Anybody with half-a-brain can see the Oilers need a centre who can win a face-off and kill penalties to make up for the losses of Jarret Stoll and Marty Reasoner. When Horcoff is the Oilers best overall centre on the dot at 51.8 per cent but is just 61-89 shorthanded, it’s tough to argue.
“Maybe now this year, with them (face-offs) being in the zone so much,” Horcoff said when asked if poor performance on the dot is the single biggest factor that has the penalty killing struggling so mightily.
“Face-offs aren’t just one guy’s stat. I think people who know the game and watch the game realize the help is and how important those 50-50 battles are . . . especially on the PK.”
-Robin Brownlee, A Man Who Learns From Watching, OilersNation
When you look at the overall lack of positive correlation, especially in 2008/2009 now that the rules have changed, you can’t conclude that a good faceoff percentage = a good penalty kill. This might be common sense but based on what I’m reading around the internet today I think people are overestimating the impact of faceoffs on Edmonton’s pk train wreck this season. What I’m trying to say is that if one conclusion can be drawn it’s that there are too many elements to a penalty kill to suggest Edmonton should be looking at adding a faceoff specialist alone - faceoff % should take a back seat to players who can actually kill the damn thing off. After all, even a successful faceoff win and dump only shaves 10-15 seconds off the penalty anyway - you still have to kill the whole damn thing off regardless of how it begins.
-Showerhead, Statzi, Irreverent Oil Fans
I’ve been trying to dig into the PK problems over the holiday break - I’ve found some interesting stuff but I’m still in the process of sorting through it all. One thing that’s clear to me: the Oilers 2005-06 PK was an unholy machine in the second half of the season. We talk a lot here about things that have a lack of sustain: that Oilers team blocked an outrageous amount of shots on the PK in the second half (something I’ll touch on when I get around to writing that is how blocked shots should be measured; I tend to think that the sensible approach is blocked shots/shots allowed*60, rather than just a pure per 60 rate, which is what I usually use) but the number of shots that they gave up in the second half was also remarkably low at 33.5/60. Columbus and Nashville were allowing more shots/60 at evens (!) in the second half than the Oilers were when they were a man down; Washington was allowing the same volume.
That’s neither here nor there though. This post is going to be narrowly focused on the topic of faceoffs and penalty killing. Faceoffs were a topic of discussion here back in the summer of 2006; call it the internet version of fiddling while Kevin Lowe ran around with a jug of gasoline and an insane gleam in his eye. I had two posts up about the topic, one dealing with faceoff zones and the other with the advantage in putting your stick down second, which attracted some comments but, quite reasonably, it seemed that most people were spending the late summer doing other things. With the putrid PK this season, people are casting about for reasons why and the departure of Stoll and Reasoner provides a tantalizing storyline.
The NHL started to do something differently last year: they recorded the game state in the play-by-play files for faceoffs. Although they were tracking it internally beforehand, they didn’t publish it. They don’t compile these stats on a team level now, but as the information is in the play-by-play files, we can go through and put together the stats on a team level - all one has to do is turn the play-by-play files into a useable database, which is a Sisyphean task of its own.
In any event, I did that and something interesting shows up on the team level. Teams on the PP enjoyed a significant edge in 2007-08 when it came to faceoffs: they won 11 out of 20. That may not sound like a lot, but there wasn’t a single team in the NHL that managed to win 55% of its faceoffs over the course of a season, so it strikes me as a pretty significant edge. Here’s the chart:

Horcoff would seem to have a point. It was, in 2007-08 at least, just more difficult to win faceoffs shorthanded than it was to win them generally. This is something that Vic Ferrari has mused about in the past, suggesting that the team on the PP is at an advantage because it has more guys who can come in to help out when a draw isn’t cleanly won.
There’s the odd team that did better on SH faceoffs than it did on PP faceoffs; that’s to be expected, I would think, when you consider that a subset of the team takes the bulk of the special teams faceoffs. For example, Horcoff has taken 49.5% of the Oilers PK faceoffs this year and Brodziak has taken 33.6% of them. On the PP, Horcoff has taken 54.4% of 285 faceoffs (winning 59.4% of them), while Brodziak has taken just 3.5%. Gagner is second, with 16.8% of the PP draws. The change in personnel is going to affect the team faceoff success rate.
