Black Box: Week 2



Every Monday we’ll be posting a weekly statistical update on how the Flames have fared with each of their individual players on the ice.  Last week was rather tame, but now that we have four games worth of data, things ought to start getting interesting.

Week 1

OZQoC Charts (explanation)

Coach Brent Sutter has taken the surprising step of using Jarome Iginla, Alex Tanguay and their centre-du-jour (David Moss or Roman Horak) in the toughest situations. Despite their historical inadequacies defensively they’ve started largely in their own zone against the best competition, although in fairness it’s not easy to pick spots on the road (the home team gets the last line change). 

Olli Jokinen, Lee Stempniak and Curtis Glencross have been the primary benefactors, enjoying the easiest ice-time, by far.  Truth be told, the Flames are doing the exact opposite of what we expected. Call it a Reverse-Vigneault: expose your weak defensive players, and carefully shelter your fastest two-way line.

Not much to conclude on the blue line, though – all defensemen are being used roughly equally, although you could argue Mark Giordano and Scott Hannan might be getting a little boost.

Even-Strength Scoring (explanation)

Unfortunately we lost the scoring chance data for their lone victory, which is going to skew the numbers a little bit, since all the other calculations are based on all four games.  For that reason, don’t be surprised if scoring chance data doesn’t correlate as well to Corsi or goal-based data as usual.

Normally all the numbers are adjusted for ice-time, and presented as a rate over 60 minutes, but thanks to the missing data that gets a little too confusing for the scoring chance data, so those are just the raw numbers.

Player         ESP/60 CEF CEA  CE% SCF SCA SC%   GF   GA    G%
Lee Stempniak    2.4   61  49 55.4% 13 12 52.0% 4.76 5.95  44.4%
Olli Jokinen     3.4   64  51 55.4% 16 10 61.5% 4.46 4.46  50.0%
Curtis Glencross 3.7   59  55 51.7% 13 10 56.5% 3.68 4.91  42.8%
Scott Hannan     1.9   47  47 50.1% 14  7 66.7% 1.92 2.88  40.0%
Mark Giordano    1.0   54  56 49.1% 15  6 71.4% 2.89 2.89  50.0%
Jay Bouwmeester  0.0   55  59 48.2% 12 22 35.3% 4.15 2.49  62.5%
Chris Butler     0.9   52  57 47.9% 13 22 37.1% 4.50 3.60  55.6%
David Moss       1.3   51  56 47.7%  6 16 27.3% 1.25 1.25  50.0%
Alex Tanguay     2.1   56  62 47.7% 13 20 39.4% 2.12 4.24  33.3%
Matt Stajan      1.4   47  53 47.2%  6  7 46.2% 2.77 2.77  50.0%
Derek Smith      0.0   49  55 46.8%  0  2  0.0% 0.00 4.42   0.0%
Niklas Hagman    2.7   50  57 46.8%  7  5 58.3% 4.04 0.00 100.0%
Jarome Iginla    0.0   54  64 45.8% 13 20 39.4% 0.91 2.73  25.0%
Rene Bourque     2.6   46  60 43.4%  6  5 54.5% 3.82 3.82  50.0%
Tom Kostopoulos  0.0   39  51 43.3%  2  1 66.7% 0.00 2.06   0.0%
Cory Sarich      0.0   40  61 39.8%  5  9 35.7% 0.00 3.65   0.0%
Tim Jackman      0.0   34  58 37.0%  2  5 28.6% 0.00 5.94   0.0%
Anton Babchuk    2.1   41  78 34.4%  6  9 40.0% 2.05 2.05  50.0%
Roman Horak      1.6   28  71 28.5%  3  3 50.0% 1.58 1.58  50.0%

Given their softer assignment it comes as no surprise that the Jokinen-Stempniak-Glencross line is the only one enjoying territorial advantage over their opponents.  Same goes to the Hannan-Giordano pairing, but to a lesser extent, even though they’ve enjoyed a lot of great luck converting that advantage into scoring chances (but unfortunately not goals).

The Bouwmeester-Butler pairing has been outplayed and seriously out-chanced, but has been fortunate enough to be the only unit to out-score their opponents.

