season if Shore can take a step forward. 

San Jose fans dreams. Flames fans nightmares. Luckily still just the stuff of fiction. 

This will definitely be one of Brad Treliving’s main concerns this off-season. The Flames have two options: sign a couple of decent UFA’s or pursue a trade. 

Calgary will have ample cap space this summer and there might be a few UFA options worth considering, including Andrej Sekera, Cody Franson and Jeff Petry. Otherwise, Calgary could dangle a draft pick and a prospect or two during the draft to try to pick up another top-4 guy. 

Right now we don’t really know who will be available on either market, so it would be foolish to be prescriptive at this point. 

Hits are kinda like blocked shots in that they aren’t bad in isolation and you’d never tell players not to do them, but nevertheless they don’t tend to correlate to wins or shot rates. That’s because, as posited, players who are hitting a lot don’t tend to have the puck. Also, it something that is difficult to test since there is no clear definition of a “hit” in the NHL. As a result, every building counts hits differently, meaning the data is “dirty” and not terribly reliable. 

Here’s a recent look at hits correlating with shots rates on the team and player level. Most of the correlations aren’t significant. There’s some suggestion that hits and shots against have a positive relationship (more of one, the more of the other), but it’s not terribly strong or conclusive. 

  • Burnward

    One quick comment regarding the Hextally. I think we come to different conclusions from the data.

    The team takes less shots than league average, with the highest percentage coming in the high slot area…an area that they shoot 40% better (?!) than league average in (if I read the shooting percentage chart right).

    That suggests to me a more selective shot taking approach and skill on display.

    It also shows the Flames allowing a lot of prime chances from the high slot…with the shooting percentage way down. This to me illustrates the shot blocking prowess of this team.

    Am I way off on this?

    • It’s a leap to go from “better than average percentages” to the causal assumption of “shot selection and deterrence”. Particularly in a league when no team (and very few skaters) have shown persistent skill in actually influencing the percentages.

      From past studies we haven’t found any defender who can influence up SV% to any meaningful degree (at least, not to a degree that is visible through natural variance). As for players, there’s a few very elite guys who can do drive SH% up over the long-run.

      • beloch

        I think what Burnward means is that the proportion of shots the Flames have taken from each area is a little skewed towards the high slot.

        ——- Total Shots —- % of Total —- League avg %
        Point: ——–1281 —– 44.6% —– 46.2%
        High Slot: — 957 —– 33.4% —– 31.4%
        Low Slot: — 631 —– 22.0% —– 22.4%

        The Flames are taking a slightly larger proportion of shots from the high slot, and slightly fewer from the point. Now take a look at the sh% from each region:

        —– Flames sh% ——– League avg sh%
        Point: ——– 1.3% ——– %1.7
        High Slot: — 5.7% ——– %4.1
        Low Slot: — 8.9% ——– %9.1

        The Flames sh% is below league average everywhere but the high slot, where it’s significantly above average. This is the same area where they’re taking a higher than normal proportion of shots from. Is this significant? Not to a large degree. If they had taken shots in the same proportion as the rest of the league, they’d have scored just 2 fewer ES goals. (This is all even strength. )

        This skew is more pronounced on powerplays, and hugely pronounced at 4on4. In 4on4 situations, the Flames’ shots from the point, high slot, and low slot were 0.676, 1.07 and 1.12 times the league average respectively. That is not subtle. The Flames are clearly choosing to shoot from more dangerous areas of the ice when 4on4, and this is probably why this team is so deadly in 4on4 situations.

        My best guess is that we’re seeing a skew from the Gaudreau-Monahan-Hudler line, and possibly a couple of the better offensive defencemen like Giordano and Wideman. The forwards in this group are good at shooting from closer in and all have very high sh%’s. At ES the effect is small because these players are not on the ice as much as the rest of the team. However, that changes on powerplays and 4on4. In these situations Monahan, Gaudreau, and Hudler play much more, and the skew towards a higher proportion of shots from the slot is more pronounced.

        • Section205

          You are right. Team Hextally For CGY shows that Flames are pretty much league average in shots taken in the high slot, and they take very few shots from the low percentage area outside.

