Shawn Horcoff, You Lucky/Unlucky Dog

Shawn Horcoff’s ugly minus-29 rating last season did not go over well with fans of the Edmonton Oilers. The veteran centre, inked to a long-term big-dollars pact, was in the richest year of that deal. The expectation was that he provide both offence and defence, but in a season where his point totals fell off and his plus/minus wandered into the sewer he became an easy target on a miserable team.

This season, things are different. Horcoff was named captain in the summer, and while the team is still struggling on the ice there is hope off of it. Despite injury, Horcoff’s offence has come around and he has a plus-3 rating on a team that has allowed 46 more goals than it has scored.

Some of that – specifically, Horcoff’s plus/minus – is illusory.

Let’s start by looking at shots for and against while Horcoff is on the ice over the past two seasons. Since plus/minus is primarily a measure of even-strength play, we’ll use Gabriel Desjardins’ five-on-five data.

Player SF/60 SA/60 Shots +/-
Shawn Horcoff, 09-10 24.9 28 -3.1
Shawn Horcoff, 10-11 24.1 30.3 -6.2

It’s interesting to note that despite the improved plus/minus, the shot differential while Horcoff is on the ice is nearly twice as bad this season as it was last season. Why then has his plus/minus improved? Simple answer: on-ice percentages.

Player SF/60 SA/60 On SH% On SV%
Shawn Horcoff, 09-10 24.9 28 0.0682 0.891
Shawn Horcoff, 10-11 24.1 30.3 0.0958 0.927

Looking at the table, we see that the Oilers’ shooting percentage when Shawn Horcoff is on the ice has risen, from 6.8% last season to 9.6% this season. Meanwhile, the team’s save percentage with Horcoff on the ice has also improved, going from a miserable 0.891 to a stellar 0.927.

But while those numbers sound significant, it’s hard to put them in real terms. So let’s take Horcoff’s 14 and change minutes of even-strength ice-time and project it over an 82-game season, with the current percentages. And to make the comparison easier, we’ll go back and do the same thing for last season. Now, despite the fact that the Oilers allow more shots this season with Horcoff on the ice, we see this difference in plus/minus:

Player SF/60 SA/60 On SH% On SV% Projected +/-
Horcoff, 09-10 24.9 28 0.0682 0.891 -26
Horcoff, 10-11 24.1 30.3 0.0958 0.927 2

That’s a tremendous gap – a +28 shift, based entirely on percentage changes.

Vic Ferrari has persuasively argued that defencemen have very little impact on the save percentage when they’re on the ice; it seems likely that forwards have even less of an impact. We know forwards can impact their on-ice shooting percentage – better passes, better shooting, etc. – to a degree, but I expect what we’re seeing here is the impact of Horcoff having Hall and Eberle as wingers rather than Patrick O’Sullivan and Jean-Francois Jacques.

We know, however, that over time on-ice save percentage + on-ice shooting percentage tends to even out to the 100 range (e.g. an 0.910 SV% and 9.0 SH%). If one were to make the argument that Horcoff is a significantly above or below average offensive player we might expect that shooting percentage to be above or below average; personally I’d suggest reality has it somewhere near the league average mark.

What would Horcoff’s plus/minus look like with his on-ice percentages normalized near the league averages? As follows:

Player SF/60 SA/60 On SH% On SV% Projected +/- Normalized
Horcoff, 09-10 24.9 28 0.0682 0.891 -26 -5
Horcoff, 10-11 24.1 30.3 0.0958 0.927 2 -11

This strikes me as closer to Horcoff’s true level of performance. His plus/minus last year overstated how ineffective he was, thanks to the combination of miserable linemates and poor goaltending behind him; this season it’s probably better than it deserves to be thanks to linemates with strong shooting ability and some good luck with the goaltending.

Whether readers agree with that analysis or not, from 2008-10 Shawn Horcoff has been a textbook example of how on-ice shooting and save percentage can redeem or decimate a player’s plus/minus.

  • Gotta get this in quick to avoid . . . you know.

    Plus minus is a pretty useless stat all things considered. I’ve always been an unabashed Horc supporter. He’s never as bad as the haters say and he’ll probably never be as good as what the supporters say.

    I think he’s a pretty good hockey player.

    *cue the people bashing Horc for getting paid too much money as opposed to K-Lowe for giving what he did when he did.