These stats can also help confirm that we’ve been seeing with our own eyes – that Anton Babchuk and Roman Horak are really struggling out there – and Cory Sarich too.  The bigger surprise is how badly that fourth line is doing given how successful Kostopoulos and Jackman were last season.

As for that top line of Iginla-Tanguay-Whoever, they’ve had some tough ice time, have generally been playing without the puck, and have really paid the price.  Once Sutter starts using them more wisely, and their luck improves, their numbers ought to bounce back – and Morrison is likely to get credit assuming his return is timed just right.

Special teams (explanation)

The power play blue line pairing of Alex Tanguay and Mark Giordano is getting the bulk of the power play time, but have been relying on luck for their scoring, instead of puck possession.  Have you been more impressed with the play of Jay Bouwmeester and Anton Babchuk?  The Flames are generating shots at three times the rate with that secondary pairing.

Player              TOI/GP PTS/60 CE/60
Jay Bouwmeester      1.3    11.6  151.1
Anton Babchuk        1.5     0.0  139.3
Chris Butler         0.9    17.1  119.4
Niklas Hagman        1.8     8.3   90.8
David Moss           1.5     0.0   80.2
Olli Jokinen         2.8     5.4   80.2
Rene Bourque         2.9     5.2   73.4
Lee Stempniak        2.1     0.0   65.6
Jarome Iginla        2.9     5.1   61.7
Alex Tanguay         4.0     7.5   48.5
Curtis Glencross     1.3     0.0   46.9
Roman Horak          0.7     0.0   46.4
Mark Giordano        3.6     4.2   46.0

The primary penalty killing duties on the blue line have been placed on the shoulders of Jay Bouwmeester, backed up by Scott Hannan, Mark Giordano and Chris Butler, the latter of which is the surprising high performer.

Player           TOI/GP CE/60
Jay Bouwmeester   3.8   107.9
Scott Hannan      3.1   120.8
Mark Giordano     3.0   114.1
Chris Butler      2.9    88.5
Rene Bourque      2.7    88.8
Curtis Glencross  2.2    82.1
Tom Kostopoulos   2.1   139.0
Lee Stempniak     1.7   117.3
David Moss        1.6   119.1
Roman Horak       1.5    99.4
Cory Sarich       1.0   183.8
Matt Stajan       0.9   185.0
Alex Tanguay      0.8   136.9

Up front Rene Bourque and Curtis Glencross have unquestionably been their top penalty killers. 

Goaltending (explanation)

Only one quality start in four starts for the Flames, and that was Kiprusoff’s game three win over the Montreal Canadiens.  Neither goalie has played particularly well so far, but at least Kipper’s almost average.

Goalie           GS QS  QS%  ESSV%
Miikka Kiprusoff  3  1 33.3%  .909
Henrik Karlsson   1  0  0.0%  .844

That’s it for this week folks.  Now that we have some meat on the bones, how does the format of our weekly statistical review look?

  • ChinookArchYYC

    “Call it a Reverse-Vigneault: expose your weak defensive players, and carefully shelter your fastest two-way line.”

    Call it – Sutter-Love: I will make you better by putting you in the worst situation possible based on your current play and available skill set.

  • icedawg_42

    I remain unconvinced that the location on the ice where a face-off is being taken has played any role whatsoever in Sutter’s choice of where to play players. There is just far too little variation from expectation to justify coming to such a conclusion.

    I have worked out actual and expected faceoff starts for all the Flames players at this spreadsheet (

    Let us take the two most extreme forward cases on each end of Zone start, Iginla’s 31.7% and Stempniak’s 62.1%. What is behind those percentages.

    Iginla had raw starts of 13 OFF, 28 DEF and 22 NEU. Also, Iginla has had something like 80 EV shifts ( doesn’t break down shift numbers by game state. He has 23 shifts per game, ~3 PP minutes per game. He only has 16 seconds SH. Call that 3 PP shifts per game, the other 20 at EV)

    If Iginla had started faceoffs in accordance with the proportion of his EV ice time to the proportion of where the Flames have had faceoffs, we might have expected him to have 4 or 5 more offensive zone starts and maybe 5 less defensive zone starts. So over 4 games, this has affected maybe 9 or 10 of Iginla’s 80 shifts. 2 or maybe 3 of his 20 shifts per game. Unless someone has some actual evidence of Sutter saying he is using zone starts, I hardly believe that in the face of all the various factors that go into the decisions that make up the coaches decision, that 2 or 3 shifts per game is evidence he is making a concerted effort to put these guys on the ice in their own end more than their playing time would dictate.