          The vast majority of Flames’ shot deficiency (the amount below average shots taken) is due to fewer low percentage shots taken from the outside. We take approximately 2 fewer (than average) shots per game from the outside.

          If Flames had taken (wasted) another 142 outside shots, of which only 1 would go in the net, then their shooting percentage and PDO would be much lower…and of course they would have lost more games. Flames are being rewarded for their patience with the puck.

  • The Real Slim Brodie

    when we talk about the flames talking higher quality shots, and then reduce the query only to a question of zone on the ice, i wonder if we’re not impoverishing the discussion severely. what about traffic? or a player moving to change an angle before taking a shot? etc. etc. surely shot quality is more complicated a phenomenon than pure location?

    • beloch

      …and what about the pass to the player who winds up taking the shot? Johnny’s tape to tape pass to Mony in the slot was so good that the puck had a excellent chance of making it past the goalie and in the net!…and all Johnny had to do on his goal at the doorstep was flip the puck in…a good play all around – lot to factor in besides the point on the ice that the shot was taken from!

      A lot of factors play into the quality of shot taken…If Johnny’s pass was skipping along the ice the chance of Mony scoring goes down. Good hockey sense also factors into shot quality – knowing where your teammates are on the ice and where they are going to go!

      Unfortunately, all this stuff would be hard to quantify…

      • Avalain

        So, I’m not exactly one of the stats guys or anything, but I think that things like this are handled by using larger sample sizes and by the fact that this is the NHL, so realistically everyone is really good. Where you shoot from matters but it’s not like our guys are just better than everyone else (or they might be but we don’t know for sure yet). So for example, the pass from Gaudreau to Monahan was great, but so was the pass that the flyers made on their goal. So, basically, these things like passes before the shot and stuff are just ignored and we hope it all evens out in the long term.

  • Burnward

    Hi Kent!

    Just wanted to add to the discussion about flames quality of scoring chances. The following graph is 7 years of NHL shots and goals plotted vs. distance of the shot.

    http://imgur.com/azdlgGZ

    It shows a huge jump in shooting percentage from say a shot from 10 ft out then a shot from 30 ft out.
    In fact, from this sample size (7 years of data) the average shooting percentage from 10 ft out is ~15% while from 30 ft out it drops to ~5%.

    I’m gonna use the Flames here to make my point, but say the flames are attempting all of their shots from 10 ft out and they get say 20 shots. The statistically probable outcome would be that they get 3 goals (0.15*20=3).

    Meanwhile if the other team gets 40 shots from 30 feet out, the statistically probable outcome is that they would get 2 goals (0.05*40=2).

    So purely talking from a theoretical standpoint with this graph as the basis of the discussion, the Flames could get outshot 40-20 and still be the much better team in terms of getting “higher percentage” shots and therefore are much more likely to win the game.
    Assuming no other shot attempts in the game, the Flames would have a 33.3% Corsi rating, while being the much more likely team to win.

    This is the xDiff stat I’ve been pumping a little bit around here.

    By weighing every single shot like this, you can get a Fenwick score that adds this expected shooting percentage as an extra measure of accuracy.

    This isn’t changing anything we know fundamentally about possession stats, it’s simply adding an extra factor to more accurately paint a picture.

    There’s a reason every single coach I’ve ever had in hockey preaches “getting to the dirty areas” and “getting good scoring chances” and this is quantifying exactly that!

    And for the record the flames are 23rd in xDiff, a much better representation of where they are as a team. They, as a team, don’t drive possession but when they do it’s closer in and therefore more likely to score.

    Have a read here is you have time about the shortcomings of Corsi and Fenwick and why xDiff is just a fundamentally better “possession” stat.

    http://fourthlineheroes.com/static/pdf/The-case-for-xGoals.pdf

          • RedMan

            Technically, it is the “circle of courage”.

            I had the privilege of visiting with Dr. Broken Leg about this as it relates to youth and human development. Belonging, Mastery, Independence and Generosity… First nations human development theory that bears many striking similarities to Erik Erikson’s better known 8 stages of (psychosocial) development when you look into it.

            thanks for noticing.

    • Southern_Point

      The issue with Xdiff

      is while it sounds nice its developers haven’t done a very good job correlating it to wins or points. As in they just haven’t yet, and I suspect if they do it will not be a significant improvement or over non-weighted shot attempts if any at all thus not worth the extra effort in attempting to calculate it.

