There is a lot of what you say, as JF Jacques, Ethan Moreau and Patrick O’Sullivan have all spent a ton of time with Horcoff, and they’re not exactly Shutt and Lafleur.

Or even Penner and Hemsky.

The more troubling issue with Horcoff is his ongoing injury issues. Can he get over them?

He hasn’t been creating as many goals this year, and he’s been a culprit on more goals against, and it’s not just because of crap linemates (though that plays some part). By my eye, he lacks agility and flexibility, almost certainly due to his back and shoulder problems. At times, when he was flying in the past, he reminded me of Boris Mikhailov. I’ve seen almost none of that class and elan in his game this year.

This, of course, is the problem with signing older players. The injuries and hard miles start to add up, and a player looks like an old broken down hockey player in the blink of an eye.

P.S. Great work on this post, by the way, and I found it particularly interesting because it strongly makes a point I’ve been trying to make about Corsi plus/minus for some time.

Corsi is a useful stat when you break it down in this way, as opposed to just saying Horc’s Corsi is .465. That’s my main issue with Corsi plus/minus is used to defend or criticize players — that it’s earned as a team, not as an individual player, and to give credit or blame one guy for “his” overall Corsi plus/minus is to be less than fair and accurate.

It’s not “his” number at all. It’s the team’s number, earned by five guys, when this one guy is on the ice, and that’s a crucial distinction, as this post so ably demonstrates.

Are JFJ, Moreau and POS’s stats better or worse without Horcoff? Seems you only looked at the effect of those guys playing with Horcoff and not the other way around.

So Horcoff’s results are less terrible when playing with Penner? How counter intuitive. Can we see Penner’s results with and without Horcoff, which might tell us who is driving that line?

I like Horcoff, but blaming his troubles this year purely on Moreau, Jacques and bad goaltending is getting close to putting lipstick on a pig. I hope he gets back to his outscoring ways, but this season has been poor.

@Scott.

Do you really think a “goal” is a team stat in the same way that a plus or minus mark on a Corsi is a team stat?

Look at this way, for a moment, if you would . . .

When a player scores a goal, or an assist, it’s almost 90 per cent certain he earned that point, that plus mark. There are very few false positives or negatives, very few times that the player wasn’t essentially involved in the scoring play, when he is handed a goal or an assist, though it does happen. For instance, it might go in off his foot by accident, or he might make some mundane pass in his own zone that ends up as an assist on a great end-to-end rush. Usually, though, if he gets a point, he’s been crucial to the scoring play, individually involved in it, not a bystander.

When a player gets a plus or minus for Corsi, though, at least 30 per cent of the time, and maybe as much as 50 per cent of the time, he had nothing to do with the play.
Nonetheless, that player is awarded a negative or a positive on every shot at net. So a player’s Corsi plus/minus number is utterly riddled with false negatives and false positives. For me, that raises questions about it’s fairness and accuracy when used as an individual stat.

For instance, the player may be in the offensive zone, but did he do anything to get his team there? Maybe he just had the good fortune to be playing with Penner or Hemsky.

Or that same player may be stuck in the defensive zone, but maybe it’s not his fault, maybe it’s because he was playing with Moreau and Jacques.

So when we look at a goal total, we know that player actually had a hand in almost all of those goals, that he was individually involved in them.

But when we look at a Corsi plus/minus, we don’t know if that player was individually involved in those plays, or if he’s been assigned that mark because of the activity of a teammate.

So I trust goals and assists as individual stats, but I’m not sold on Corsi plus/minus in this regard. It’s a helluva stat for analyzing a team, but it’s hard to separate out the individual player, though Derek has given it a good shot here.

O’Sullivan has played mostly RW this year. You might want to include Nilsson (depending on how much he’s been paired with Horcoff) instead. I don’t know how much that changes things, but I doubt much if any.

er…nevermind. I didn’t read far enough into the comments, I guess.

I predicted 13 goals and 38 points from Horcoff before the season started.

