You keep sidestepping the question. Does playing picture-perfect defense, as I described, count as a net negative for that entire shift?
- jmatchett383

Let's take this to PMs, we're probably annoying the other denizens.

(frank)ing communists. IMPEACH BETTMAN!!!
- jmatchett383

Nothing that radical; it's just HB talk. But lotteries and weighted percentages really make for unsettling scenarios in pro.sports, some of which are unfortunate, like the Sixers lat.year.

Let's take this to PMs, we're probably annoying the other denizens.
- Tomahawk

Well just say yes or no, whether that shift is a net negative or positive on your rating, and it ends. I'm just looking for a yes or no.

That's my problem with a lot of stats. They give values to our perceptions. Players don't play based on those numbers, the numbers are simply a trend based on what happens on the ice. That, and to tell us how worthless defensemen who can actually play defense are.
- jmatchett383

No doubt. Anyone can use those stats to validate what they want to see. It's a team number used incorrectly to assess individual players.

Well just say yes or no, whether that shift is a net negative or positive on your rating, and it ends. I'm just looking for a yes or no.
- jmatchett383

Yes, but I'd say it's not common.

Yes, but I'd say it's not common.
- Tomahawk

Agreed.

No doubt. Anyone can use those stats to validate what they want to see. It's a team number used incorrectly to assess individual players.
- MJL

There is some validity to it, but it's not a crystal clear, clear-cut picture.

You keep sidestepping the question. Does playing picture-perfect defense, as I described, count as a net negative for that entire shift?
- jmatchett383

Corsi isn't a reliable metric for individual shifts. Over the course of a season it paints a clearer picture. It's great if a team has tremendous defensive zone coverage, but if they have 15 shifts per game like that and generate no offense, they aren't helping the team. Even the best defensive players will make defensive zone mistakes that lead to goals or penalties.

That's a bit of a misconception.

CF/FF captures how the entire team does when he's on the ice. How he handles the puck is part of that, but all the other "little things" he does to contribute to the team's success (or failure) to control play while he's out there is captured as well to some extent: reads, positioning, support, good decision-making, etc.

It's not just a measure of puck skill.

There are plenty of examples of comparable players who aren't naturals with the puck but still managed not to hurt their teams in the possession game this past season: Scandella, Dillon, Methot, Volchenkov, Allen, Ericsson, Jackman, etc. Lot of those guys have the same mobility, injury issues, too.

To me, what separates them from Grossmann could be tactics (out of a player's control) and decision-making... I don't believe 8 is consistently making the best plays w/ and w/o the puck.
- Tomahawk

Same old problem. The numbers can't tell you if it's because those players managed not to hurt their team, or they were just carried along by the situations of who they play with and the situations they play in.

If you're questioning Grossmann's decision making and defensive reads. I think that is entirely false. He is very sound in both departments. He just lacks the physical ability to do a lot with the puck.

So, if the other team is cycling the puck, and the Flyers play fantastic defensive zone defense which is preached by coaches where they keep the puck to the outside, get their sticks in passing lanes to the inside, block shot attempts, lay a few nice hits, and then finally get the puck and clear it to center for a line change, that's considered "bad" by the rating system, correct?
- jmatchett383

The more important issue that the "rating system" faces is that defenseman A makes an errant pass, that leads to extended zone time for that shift, and 4 shots attempted against. The entire shift for all 5 skaters is spent defending until they can clear the puck and change. All 5 players skate to the bench with a -4 rating in shot differential because one player made an error.

Let's take this to PMs, we're probably annoying the other denizens.
- Tomahawk

NAH... let's see this playout. It's August.

But that's what I mean.

Now, granted, that's one instance over the course of a game and season. But on that shift, the defense would be doing everything right, but as far as the stats are concerned, that's a net negative rating on all 5 players because they did not have puck possession and when they did get the puck, they did not create a zone entry. That's an example of my issue with the possession ratings, they only tell a small part of the story.
- jmatchett383

Other wise known as false data!

NAH... let's see this playout. It's August.
- mayorofangrytown

Too late.

You keep sidestepping the question. Does playing picture-perfect defense, as I described, count as a net negative for that entire shift?
- jmatchett383

In an example even if the team was hemmed into their own zone, and the other team attempted 4 shots that were non threatening shots from the outside, or shots that were blocked, all 5 skaters would get a negative rating for the shift. Even though they defended well.

Yes, but I'd say it's not common.
- Tomahawk

It's very common.

