I didn’t watch this game as closely as I usually do as we had people over. Two things:
*Other people have already beat up on Darcy Hordichuk for his thoroughly useless presence in the lineup. I was intrigued by something that Kelly Hrudey said in the broadcast after the whole asinine “Hordichuk pointlessly runs Dennis Wideman into the net, takes penalty” sequence:
(Hordichuk) is normally very dependable when he’s inserted in the lineup. He rarely does take minor penalties so I’m sure that’s one that he’s going to remember, that you can’t take those penalties when you’ve just been inserted in the lineup.
The thing about that…it’s totally not true. Gabe Desjardins’ tracks penalties taken per sixty minutes of TOI. There’s no explicit explanation on the website but I went through and did some checking and it looks to me like Gabe filters out misconduct penalties and coincidental penalties. Hordichuk was 15th in the NHL in this category last year, which is awful. He was ninth the year before, which is worse.
The interesting thing, for me anyway, is that this is a relatively new problem for Hordichuk, which is presumably why Hrudey had this idea. In the 2008-09 and 2009-10 seasons, he’s not a particularly notable taker of penalties – a bit high, but nothing outrageous. In 2007-08, he was notable but for a completely different reason: he took 60 PIM in 232 minutes of ice time but not a single one of those put his team shorthanded: he had a twelve fights. Somewhere along the line, he’s evolved into a guy who takes penalties.
My back of the envelope math based on the data at BTN suggests that NHL forwards averaged .7 penalties per 60 minutes of 5v5 play. Hordichuk took 12 penalties in his 187 minutes, or 10 more penalties than the average forward would have taken in that time. If you assume that each penalty costs you 0.2 goal difference, the cost of Hordichuk’s penalties is two goals, or 2/3 of a point in the standings. Hordichuk drew a ton of penalties last year; he more than wipes out his cost in PIM, actually – but you kind of wonder what the point of having him in the lineup is, when he takes a lot of penalties and barely plays. Even if you think that his penalties drawn should be expected to wipe out his penalties taken, what’s the point of having a guy in the lineup who you don’t trust to play? For all Hordichuk’s talk about protection and hitting people, he’s had a hell of a time in his career finding a coach who will put him on the ice to do that.
*The Oilers have played 240 minutes of regulation time so far this year. Of that, they’ve trailed for 180.83 minutes, been tied for 59.12 minutes and led for…carry the 6…zero minutes. More than half of the time that they’ve been tied was in the Los Angeles game, which was marred by a never-ending sequence of penalties that saw 16 of the 33 minutes before the Kings scored played on special teams.
In other words, we’re a week into the season and we still haven’t had a chance to see Krueger would run his bench at ES for an extended period in a tie game or with a lead to protect. The Oilers have played two games that were close almost all of the way through, with LA and Vancouver – and two games that had a lot of 2+ goal deficit time, with San Jose and Calgary. It’s interesting how the ice time of Ryan Smyth and Shawn Horcoff has dipped noticeably in the two games that were blowouts, although both saw even less ice in Calgary than they did in the games in which the Oilers were close – trailing by two in the third period isn’t really close.
Krueger’s forward rotation in the third was pretty interesting. I’ve kind of annotated it here:
20-37-91 (Note that he was fine with coming back with MPS on the third shift but that he left Smyth/Horcoff on the bench)
10-20-94 (I had guessed that this was a defensive zone faceoff; in fact, it appears that this was due to Gagner/Hemsky staying out for 1:42 on their preceding shift. Horcoff went to the bench almost immediately, playing just 26 seconds. He was replaced by Gagner and presumably the rest of that line would have followed. This was negated by Belanger taking a penalty.)
10-94 (7:40, 4v5) (Horcoff again went to the bench almost immediately. The Oilers have a demonstrated preference for Belanger and Horcoff taking SH faceoffs this year – they’ve taken 36 between them, with the rest of the Oilers combining for eight. One wonders if, when one of them is in the box, Krueger intends to use the other as exclusively for faceoffs as possible, having him come to the bench as soon as the zone is cleared. Belanger played the entire two minutes of 5v3 when Horcoff was in the penalty box against LA, aided by a timeout in the middle of the penalty kill.)