    If we look at Stempniak, the story is much the same. His actual starts are 18 OFF, 11 DEF, 29 NEU. He takes 21.3 shifts per game, call it about 19 EV shifts (2 min per game PP time). If he had taken faceoffs in direct proportion to his EV time on ice he would have had about 5 more offensive zone draws and 6 or 7 more defensive zone draws. So we are looking at 11 or 12 shifts over 4 games, or about 3 per game. 3 out of 19 shifts. Again, taking into account all of the other competing reasons to put a line on the ice, I don’t think this demonstrates anything about a Coaches usage of players in a game.

    I also went back and looked at last season’s numbers. Aside from Babchuk (and I am still not sure I can get his data right as it appears puts all of his zonestarts from the season into his Calgary numbers), I can see no player whose zone start totals in the offensive zone or defensive zone appear to have been influenced more than 1 draw per game more than what we would have expected given their ice time and where the Flames started their shifts.

    I don’t dispute that you should use zone start numbers to put corsi/fenwick possession numbers into context when comparing between players. But until we start getting in Sedin/Malhotra like zone starts over a whole season I see nothing at all that demonstrates that the zone start numbers say anything material about how Sutter is deploying players. To keep trumpeting these numbers as if that is the case is just postulating a false narrative of the kind that I see so frequently decried on this site.

    • Interesting work. I think you make a good argument for taking these things less seriously in the extreme, particularly in small samples.

      That said, I’m wondering if we can properly project potential face-offs based on actual face-offs taken by ES ice time. By that I mean it may be that coach’s decisions influence the number of offensive and defensive zone draws by the match-up and face-off choices they make.

      For example, the Flames have had a lot more defensive zone draws than offensive so far. In part, because everyone playing against decent players or starting out in their own zone for the Flames are getting beat up. The puck spends more time in the defensive zone and results in more defensive zone draws.

      If match-up and face-off decisions shift puck location and o-zone finish rate, then bad decisions are necessarily self defeating and wouldn’t manifest well in expected o-zone and ZS% by ES ice if we take the number of o-zone and d-zone faceoffs as more or less constant. I assume with better decision making there’s more offensive draws to go around and less defensive zone face-offs.

      That’s theoretical and I have no study to back it. It’s possible it merely means teams with better players who drive the play north simply have better ZS% and the coach’s decision making in regards to faceoffs is meanigful only in aggregate over the long-term.

      All that said, I’d much, much prefer to see Iginla have Sedin or Tavares like ZS percentages already:

      if only because that’s role I’d prefer to see for Jarome over the long haul.

      • I agree that it would probably be a good strategy for Sutter to use ZS to try and give Iginla easier ice time. I just don’t see what I would call significant evidence that he is attempting to use particular players in particular face-off situations over the last season or the start of this season. If he is using that as a decision making criteria it appears to be secondary.

        I also agree that there is a “Which comes first, the chicken or the egg” element to these numbers. It may be that if Calgary starts earning more O zone starts, Iginla’s ice time gets cut back so other get those O zone starts. In which case Sutter is being stupid.

        The NYI numbers are interesting but their team zone start percentage is at 58%. I will be interested to see how Tavares’ ZS numbers look once the (what I assume to be) invevitable shift occurs.

        The Vancouver numbers are most fascinating though. They are right around 50% Zone Start as a team, but both the Sedins’ and Hodgson have high Zone Start and then Lapierre and Malhotra are absolutely buried. I can only assume that Vigneault almost totally ignores all other factors in which lines to put on the ice. (I would be damn curious what the 12 D Zone starts are for the Sedins. 4th line? Icing?)

        Trust me, if the Flames numbers looked like this you would be seeing no argument from me as to strategy.

        I also notice that the Sedin’s only have 16 EV minutes per game. Perhaps they would have more if Vigneault was willing to use them for more D Zone draws?

        • Related: last year I noticed Sutter making offensive zone faceoff decisions based on which trio seemed to be “on” that night. For instance, when David Moss joined the 4th line and started crushing things, he had a couple of games where he started them a bit more often in the offensive zone. A kind of “you look more dangerous, so get out there” type of thing.