  • Hall for President

    I guess Horc’s contract value will never be recoverable…

    The contract was a bad decision, no point continually revisiting it over and over again

  • Quicksilver ballet

    If Daryl bought him out after the 2012-2013 season, the Oilers may benifit from having that 5.5 cap hit trimmed a little by then, he’s due only 7’ish over the last two years of that deal. What would that do to the cap hit over the last two years of his deal after that 4.6 mill buyout?

  • Romanus

    Didnt finish reading the whole article..I dunno, I just think Horcoff along with almost every other Oilers had a terrible season last year. Deep down in my mind I keep on getting more and more pissed off every time someone mentions that Omark might get traded because the team might get something for him and he;s small..
    Hello..Penner plays smaller than Omark..Omark plays bigger than most guys over 6 feet. Keep the kid, hes a future Martin St.Louis.

    I know Willis didnt mention Omark, but I just finished reading R.Cloutier from Hockeybuzz..what the hell is that guy smoking..

  • Just finished watching the all star draft and have two thoughts:

    1: team Lindstrom = way better than team Staal
    2: Ovechkin taking pictures of Kessel sitting all by his lonesome was one of the funniest things I’ve seen in a while.

    Now I’m out of thoughts.

    • Listening to Kypreos and Millard talk about it afterwards is sad. They were going on about how emberrassing it was for Kessel and how strange it was to see a Toronto player be last.

      1) It takes a lot more than this draft of All-Stars to make me feel bad for Millionaires.

      2) I hope the player selected last for the next 10 years hails from Toronto. In fact, that should be a mandate of the new CBA.

  • Oilers4ever

    It’s not Horc’s fault the KLowe was a dumb arse and gave him too much money… if someone whipped that in my face for my job I’d say yes too regardless whether I think I am worth it or not.. Stupid is as stupid does.. or however that saying goes…

    And agreed.. those stats hurt the brain way too much… you guys and all yer damn stats, CORSI numbers.. blah blah blah.. what matters is what you see on the ice… and Horc’s having a good year so far minus the injury.

    Lay off the guy already.. like YOU could do any better.. obviously not.. or you’d be out there playing for the same big bag of cash…

  • O.C.

    I detest plus minus.

    The player who took a penalty that the other team scored on should get a minus 5 by the logic of how it is currently calculated.

    And the player who drew a penalty should get the plus five if his team scores on the PP.

    Players shouldn’t get minuses for empty net goals.

    Bloody plus minus sucks. As do shots. We should focus on legit scoring chances. Do you know if you hit the post that isn’t a shot on goal? I guess bloody shots on goal stats suck too.

    Short answer. Horcoff having a better year. We don’t need to see the plus minus.

  • I kind of feel bad for you Jonathan. If you posted this article over at C&B, you’d have 50-60 comments already, the majority being on topic.

    As far as it goes, I can’t help but wonder if your final assertion might be questionable in that you’re attempting to normalize to league averages. That would be fine if Horcoff had anywhere near league average linemates the past couple of seasons. We all know what happened last year, and this year he’s been carrying two high powered rookies with limited defensive acumen. On top of which, the goalies he had to work with last year were far from league average (so far below – *sigh*).

    Basically what I’m trying to say is that it appears to me you’re using extreme anomaly performances for comparators to league baselines. I don’t think that’ll give a realistic analysis because those anomalies are so far off the mean that equivalency may not be possible. The results are just too far off the grid. In other words, the results from the recent past weren’t bad on a linear scale, perhaps more on a geometric scale instead (they weren’t just bad, they were bad squared).

      • So you’re saying the quality of player has nothing to do with his team’s quality of chances for and against? That’s like saying there’s no such thing as good and bad players. An extra step or two around a defenceman gives you a much better shot. A forward not trying to poke the puck away gives you a better shot.

        Horcoff has looked a lot better this year than last year, and that’s evidenced in the shooting percentage gap which, considering that there are 11 other players on the ice at all times, is pretty amazing.

  • Horcsky

    Two things about this year suggest Horcoff is playing effectively.

    1. The team looks lost without him. Also, their record indicates as much.

    2. His line actually spends time in the offensive zone. When both Horcoff and Hemsky were out, our time on attack must have been ridiculously low.