      The xGoals stat xdiff is derived from is somewhat more interesting than xdiff. But the interesting part of it (predicting offensive output year to year) is just as easily obtained by looking at PDO and Shot attempts in combination.

  • Burnward

    Can FlamesNation please update the scoreboard on the right column in a timely manner. It looks extremely poor when games from many nights prior aren’t even updated and upcoming games are not displayed. This is basic stuff.

    More frequent polling wouldn’t hurt either.

  • joey joe joe jr shabadoo

    I get that teams with high Corsi are more likely to be successful. Nobody argues that.

    What I don’t get is why so many writers feel that teams with low Corsi have NO chance. If I was a coach and had a team that lacked elite talent, I would look for a way to maximize what I do have, and look for unconventional ways to win.

    Teams like the Flames, Avs and Leafs (when they were good) have proven that you can win even with low Corsi for long stretches. The Avs had a 112 point season last year!

    So what if the strategy is sustainable tomorrow? At the end of the day, aren’t the coach and players being paid to win tonight’s game in whatever way they can?

    Isn’t 2 points with low Corsi still worth the same as 2 points with high Corsi?

      • RedMan

        first, how? you’re allowed to say what you want within reasonable limits. what really discourages it is having the authors on this site put in work to make articles and every single time a new one is posted there are people in the comments telling us why it sucks.

        I don’t write any more because I was tired of having that happen.

        • RedMan

          Two points:

          1) It doesn’t take much of a look back (last season) to see that Flames fans embraced analytics when it was the Avs and Leafs as the outliers. Arguing against fanboy homer-ism is a fool’s errand. Flames are winning, so every excuse in the book (ie. perfect passes to wide open nets – HAHAHAHAHA!) gets trotted out.

          Personally, I find I can cheer on the team and enjoy the wins without having to bury my head in the sand. Year two of a rebuild people, it’s ok to win by being a statistical outlier.

          2) I liked your writing. Quite a bit. But you need thicker skin. Criticism generates discussion. Otherwise you get the CP boards.

          I suspect increased fan worship is the reason we no longer see long time posters such as Vintage Flame, schevvy (one of my favorite posters) and John Deere Green.

          • Section205

            Colorado was giving up 33 shots per game last year (25th). Toronto was giving up 36 shots a game (30th).

            Calgary gives up 29 shots per game. 11th best in the league.

            These facts sponsored by commenters labelled as “fan-boy homers, worshipping with their heads buried in the sand”.

          • RedMan

            where are they in shot generation? maybe look at both halves.

            @the-wolf

            I have no issues with criticism. I have an issue with the same criticism coming up again and again and again and again in perpetuity.

          • RedMan

            what? shot generation is the amount of shots the team has generated. they are 28th in shots for and have a -114 shot ratio, even though they are 11th in defending shots and have an insane PP/PK ratio.

            you are arguing the wrong point.

          • Section205

            Are you kidding? We were just talking (before all the sighing, whining and crying) about the shot deficiency being largely due to 142 shots fewer than average being low percentage outside shots. You’re lamenting -114 shots overall. How is that the wrong point? Unbelievable.

            The “insane PP/PK ratio” is a credit to the team. The Flames don’t get an “insane” amount of Powerplays. So why would you expect extra shots generated there? They take very few penalties, which helps them keep shots down. Which is why Colorado and Toronto are not really comparable to this team. They leaked shots on goal and it caught up to them.

            You like many others are focusing on the “ratios” or the “net plus/minus” values instead of recognizing the two absolute values. Toronto/Colorado’s problem was at the other end of the ice.

            Calgary’s problem is on the offensive side, but just look at the hextally and you can see why I don’t worry about 142 of those outside shots. Their shooting percentage should be above average because they waste far fewer shots from outside.

            Obviously this team is developing, and in a few years should have better corsi and fenwick too.

          • Megamind

            Then don’t keep posting the same article or point of view.
            And EVERYONE here can understand that your “sigh” is you saying that you are smarter than the rest of us.
            Statistically that may be possible, but I believe that it may be improbable.

          • RedMan

            The team is defying the stats.