          Maybe that’s why we see the Jokinen line with the edge over everyone else so far. Once or twice a game Sutter has thought “they look better than anyone else so far” and opted for them for another o-zone start or two.

          Just a guess.

          • icedawg_42

            This is as good an explanation as I have heard for the uptick in zone starts for the Jokinen-Glencross-Stempniak unit.

            Does that mean that Iginla-Tanguay-Moss aren’t getting high ground because they need their minutes, but the limited O zone starts have gone to Jokinen’s line?

            If that is the case we can’t be rid of Sutter soon enough. I think this homestand will be a real litmus test in that regard. If he is still not getting Iginla et al. high ground, bring out the lynch mob.

          • It’s something worth keeping an eye on. I have my hands full with chats and scoring chances during games, so I actually miss a lot of other nuance unfortunately. Hopefully others around here can take note.

        • I think I can agree with you that Sutter is likely ignoring faceoffs absent rare factors (trailing in the third period and I believe he actively sheltered Backlund last year).

          Sutter is clearly following other information to make his face-offs decisions. I think that would be “who is the other coach playing?” (one of the top two lines) and “get Iginla as much ice time as possible”. Which would give rise to the apparently random scattering of ZS percentages we see so far I think.

  • icedawg_42

    I don’t see why everyone’s all over Brent Sutter. For my money, & yes I personally pay for the tickets, it all on the players. There is no leadership in this group, no sense of urgency when they play etc….

    This is an old team going no where except to the Bank !

  • icedawg_42

    Yeah, I dont get it myself, I expect better out of Brent – it’s easy to see that Iggy is NOT a 3 zone player. He’s an offensive weapon, so use him as such and put him in a position to succeed. The Chud Moss Scoreface trio has proven time and again the ability to work well against the heavies. Hopefully this gets sorted out really quick. 6 points back after 4 games is getting ugly.

  • Also of interest: Stajan playing nobodies, but having one of the worst ZS ratios on the team.

    Of course, with the Flames being really lousy at driving play thus far, they aren’t a lot of offensive zone draws to go around. That’s probably why you see all the 4th liners below 50%.

  • jeremywilhelm

    i hate to sound like a broken record, however i must state my opinion. butta is a highly overrated bench boss. the world junior success schews things, a fire hydrant could have coached those teams to golds.iggie gets man handled and mauled against the heavies,yet butta continues to implement the same unsuccessful strategy. time to move the much maligned matt cottonelle to the top unit and play the favourable matchups card butta. as of now everything butta has done has been like lipstick on a pig,its still a pig buttabing.

  • xis10ce

    I’m curious to see what this 6 game home stand does to the numbers. Given that we will have home ice advantage and last change we get the luxury of choosing who plays when.

    I’m hoping Sutter start taking advantage of that last change to shelter Iggy.

    • He always has the choice of who to play when. The question is who do they play against. Why he is defaulting to the Iginla line as tough minutes line on the road is still a question mark. Especially with Moss struggling on face-offs.

  • Really eye-opening stuff for me. And it leads to the question…

    What the hell is Brent Sutter doing? How many times does he have to re-visit the “Iginla as hard minute forward” failed experiment? This is what he did with Jokinen-Iginla-Moss to start his first season here and the results were pretty much exactly the same.

    Say it with me Brent: Iginla needs the high ground or he gets beaten up.

    I don’t know. Maybe he’ll go back to splitting the tough assignment with the Jokinen line when Morrison or Backlund returns and Moss goes back to wing. I still don’t get it though.

    Now we know why the second line has looked like world beaters and Iginla has all of zero ES points in four games.

    • icedawg_42

      No Kidding.

      Hell… that trio helped Phil Kessel look like he was channelling the Sidney Crosby Forza the NHL is missing right now.

      There are times that I legitimately wonder if Iggy, Tanguay + other aren’t over their heads playing the second line of very strong teams like Vancouver vs the Sedins, Or San jose vs Marleau, nevermind the top comp.

      Playing the toughest….. man… that’s just a bad recipe.

      We’re 1 and 3. It’s early, but it’d be easier if there were signs pointing the right direction.