    • And for what? If I’m not mistaken wed be responsibly for what 1/2 the cap hit over the next… 4-6 years…. I doubt we are going to pick up a free agent that has horcoffs worth for half his cap hit. It’s not like we need the extra cap space at this moment.

  • Kodiak

    I know there are other factors involved, but I’d like to see the teams win/loss record with Horcoff in/out of the lineup as opposed to Hemsky in/out of the lineup. I think the Horcoff haters and Hemsky lovers might be a bit surprised by the findings.

  • @ MattL:

    I’m saying that there’s absolutely no repeatability in on-ice save percentage results. Horcoff’s 0.891 last year wasn’t his fault, any more than his 0.927 this year shows a new commitment to offence.

    If there’s no repeatability – and the data shows this – than there’s absolutely no sense in blaming or creditting a player for his on-ice save percentage.

    On-ice shooting percentage is a different story, as I acknowledge in the article.

    • To be semi-fair, the story you linked (and the one you wrote) was only a 2 year sample, so I think it might be jumping the gun to say the data shows anything.

      So do you think Horcoff’s been better this year, but that’s not why his stats are different in this case? Or do we just think he’s having a better year because we’re fooled by stats?

  • 9 Inches Uncut

    Sorry. Horcoff was terrible last year and a key contributor to their last place finish.

    He’s been much better this year but also missed a chunk of time with the knee injury.

  • Champ

    While I agree that extreme shooting percentages tend to normalize over time, there are cases where you have to look at how a player plays too. We were all waiting for POS’s sh % to improve but it never happened. It was easy to tell why just by watching him play a couple of games. He shot from everywhere! Any crappy angle shot, any 0.0001% chance shot he could take, he would. That makes his line’s SF/60 look a lot better but in reality, they did not spend a lot of time in the opposing zone compared to their own.
    Your assumption was that that line’s +/- was off and should be adjusted to reflect SF and SA/60. I think that while most of the time +/- is way off, and Corsi’s are a much better reflection of 2-way play, this is one situation where the +/- was probably more indicative of that line’s play.

    Advanced stats are great, but I think they work best when they’re combined with what you have witnessed on the ice. And I’ve seen a better Horcoff this year.

  • Horcoff’s points and play have very little to do with Horcoff and a lot more to do with people he plays with.

    I for one consider Horcoff to be a PR(ON veterans know what that stands for) type a player.

    Just take a look at the stats above.

    Last year Horcoff was in the minus for SF/SA, no suprise. The reason Horcoff’s SF/SA have decreased this year is the rookies he is playing with. The rookies are not as responsible in their own zone leading to more shots on Horcoff’s goalie. The rookies are a lot smaller and less experienced than NHL defensemen, therefore they don’t generate as much shots as their opposition.

    The reason his plus/minus has improved is because Horcoff is PRing two rookies with a lot of skill and potential. Eberle and Hall are a lot better than the pairings of players Horcoff went through last year. Horcoff don’t carry, he gets carried. OK! OK! Don’t come back with the “Horcoff is very responsible in his zone and carries his linemates for most part blah blah blah” I don’t see it.

    Same is to be said for shot percentage, you can do the math.

    The only way Horcoff is a textbook example is if, you put in him the textbook, force someone to read it and they chose to believe it.
    Horcoff will always be as good as the the people he plays with, and sometimes not as good. Horcoff him self is not much of a difference maker or a game changer, he is Horcoff.

    • OB1 Team Yakopov - F.S.T.N.F

      Thanks for the laughs man!

      Sidney Crosby’s wingers consistantly struggle to put up 50+ points yet Horcoffs #’s are propped up by a couple of rookies that may not hit 50 themselves.

      A true PR of an opinion!

      • No problem OB. I knew you’d have something wise to say. But you forgot son, I am wise you are otherwise.

        Crosby’s wingers rarely get as much PP time as Crosby does. I really dont’ believe your “consistantly strruggle” idea. Show me some math.

        Why do you always come with something to do with Crosby when someone says something bad about Horcoff? Horcoff is not Crosby, he isn’t near Crosby. So why?

        And yes, I would say his #’s are propped by the rookies, seeing as they both have more points then him.

        Hall will get 50+ points, Eberle’s injury may prevent him from doing so.

        • OB1 Team Yakopov - F.S.T.N.F

          Show you some math? Go through the Pens scoring pages for the last 5 years, he rarely has a winger putting up 50+.