93-14 (Iginla penalty at 9:05)
20-37-91 (First shift after PP: out come all of the forwards who didn’t play on the PP, except for Hordichuk and, curiously, Smyth.)
20-94-91 (As I’ll explain below, this is the “Where’s Horcoff?” shift that sort of puzzles me.)
20-94-91 (Horcoff’s absence on this shift is a bit more understandable than his absence is from the preceding 20-94-91 shift. As you’ll see, he was about to come out and if there’s a marginal difference between him and Belanger in the offensive zone, I’d rather have Horcoff fresh when he comes out to face the other team than having just played a shift.)
20-89-64-83 (Belanger’s preceding shift was short and he was left out to take the faceoff to Kiprusoff’s right. Looking at the ES faceoff numbers for Belanger and Horcoff (SH faceoffs are harder to win than ES faceoffs which are harder to win than PP faceoffs, so you need to look at them by game state), Belanger was probably the right choice. Over the last three years, he’s a 56.4% guy to Horcoff’s 49.3% at ES. Chase every edge, I guess. Belanger won the draw.)
89-93-4-14-64 (Krueger let Gagner take two faceoffs in the Calgary zone here with 1:09 and 0:58 remaining. Gagner lost them both and Calgary iced the puck each time. I initially thought this was strange, having him take them but, looking at the numbers, Gagner improved significantly at ES faceoffs last year and this was effectively a PP faceoff. In that light, this is a less curious decision.)
20-89-94-4-14 (Having seen enough from Gagner, Krueger sent Belanger out to take the faceoff with 0:50 remaining. Belanger won it.)
One thing at least that we can take from this: Belanger is probably going to be the guy on important faceoffs this year, with Horcoff in a secondary role and Gagner possibly allowed to mostly take his own draws until Krueger loses patience with him. Belanger, of course, won the faceoff with ten seconds left against LA that led to Yakupov’s goal.
The thing that’s trickier to tease out of this is why Horcoff missed some time on Saturday night. I mentioned the “Where’s Horcoff” shift and I can’t figure out why he wouldn’t have had a shot in the rotation there. My best guess is that Krueger is looking to reduce the minutes for Horcoff and Smyth this year by transferring some of the “lower leverage” minutes, when a goal against is less critical (a two goal deficit on the road meets this description) away from Horcoff and Smyth and towards Petrell and Belanger. Even if you think that there’s a marginal offensive difference between Belanger and Horcoff, I think it’s more difficult to argue that there’s a marginal offensive difference between Petrell and Smyth and Petrell was taking shifts that Smyth could have taken.
Both Smyth and Horcoff have played a lot in the two games that the OIlers have played that weren’t blowouts. They’ve played markedly less in the games that were blowouts. If you look at this game, garbage time essentially started at 14:57 of the second, when the Flames restored their two goal lead. Assume, for the sake of discussion, that Edmonton was a team that, against the Flames over a million games, would average 2.5 GF and 2 GA. They probably aren’t that good in the real world – the betting market says Calgary’s better than Edmonton when the game’s played in Calgary. We’ll pretend score effects don’t exist either.
In that world, Edmonton wins or takes the game to OT ~22% of the time. Then you have to ask yourself how upping Belanger’s ice time at the expense of Horcoff affects those odds – the answer is probably “Not by much”; in the world where the Oilers get a point out of the game, it’s going to be the RNH or Gagner lines that are most likely to score.
If you look at the game up until Calgary made it 4-2, Horcoff had played 9:20 to Belanger’s 6:32. From that point forward, Belanger played 5:36 to Horcoff’s 2:58. The best, most reasonable guess, seems to me that Krueger’s identified garbage time as a way to remove TOI from Horcoff.
The other thing that’s striking about this that I guess I knew, to an extent, is just how exposed you are if you have two lines that you can reasonably expect to score and you’re down on the road. The other coach can simply hold his preferred line back to face yours – if you want to run your third and fourth lines out there to burn time, waiting for him to give you a matchup you can live with, so be it. There’s an interesting idea for a study here, I think, about just how much value last change has by looking at games in which a team is leading by X goals, home and away, controlling for time since previous game and seeing the differences in winning percentage. That’s probably a pretty good measure of what value line matching (and, by extension, home ice advantage) gives you.Email Tyler Dellow at firstname.lastname@example.org