Looks like I gave him too much credit.

Scott.

Well, we’re not going to agree here, not today at least ;).

 From what I see, people make as many mistakes and overstatements when they talk about Corsi plus/minus as they do when they talk about goals and assists. I see Corsi plus/minus constantly used as a summation of a player’s worth, as if it were the last word on all matters and to the exclusion of all other considerations.

I see it as a great team stat, one that really does help us identify superior teams, and one that is pretty good for rating line play, as well, or even maybe defensive pairing play, but not really so useful when it comes to rating individual players (though it’s fair to say that great players will drive a strong Corsi plus/minus and a weak player will drive a bad Corsi plus/minus, it’s just difficult to separate out who the real driver is, just by looking at Corsi numbers).

One confounding issue is that it’s not been proven in convincing fashion that goals necessarily come from territorial dominance, and that might limit the usefulness of Corsi plus/minus.

For instance, Visnovsky has strong Corsi because he usually plays with top offensive players, he’s one himself, and he’s overall quite effective. At the same time, he is prone to horrendous turnovers and defensive lapses, which show up just as a -1 on his Corsi plus/minus, but also -1 on the score clock. Corsi doesn’t give enough weight to major mistakes, as all mistakes are treated equally.

As for the relationship between territorial dominance and goals scored, I’m not so sure that it matters what zone a player starts in when it comes to goal scoring. This season, I’ve been tracking this, looking at how goals get scored.

Is it really such a kick in the groin to be out for so many defensive zone faceoffs, at least when it comes to your plus/minus? Or because of the fluid nature of hockey, is there any real disadvantage when it comes to your scoring and goal-causing error stats?

I can tell you that the play that kicks off an Oilers goal scoring play very often happens in the defensive end, with a defenceman winning the puck and moving it forward. Or it happens because of a neutral zone turnover. Or it happens because of a great end-to-end rush (less often).

Many goals, I’d estimate at least half, are scored off the rush, or off a rush and then a short amount of cycling.

I haven’t tabulated the results, though, of this season long survey, so I have nothing definitive to say here.

Scott. I know there is seemingly a relationship between Corsi plus/minus and goal scoring plus/minus for teams.

I know that good teams have territorial dominance and they also tend to outscore the opposition. I know that you will rarely see one without the other over time.

What I wonder, though, is whether or not good teams also outscore the opposition because they also have more strong breakout plays and up-ice rushes, plays that very often originate deep in their own zone.

Could it be that these uneven man rushes are just as important to goal-scoring, if not more important, than territorial dominance (as measured by Corsi plus-minus)?

I think you might find a small list of things that good teams do, and some might be more related to outscoring than others.

Essentially my question is this: is there a stronger relationship between uneven man breaks and outscoring than there is between shots at net differential and outscoring?

Now, no one tracks uneven man breaks, but I am tracking with the Oilers this year which zone each scoring play starts in. I’ve been recording it on every goal, the moment that the Oilers gain the initiative that leads to the goal being scored. Most often, the goal is scored within 1 to 15 seconds after this momentum change, I estimate. Much less often the goal comes through sustained pressure.

Very often the goal scoring starts in the offensive zone, with a faceoff won, a battle won, an interception of a pass. But very often scoring plays also start in the Oil’s defensive zone with a faceoff won, a battle won, an intercepted pass. The Oil get the initiative, they keep it and they score a goal.

So this is why I’m skeptical that what zone a player takes a faceoff in will have much impact on his scoring stats, even as they do impact his Corsi plus/minus. This could be the case, but I’m not sure it’s proven to be the case.

Because of the speed of skaters and of passing, NHL attackers quickly punish glaring errors, wherever they happen on the ice, and it could well be that those glaring errors happen just as often, if not more often, when a team is attacking, as opposed to concentrating on defence in its own end.

 I can’t say for sure, but it could well be, as hockey, IMO, is far more of a fluid game than a static one.