In fairness to Robin, I note that there was some moderate degree of correlation between PK faceoff success rate and PK% last year, at .449. The difficulty that I have, as I’ll set out below, is that the Oilers PK has absolutely cratered this year. I’m not opposed to the argument that their faceoff woes are part of the problem but I have a harder time believing that it’s the most significant problem facing the PK unit. I think that Showerhead has it when he writes “…there are too many elements to a penalty kill to suggest Edmonton should be looking at adding a faceoff specialist alone - faceoff % should take a back seat to players who can actually kill the damn thing off. After all, even a successful faceoff win and dump only shaves 10-15 seconds off the penalty anyway - you still have to kill the whole damn thing off regardless of how it begins.”
You’ll note that the Oilers won 47.6% of their shorthanded faceoffs in 2007-08, good for 6th in the league. The current edition of the club has won 126 of 303 faceoffs while shorthanded this year, good for a 41.6% success rate. The Oilers took 8.03 PK faceoffs/game last year; this year they’ve taken 8.4 PK faceoffs game. If the Oilers were rolling along at last year’s rate, they’d have won 144.2 faceoffs so far this year instead of 126. So what we’re talking about here is an extra lost faceoff on the PK every second game this year.
To tie it all back to the crack at Brownlee that I opened this with, through thirty games, the Oilers were giving up an extra 10.6 PPSA/60 over their rate from last year. At last year’s rate of 49.9 PPSA/60, they would have allowed 182 PPSA through 30 games instead of 221. Does it really seem plausible that an extra 15 lost faceoffs (through 30 games at a rate of one extra loss per two games; some of which wouldn’t have been in the Oilers end anyway) is responsible for an extra 39 shots against? The logic doesn’t seem to be there to me; it might be a contributing factor but I just can’t see it being the most significant problem.
The failure of the current PK is probably due to a lot of different factors but it’s hard to overlook the drop in save percentage. Last year, the Oilers posted an .892 on the PK; they’re currently sitting at an .863. Over the course of 265 PK shots, there’s 7.7 goals on the difference in save percentage alone. Considering that the Oilers are currently about 14 goals off of last year’s pace on the PK, more than half of that is probably tied to the dip in PK. Whether that falls on the goaltending or the quality of chances given up, it strikes me that any argument that the inability to win faceoffs is the key factor in this drop needs to somehow link 18 lost faceoffs with a massive drop in save percentage. I have an awfully hard time seeing how that can be done and so, at present, I see the faceoff problems as likely being, at best, a tertiary factor in the general decline of the PK, somewhere behind the bounces and not stopping as many pucks as they did last year.
Just curious… have you seperated own zone PK draws this year and last year?
I mean it’s one thing to lose the PK draw in the neutral zone, it’s another to lose it in your own end.
I’m not sure of the value in terms of goals against, but with more obvious in zone face-offs than ever before that may have an impact on the end results.
But I do agree with the general point… there is a lot more wrong with the Oilers PK than just the face-off %. It’s not that they are giving up goals right off the draw, but they are giving up goals off the rush, off a dump in, virtually any way you can think of to get scored on the PP they are.
Obviously guys lose more draws 5 on 4. Even more 5 on 3… the fact Brownlee almost called Horcoff out on that is just plain dumb. 5 on 4, guys will be cheating more towards tying it up, while the PK’er has to win clean.
Winning 5 on 3, as the three, is damn near impossible. Given Horcoff has taken nearly everyone of those this season, I’m sure that has impacted those numbers even more…
Garon’s strength in his game is his lateral movement and ability to take away the bottom half of the net. This is arguably Roli’s weakness. Nowhere is this need more prevalent than on the PP. could that account for some of the sv% drop. You said “2005-06 PK was an unholy machine” and I don’t know what the sv% was when Roli was in net in 05/06 so that might be not the case but the old adage “your best penalty killer is your goalie” probably holds some weight.
Glad the evidence shows that a face off guy isn’t paramount to fix the PP. I’ve been fighting off the “Anybody with half-a-brain can see the Oilers need a centre who can win a face-off and kill penalties crowd” for some time.
