• In defence of Andrew Cogliano

    by  • September 29, 2010 • Uncategorized • 14 Comments

    Andrew Cogliano is a horror show when it comes to taking draws. Hopeful quotes aside (there really ought to be some sort of a name for those quotes when a player comes to training camp or spring training and announces that he’s in the best shape of his life, never felt better, lost 15 pounds so he’s quicker, put on 20 pounds so he’s stronger, worked on faceoffs all summer…if sport is replacing religion as the factor unites society together, these quotes are the equivalent to the portion of the mass in which the priest goes through the ritual of blessing the eucharist)…sorry, I’ve lost my train of thought.

    Andrew Cogliano and faceoffs. Right. While last season was an improvement for Cogliano, he’s still simply horrible at them, three years into his career. After being a 37% or so guy on his career at ES to that point (with faceoffs, it’s sensible to simply look at ES because SH faceoffs are harder to win and PP faceoffs easier; just looking at ES permits you to compare apples with apples), he won 42.3% of his faceoffs last year. Improvement! Except that the improvement had him tucked neatly into 170th out of the 189 guys who took at least 100 draws. So still a nightmare.

    One significant thing has changed though: Andrew Cogliano is, for the purposes of faceoffs, basically no longer used as a centre. I was fiddling around and it strikes me that the proper metric to use is probably ESFO/60 – how often is a coach sending a guy out for a faceoff? Cogliano took 19.45 ESFO/60. For some context, Zach Stortini (120th out of 189 with a 48.0% winning percentage) took 15.53 ESFO/60. Horcoff took 51.38 ESFO/60, despite having several games in which he took almost none, presumably because his shoulder was dinged up.

    Just a bit of context on this – amongst guys who played any degree of time, Sidney Crosby led the league last year with 58.39 ESFO/60. Horcoff was at 59.50 ESFO/60 in 2008-09, when MacTavish basically used him every time he could. That’s sort of the upper bound here. Last year represented a big drop for Cogliano from his first two seasons – he started out taking 31.09 ESFO/60 in 2007-08 and saw that jump to 38.01 in 2008-09.

    Last year’s figure of 19.45 ESFO/60 essentially means that Cogliano basically isn’t a centre anymore, at least for faceoffs. That rate puts him at 133rd in the group of forwards who played at least 500 ES minutes. He was fourth on the Oilers, more part of a group that includes Stortini, Brule (18.4) and Penner (14.6) than the group that includes Horcoff, Potulny (46.8) and Gagner (38.3). To a certain extent, he’s already been tested and found wanting. It’s a shame that the league doesn’t track faceoff data better – I have some suspicion that with his massive drop in faceoff rates, he probably faced more favourable competition which might explain the uptick in his success rates.

    Let’s assume though that last season was a momentary blip brought on by a desire to see what Ryan Potulny could do in the faceoff circle and that the plan is for Cogliano to be a 45.0 ESFO/60 guy going forward. Does his ineptitude in the circle matter? I would argue that, on any reasonable analysis, it doesn’t.

    If we assume that Cogliano is a 45.0 ESFO/60 guy, that he’ll play 82 games a season (as he always does) and that he’ll play 14 ES minutes a night, we’re talking about a player who takes 923 ES faceoffs annually. For some context, that would put him in 28th in the league last year in terms of faceoffs taken. It’s a lot.

    The difference between a fellow who wins 40% of his draws and one who wins 50% of his draws, even over 923 ES faceoffs, is pretty small – it’s 92 wins over the course of a season. While I can’t find anything off through a quick search that provides the goal value of a lost draw, this analysis from Gabe Desjardins suggests that in terms of shots, it’s pretty negligible when it happens in the neutral zone – if 1/3 of those lost draws were in the neutral zone, you’re probably talking about a swing of 2 shots against you. While I can’t quite extrapolate from Gabe’s numbers there as to the size of the swing from the (say) 2/3 of faceoffs that would be in the offensive or defensive zone, it looks like for every four faceoffs you lose, you give up an extra shot in the defensive zone and cost yourself an extra shot in the offensive zone, respectively, in the next thirty seconds. So, call it a negative swing of 15 shots, plus whatever additional shots against you give up when you lose an offensive zone faceoff and additional shots for you lose when you lose a defensive zone faceoff in the next thirty seconds, which I would expect to be minimal.

    In short, I’d have a hard time arguing that, even if he’s horrible on faceoffs, Cogliano is going to cost you much more than a goal or two of goal differential annually because of it. That has some value, to be sure, but it’s not going to be the difference between a player who you want to keep and one who you want to lose, particularly when this entire analysis has been premised on Cogliano getting a huge volume of faceoffs, despite being terrible at it, something that simply will not occur, because coaches aren’t stupid.