            I myself have asked for explanation and insight beyond “luck.”

            But a lot of the comments are ridiculous attempts at working around the facts. Avs and Leafs fans stated different/but the same statements all last season.

            I just don’t get where the angst comes from to cheer for a team that’s playing over its head? As I said, year two of a rebuild. Next year the 3rd pairing D should be a lot better, Bennett should help, others will have another year of experience.
            Why the insistent need to justify?

          • RedMan

            is there angst here over cheering for a team that is achieving over their heads?

            where is this angst?

            I think we all are cheering for a team that is playing over their heads, which is making us all proud.

          • RedMan

            by and large the vast majority of people here HAVE “embraced analytics”.

            I don’t know why one assumes that because people got tired of having blog after blog telling us why the Flames actually suck and why we are stupid fans, or that some don’t agree with the “luck” conclusion, that the fans here are fanboys and don’t embrace analytics.

            It is too bad that Justin was discouraged from writing – he was awesome and is missed, but in my opinion, it was the blog authors that need to adapt, not the fans here.

            Are people really so stupid that they “embrace” new stats when it makes colorado and Toronto look bad, but then quickly hate it when the same identifies the Flames as the new outlier? again, that amounts to another backhanded insult of people who keep coming here – and it’s these kinds of insults that generates the arguments, not the fact that people don’t believe in them when they are less favorable to our team.

            If you notice the last couple weeks there has been a whole spate of blogs that are not focussed on CORSI and FENWICK – and the fans here have loved them and appreciated them. But to make the accusation that the people here have their heads in the sand, are dippy homers with flaming red glasses, to outright call the fans here idiots – either by implication or outright, over and over and over… this has lead to the constant negativity that has really soured this sight. and you can’t blame the fans that keep coming and paying the bills here with the click-throughs – it is the fault of the authors and sight management who kept hammering the same stats with the same “Flames suck even if they are winning, by luck” conclusion.

            everyone knows they are rebuilding, everyone knows they are overachieving, everyone knows that by the numbers, on paper, this team should finish 25th in the league.

            You don’t like peoples “better passes, better scoring position” arguments, but others don’t like the “science has proven that luck is the only answer”, and the truth is nobody knows for sure, except to say that “luck” is a used when we don’t know what the actual cause is.

            and on that note, what is luck? I’d suggest luck is when a puck careens off of 4 bodies and goes in off of the net and then the goalies back. Flames have had their share of luck, but as statisticians who keep preaching regression to the mean, surely there is a comprehension that this type of luck – the things that truly are luck, generally balanced out, just like PDO, and teams get and give generally speaking in equal amounts over longer sample sizes. I think that by now – after over 85% of the season played we can agree that yes – the Flames are beating the odds, but no, it is NOT luck, which averages out.

            and there are some here who are skeptical of CORSI – count me in. So what? isn’t that OK? CORSI may turn out to be the stat of stats, and we will have to eat our hats, and it may not – time will tell how valuable it becomes, but there is skepticism by some, so what? the authors should embrace the skepticism and not call everyone homers and so forth.

            OK, this is about 1000 words past the TL:DR threshold – I just hope this site gets back on track as the best go to site for flames stuff and quits this nonsensical fighting over new stats.

            let’s just all get along, those who enjoy corsi and those who don’t, yet.

          • Megamind

            Valid points with the exception that my comments were directed to those who make those types of comments and not the entire board. Again, angst.

            Plus, there’s a fundamental difference between asking for an explanation of “luck” or presenting a comprehensive alternative view on the stats vs an endless variation of “the Flames have found a new way of playing hockey.”

    • Robear

      Hey Gang,
      Not sure if this helps, but the recent descent of some of the threads into “us” vs “them” on this site I believe stems from the challenges of communicating with people who are emotionally invested in subjects.

      Readers, posters and writers alike are writing for or viewing this site as Flames fans (for the most part, trolling aside). So everyone has an emotional connection.

      The current analytics (corsi, fenwick PDO etc.)purists believe that the analytics prove that the Flames play is unsustainable and is doomed to regress.

      Those who are not as convinced by the analytics are either a) struggling to enjoy the Flames current results while hearing that its doomed to regress; b)trying to find an alternate reason/justification to the success that incorporates analytics; c) trying to ignore ALL the analytics, or (my favourite); d) trying to do all 3 at once after drinking too much beer while watching the game.