I kind of hit it obliquely but Garon is getting torched on the PK this year…save percentage of like .826. Small sample size and all that but, if you believe that goalies don’t impact on the shots all that much, he’s the one that’s getting killed, not Roli.
It’s what guys ask me about what I consider to be a scoring chance or not but I’d really like to get into the quality of shots that the Oilers are giving up on the PK.
Just going from recent memory we know that Iginla and Cammallari had all day from great positions to score on NYE and we also know that Brian Lee had a great look for the GW last Tues. I doubt any goalie’s gonna make consistent saves on those looks and here’s where I think we can find the problem.
The Oilers PK of my youth was pressuring down-ice and getting down to block shots and Stoll in particular was a real dandy when it came to having his stick in primo passing lanes.
But right now Everything is wrong.
I agree Dennis. And I’m not even sure if this team is unwilling to block shots because they aren’t even in the right areas to do it on a lot of them.
I’m making this comment to my hockey blog from my blackberry in a starbucks in downtown toronto. I need to pick up my skates from the office because I’m on the ice in an hour. I have been to the mountaintop and seen the Canadian yuppie dream.
Re db’s point about shotblocking: I tend to agree. My sense this year is that the Oilers haven’t really been in position to block a ton of shots. The screened goals against CHI a few games back were noteworthy in part because they weren’t completely open looks.
I think you’re wrong Dennis.
PK is such a small sample size that things go sideways and rightways in a hurry. Your own numbers will confirm that when we get people to separate the 5v3 from 5v4 chances, Dennis. Bet real money on that.
You had me convinced in 03/04 as well, when they were giving up one against a game, and rarely more. I ran numbers to make you point, and ended up showing that all PKs are as erratic as fuck. You can check for yourself, and you can use rollong averages (my fave) and play that against dice. I’m telling you Dennis, there is nothing there.
There are things that folks think are regular that spread way outside the norm for things that nobody but RiversQ will peg. And Cuban has deleted his old posts anyways. But hockey is hockey.
At the risk of plagiarizing Al Arbour: If you have enough good hockey players, eventually you’ll win more than your share of games. Otherwise not.
Tyler, great post. The save percentage numbers are ones I’ve been curious about but never dug up and the personnel change + looking only at SH faceoff percentages makes your work quite a bit more accurate than mine. Also, I’ve noticed some pretty dry and hilarious one liners left in comments around the Oilogosphere over the break (I especially enjoy the yuppie dream line) so you’re in fine form right now.
I’ve made some very theoretical ramblings in the comments in the most recent two IOF posts - I’d be very curious to read your take on them as well as the people they were originally pointed at.
Something that was said recently by Moreau when asked about the PK woes, tips us to where the team has been failing IMO.
18 said they were goint to start things over and keep it simple. As much as a cliche as that sounds, its pretty clear that the guys were losing their assignments when faced with rotation and switching that comes with more intricate defensive strategies.
Last night on the Pkill, I swear the oilers defensemen were tethered to their respective posts with a bungee cord as soon as DAL had the puck. There was minimal pursuit within the defensive zone. It gave up alot of posession time, but it limited mistakes by sticking to a zone scheme. Much like basketball, zone defense is vulnerable but simple and effective.
I’ve always thought/commented that faceoff wins were largely overblown on special teams when we consider the range of +/- 5% and the ensuing effect. I suspect however it is more of an achilles to a PK to lose the faceoff, as it accomplishes entry into the zone and posession. Two things the attacking team must accomplish seperately if they lose the draw… whats the percentage of entry attempts resulting in offensive posession for powerplay teams? I have no idea obviously but I’d guess 50-60% maybe?
If you lose the pkill faceoff, you surrender possession immediately, whereas a win gets you 15-20 secs immediately, in addition to subsequent entry attempt time that you can kill…
A great stat would be to have available the % of time a team spends in each zone during PKills.
Losing 1 extra faceoff per 2 games is, by itself, pretty insignificant however to the grand scheme of pkill success.
I cringe at the sv% argument a little as well, though because when somethings broken in front of the goalie their stats will falter accordingly.
call it the internet version of fiddling while Kevin Lowe ran around with a jug of gasoline and an insane gleam in his eye.
Just spewed coffee into my monitor and will probably be fired now. Thanks.