    None of this is to say that I see Cogliano as a guy who you want around in the long term. He isn’t really a guy who creates a lot of offence or one who is particularly good defensively. His announcement that he’d decided to become Shawn Horcoff struck me as a bit laughable, as I think that there’s an awful lot of learning that would have to take place before he could be Horcoff, if he ever could be. He’s a long term project. Unless someone can show me that I’m wrong about the value of a lost draw though, it seems to me that there are far better reasons to trash him and his value to the Oilers than his ineptitude in the dot.


    14 Responses to In defence of Andrew Cogliano

    1. R O
      September 29, 2010 at

      Solid. I’ve been arguing this for a while based on Gabe’s “15 seconds of 5on4″ conclusion + my own back-of-the-envelope calculations: basically, an entire team of 45% FO men probably won’t lose more than 3GD at EV.

    2. Triumph
      September 29, 2010 at

      while i don’t think the deficit is that great either (i’ve surmised 6 goals between 60% and 40%), you are leaving out future defensive-zone faceoffs and penalties taken as a result of faceoff losses. that probably adds up to an extra goal against for the very worst faceoff men.

    3. Saj
      September 29, 2010 at

      Lost draws can also cost you powerplays (both ones you would have had, in the offensive zone, as well as ones you did get, in the defensive zone), and the scoring rate on those is like 20%. But I don’t know how to quantify how many PPs are caused by lost draws.

      Regardless, I agree with your last paragraph, in that he is just not that useful in all aspects, not just faceoffs.

    4. RyanV
      September 29, 2010 at

      I ran a quick check from the playbyplay data, and it suggests that Cogliano wasn’t facing off against weaker faceoff men in 09-10 than the previous years. In fact, his “faceoff qualcomp” probably increased. Here are the numbers:

      07-08: 50.6%
      08-09: 50.6%
      09-10: 51.2%

      (I got these numbers by adding up the total number of faceoff wins and losses by the guys Cogliano took draws against multiplied by the number of draws he took against them, then taking the win%. This isn’t the best way to do it, but it’s the easiest.)

    5. B.C.B.
      September 30, 2010 at

      There is one thing about the even strength offensive zone draw that I do not think you have taken into in consideration. If you lose the draw in the offensive zone, it may only ‘cost’ you one shot on net, but it allows the opponents to get the puck out of their zone and make a line change. This is important if you are trying to employ a hard line matching strategy. That is all, have a good day.

    6. speeds
      September 30, 2010 at

      I have a fuzzy recollection that Tom Benjamin wrote about this a long time ago, that faceoffs are a stat given too much importance simply because they are counted.

      I’d provide a link if I could find it, if anyone else has it, and could post it, that would be great.

    7. Robert Cleave
      September 30, 2010 at

      Speeds, I believe this is the post you’re looking for.

    8. speeds
      September 30, 2010 at


      Thanks, I think that’s the one as well.

    9. Gerald
      September 30, 2010 at

      One of the rare times where Benjamin and I would agree on anything.

    10. Vic Ferrari
      September 30, 2010 at

      I think if Gabe broke down the neutral zone faceoffs further he’d find that the ones taken on the dot by the blue line, like just after an icing … it really doesn’t matter whether you win those or not.

      This was one of the findings of Andrew C. Thomas when he studied the Harvard mens varsity team. And since the rest of his data seems to mesh really well with NHL numbers, I assume that this is the case for the NHL.

      I suspect that if we look a the rookie year for Cogliano and Gagner, they took a disproportionate amount of their faceoffs after offsides. And Stoll, Reasoner and Horcoff disproportionately more in the ends of the rink.

      It may be a small gain, but a coach has to earn his keep. A bunch of small advantages like that may add up to an extra win or two on the season, I think.

    11. Vic Ferrari
      September 30, 2010 at

      Oops. In the first sentence that should read “like just after an offside”

    12. Adam D
      September 30, 2010 at

      I’ve often suspected that face-offs in general were blown out of proportion, and this seems to confirm that. This also is relevant in regards to the constant braying that we need “another C that can win face-offs!”.

    13. Saj
      September 30, 2010 at

      Has anyone ever done a regression of winning percentage against faceoffs and other variables like GF, GA? If the faceoff coefficient turns out to be significant, maybe that can quantify things in the W/L column a bit better than our guestimates here.

    14. October 3, 2010 at

      I would imagine that being a failure on the dot would have more of an impact at special teams, simply because a lost faceoff means either a lost opportunity to clear (PK) or a gained opportunity to clear for the opponent (PP), but even at that, how many GD is that worth, with the special teams minutes Cogliano’s likely to get? Surely less than five?

    Leave a Reply

    Your email address will not be published. Required fields are marked *