      I can see how the constant questioning would wear on those who are converted to the Ways of the Corsi (poor Star Wars analogy.. my bad…its late), but I can also appreciate the challenge faced by those who aren’t as convinced.

      The ongoing debate reflects those viewpoints. But its emotionally challenging to expose yourself to alternate view points.

      And when those view points are different from ours and expressed in a way that makes people feel like their views are less valid, then people feel attacked and their ability to respond on a purely intellectual level decreases The more emotionally charged the subject then the greater the emotional response.

      To illustrate this point apparently when people are given life altering information (cancer diagnoses etc.) their ability to respond to the information decreases by over 5 grade levels. That means that the average high school graduate will respond to emotionally charged information at a Grade 8 level!!

      Most of this reflects the adage that people generally dont remember what you said so much as how you made them FEEL. Totally fluffy, I know, but very real. This explains why people gravitate to like minded forums where people share the same view on subjects.

      Further complicating all this is that all communicating on blog sites is by written word when its well known that true human communication is 80 to 90% NON VERBAL. That’s right folks… its not what you said… its HOW YOU SAID IT.

      I’m starting to ramble a bit, but my point here is not to throw stones, but to encourage everyone (authors, posters and readers alike) to remember that there are real people behind these words and to cut them some slack.

      Hopefully (if you made it this far into my post) this helps.

      Go Flames!

  • Section205

    I just wanted to comment on this and a few other things:
    Point 1: The problem with the advanced stats crowd is they think that the only stats that matter are the ones they agree on; lets not forget that just a few months ago they were still voodoo and unknowns to many in the hockey world. While Corsi and Fenwick are good tools and can help indicate which team should win over the long haul they are not always right. a team like the Flames and I would suggest the Haabs don’t have as many big bodies up front to drive possession so they rely on other traits. These traits slash style and expectations will not always show up in the advanced stats but I believe each organization will be tracking different things for their own use. Measuring quality shots(rebounds), hits and successful breakout passes could all be used by a team like the Flames and Habs to measure successs.

    Flames need to sign Backlund to a three year deal for the following reasons: Hartley trusts him with the dirty jobs and his presence will allow Bennett time to develop into the center we want him to be. If Backlund is here Bennett will be sheltered and allowed to develop next year. This also is related to the depth at center in the organization: Monahan, Backlund, Bennett, Stajan, Granlund(Hartley trust him too),Arnold and the guys who can fill in like Jooris and Colborne.

    How to help the depth on defense: start by signing a 6’2 200lb R handed shot NCAA free agent. I have no idea how good Morrison will be but he adds depth to an organization that needs it. If he is near the level of the guy the wings signed a couple of years ago he will be an upgrade, BT add a couple of more and look at the UFA market, draft well (we have 6 in the top 90 make at least 3 defenders), develop the ones we have, get the Europeans over here and if needed make a trade for a 4/5.

    No matter what the results of this yer end up being lets remember this is year 2 of a rebuild and it will take a couple of more years to be a real contender(who knows we may slip like Colorado did) keep building, keep turning over rocks to find gold and keep developing the kids don’t chase the cup let it come to you.

  • Burnward

    Does anyone know anything about the big monster the baby flames have signed to a PTO. Olag Y from belerus 6’7″ 230+lb defender who played at UMass. Also looking at the Baby Flames roster makes we understand why Spoon is not up here. I hope the baby Flames I hope they make the playoffs. I tried to find news on when Ortio might return to them but had no luck.

  • Megamind

    Helps when I spell my name right:

    Valid points with the exception that my comments were directed to those who make those types of comments and not the entire board. Again, angst.

    Plus, there’s a fundamental difference between asking for an explanation of “luck” or presenting a comprehensive alternative view on the stats vs an endless variation of “the Flames have found a new way of playing hockey.”

    • Section205

      I hear what you are saying but you must accept that over the decades the game does take massive shifts.

      Everyone in the 70s played tough physical defensive hockey. 80s and 90s were a much different “way of playing hockey”

      Then we get the left-wing-lock and the dead puck era begins.

      Things do change.