Corsi isn't a reliable metric for individual shifts. Over the course of a season it paints a clearer picture. It's great if a team has tremendous defensive zone coverage, but if they have 15 shifts per game like that and generate no offense, they aren't helping the team. Even the best defensive players will make defensive zone mistakes that lead to goals or penalties.
- PhillySportsGuy

A larger sample size does not take care of the false data issue. Corsi doesn't measure goals or penalties.

[img]

if you can manage to watch the whole thing, you get a gold star. I couldn't last that long [/img]

Corsi isn't a reliable metric for individual shifts. Over the course of a season it paints a clearer picture. It's great if a team has tremendous defensive zone coverage, but if they have 15 shifts per game like that and generate no offense, they aren't helping the team. Even the best defensive players will make defensive zone mistakes that lead to goals or penalties.
- PhillySportsGuy

It's not worth it, people want to dismiss stats, let them.

There's no stats advocate that says they are perfect, no stats advocate that feels the numbers can replace watching the games, no stats advocate that feels they are 100% fool proof.

The only people that seem to feel something has to be flawless to be effective are the ones that bad mouth the stats. It sucks, but if somebody wants to engage in the endless debate, why bother?

Corsi isn't a reliable metric for individual shifts. Over the course of a season it paints a clearer picture. It's great if a team has tremendous defensive zone coverage, but if they have 15 shifts per game like that and generate no offense, they aren't helping the team. Even the best defensive players will make defensive zone mistakes that lead to goals or penalties.
- PhillySportsGuy

The truth is Corsi and Fenwick will be gone in 3 to 5 years when they get the tracking system in place. SportVU or something like it will make the guesstimating of Corsi and Fenwick obsolete. Why do you need something that approximates possession if you have actual possession statistics right in front of you?

That's why I try to steer clear of these kind of discussions. No NHL team uses the raw Corsi and Fanwick numbers the schmoos like us on the internet use because at best they're trends. Are they useful. Maybe when they started but they're already out dated and will be easily pushed aside like +/- in just a few seasons. It's really a fruitless discussion, analytics aren't going away, they're going to change. They already have. What they're offering us with Corsi and Fenwick are glorified plus minus numbers.

Too late.
- Tomahawk

You even got me to weigh in and I'm usually a bystander.

You even got me to weigh in and I'm usually a bystander.
- mayorofangrytown

So it was like a case of being denied something leading you to want it more (even though it's the last thing on earth you really want)?

The truth is Corsi and Fenwick will be gone in 3 to 5 years when they get the tracking system in place. SportVU or something like it will make the guesstimating of Corsi and Fenwick obsolete. Why do you need something that approximates possession if you have actual possession statistics right in front of you?
- mayorofangrytown

I suspect CF and FF will still live on, since shot attempts will always be relevant. The raw data will be a bit cleaner, though, since the human element will be taken out of scoring.

The truth is Corsi and Fenwick will be gone in 3 to 5 years when they get the tracking system in place. SportVU or something like it will make the guesstimating of Corsi and Fenwick obsolete. Why do you need something that approximates possession if you have actual possession statistics right in front of you?

That's why I try to steer clear of these kind of discussions. No NHL team uses the raw Corsi and Fanwick numbers the schmoos like us on the internet use because at best they're trends. Are they useful. Maybe when they started but they're already out dated and will be easily pushed aside like +/- in just a few seasons. It's really a fruitless discussion, analytics aren't going away, they're going to change. They already have. What they're offering us with Corsi and Fenwick are glorified plus minus numbers.
- mayorofangrytown

The idea behind advanced stats is to paint as accurate a picture as possible. Rather than go away, the benefits of SportVu will be hyper-accurate tracking and can eliminate the the questions about the data.

I'd say the large majority of stat heads look forward with excitement at the dawn of the NHL using a SportVu system to further refine the analytics.

I suspect CF and FF will still live on, since shot attempts will always be relevant. The raw data will be a bit cleaner, though, since the human element will be taken out of scoring.
- Tomahawk

In some of the interviews I listened to there was a lot of mention about pass attempts/completed, quality of turn overs leading the way in possession time so they would steer possession analytics away from Corsi and Fenwick numbers.

Just what I heard more than once.

Hey, at least I don't glaze over now.

In some of the interviews I listened to there was a lot of mention about pass attempts/completed, quality of turn overs leading the way in possession time so they would steer possession analytics away from Corsi and Fenwick numbers.

Just what I heard more than once.

Hey, at least I don't glaze over now.
- mayorofangrytown

Yeah, the shot stuff will probably get decoupled from the idea of possession for sure. Probably just go by pure puck-on-stick times for possession... and then all the cool ancillary stuff like shot attempts, passing accuracy/distance/reception, entries/exits, etc will follow that.

Fun time to be a (nerdy/obsessive) fan.