          I bring up Crosby because he is the #1 player in the league and if he isn’t pumping up his wingers #’s then you can be damn sure Eberle/Hall aren’t doing it for Horcoff.

          • OK, genius. Riddle me this. If your options are Eberle, Hall and Horcoff. What two would you chose to be on your team?

            I tell you again, Crosby is an elite player is his own right and class. Horcoff is no where near Crosby. Understood?

            Do you think Horcoff would have the same numbers this year, if he played with Jacques and Stortini all year? Would Jacques and Stortini’s numbers be better than playing with Fraser?

          • OB1 Team Yakopov - F.S.T.N.F

            “OK, genius. Riddle me this. If your options are Eberle, Hall and Horcoff. What two would you chose to be on your team?”

            Irrelavant.

            “I tell you again, Crosby is an elite player is his own right and class. Horcoff is no where near Crosby. Understood?”

            Exactly! However, unfortunatly for you and your theory, Eberle/Hall are no where near Crosby’s class either….. so it seems pretty weak to suggest that those two are doing something Crosby hasn’t been.

          • OB1, please do not reply to my comments any more. Read them and forget them or just ignore them.

            I do not respect you half @$$ed opinions or your PR points of view.
            Keep coming here to PR and prop things you couldn’t think of your self, but do not reply to me.

            In my time here I have not read anything of value or originallity from you.

            Leave me alone with your narrow minded ignorant comments.

            Thank You

            Sincerely Kip

          • OB1 Team Yakopov - F.S.T.N.F

            That’s a strange responce? You thik Horc has his numbers propped up by rookies and I don’t.

            Your position has been mentioned countless times by countless people over the last couple of years.

            I can show you that the best player in the world doesn’t seem to be able to prop up his linemates numbers to support my position, something that I’ve never seen anyone else bring up before me, while you really can’t support your position at all.

            That makes it kind of strange then that I would be the ignorant, narrow minded and unoriginal one.

          • You’re d**b! You should think before you write, and then, read what you wrote, and then, think it through.

            How many rookies has Horcoff been playing with in the last couple years? Since people have mentioned it so much, tell me who these rookies are.

            You can’t show me, I asked you to and you didn’t. So, quit lying! You didn’t bring nothing up. What are you taking about?

            Since you know so much, tell me. What % of Crosby’s #’s come on the PP?

            You had nothing to say about my first comment, just something about Crosby. Tell me what was so original about that? You’re a PR, always have been and always will be.

            Also, stop going back to edit and add things to your comments after you had time to read them and realize how stupid they sound or after someone told you something you didn’t know.

          • OB1 Team Yakopov - F.S.T.N.F

            “How many rookies has Horcoff been playing with in the last couple years? Since people have mentioned it so much, tell me who these rookies are.”

            People have been saying Horc’s #’s are because of his linemates for years. First he was only productive because of Hemsky, now he is only productive because of Hall/Eberle.

          • OB1 Team Yakopov - F.S.T.N.F

            I already told you what I have to prove otherwise. Crosby doesn’t pump his teamates #’s, so I highly doubt Hemsky or Hall or Eberle do.

            Do you think Hemsky, Hall, Eberle can boost players numbers better then Crosby?

  • JW

    Good article. One quick note, because you’re doing so much writing… “off of” doesn’t happen in English.

    But, much more to the point, I agree with pretty much everything except a couple of sentences:

    We know, however, that over time on-ice save percentage + on-ice shooting percentage tends to even out to the 100 range (e.g. an 0.910 SV% and 9.0 SH%).

    So a player with an historical 14% shooting percentage can only expect .894 goaltending behind him?

  • @ spOILer:

    No, of course the two aren’t connected like that.

    For our hypothetical “average” NHL player, we’ll see him regress to the league mean. For individuals, we expect to see them regress to their goaltender’s EV SV% and their SH% to the skill level of themselves and their linemates.

    That was offered up only as an example, if we assumed Horcoff to have a league average goalie behind him and league average linemates, while being a league-average offensive player.

  • OB1 Team Yakopov - F.S.T.N.F

    “You can’t show me, I asked you to and you didn’t. So, quit lying! You didn’t bring nothing up. What are you taking about?”

    Well Kip, I trust that you can go to NHL.com and search Pitsburg scoring stats 06/07/08/09/10/11 ect on your own, I really shouldn’t have to physically show you the numbers.