NHL.com will start tracking "advanced" stats

[FBI]

I couldn't find penalties drawn.. How did Johnny do on that one?

[Finger Cookin]

Originally Posted by FBI

I couldn't find penalties drawn.. How did Johnny do on that one?

The stat is only shown as PenDr/20 and is under the Skater Scoring report. Johnny is 6th on the team with 0.33 PenDr/20. The leader on the team in this category is Lance Bouma at 0.51. The league leader is Tom Setsito at 5.34, but in only 3 games played. For regular players, the league leader is Tom Wilson at 1.55.

As getbak mentioned, the stat seems pretty skewed right now as it seems to include coincidental penalties. I'm also not clear if the measure is penalty minutes drawn per 20 minutes of play, or *penalties* drawn per 20 minutes of play.

It's early days.

[Resolute 14]

Originally Posted by Joborule

Considering I don't even remember what those names stood for half the time, it's a no brainer to go with descriptive stat names since only a very small minority knew what the names represented.

True, but at the same time, I've known of Corsi for about six years now. Changing to "SAT" will take some time to get comfortable with. And, with the aforementioned ego problem, I expect analytics guys will resist.

[Quincy Egg]

Calgary appears to have benefited a lot from luck going by their Corsi and Fenwick- pretty much parallels the Avs last season.

[Resolute 14]

Originally Posted by Quincy Egg

Calgary appears to have benefited a lot from luck going by their Corsi and Fenwick- pretty much parallels the Avs last season.

Yeah, we discussed this in the other advanced stats thread, but most playoff teams will have higher Corsi/SAT numbers. Those that don't usually outscore or outsave their Corsi. Colorado last year did it through both. Montreal, who had a worse Corsi than the Avs, did it through Carey Price. The Flames are doing it this year through offence.

The Habs are actually the cautionary tale against regarding winning despite poor Corsi/Fenwick as being the result of "luck". The Avs and Habs were in the same place last year, but the analytics community allowed expectations to colour their view. They didn't expect Colorado to do well, so became offended when the Avs won their division. They did expect Montreal to do well (despite the stats), so were utterly silent about "luck".

[MrMastodonFarm]

Dude, just report his off topic post, don't give into the obvious troll. Guy hasn't posted in over a year, signs in just to take an obvious swipe at the Flames. The thread is about the NHL adding new stats, not how the Flames will suck next year because of it.

When the troll is that obvious, report and move on. Not worth the typing.

[Resolute 14]

Perhaps, but real analytics guys think and say the same. It's always nice to take a swipe at pointing out how they are equally as bad as anyone else for using the eye test to frame how they view their own stats.

[Enoch Root]

Quincy Egg is a Nuck troll - and not even a very good one

[CorsiHockeyLeague]

Originally Posted by Finger Cookin

The stat is only shown as PenDr/20 and is under the Skater Scoring report. Johnny is 6th on the team with 0.33 PenDr/20. The leader on the team in this category is Lance Bouma at 0.51. The league leader is Tom Setsito at 5.34, but in only 3 games played. For regular players, the league leader is Tom Wilson at 1.55.

As getbak mentioned, the stat seems pretty skewed right now as it seems to include coincidental penalties. I'm also not clear if the measure is penalty minutes drawn per 20 minutes of play, or *penalties* drawn per 20 minutes of play.

It's early days.

It is.

You can basically ignore the NHL's penalties drawn stat, because they screwed it up. The reason Sestito and other goons are on top is because the NHL for reasons passing understanding decided to include fighting majors in the equation, which is obviously ridiculous.

The rest of the page is so far basically useless because it doesn't filter by game state (are the stats we're seeing just even strength stats, just 5 on 5 stats, or in all game states?) and have no option to filter by GP/TOI.

This is going to take some tweaking before it's worth using.

[Cflames_12.5]

I'm not a fan of the enhanced stat's. Mostly because I don't understand them. I don't hate them. The pro advance stats guys I'm sure are nice guys but they are very annoying. They are always saying team A is playing above their head just watch the enhanced stats say so. And when you lose they are always there to rub dirt in your face face with a big I told you so. The Flames are a young team and they will come around. Most of the teams on the top of the enhanced stats list have been together for a long time.

I for one have been thoroughly entertained this year. And would rather take these win anyway they come. I would rather have bad enhanced stats and playing meaningful hockey, then have good enhanced stats and a horrible losing record. I'm sure though if the flames had good enhanced stats I might care more about them.

[mikeecho]

WRT to the names of some of the "enhanced Stats" taking on descriptive names... EXCELLENT!

I wonder if the people (likely league employees way back when) who thought of measuring +/-, PIM, SOG, GAA, etc... are bitter the measurements weren't named after them? It's a stupid concept.

Imagine if Backlund has a Morrison of +3, with 5 Smith's and 64 Davidson's.

What about the people who pioneered some of the "enhanced stats" that already have descriptive names. Someone though of Zone Starts, but didn't plague it with their last name or some other cutesy name. Respect to these people who I suspect realize that it is and should be about the product on the ice.

[driveway]

Originally Posted by Cflames_12.5

I'm not a fan of the enhanced stat's. Mostly because I don't understand them. I don't hate them.

As they come to be talked about more and more, you'll come to understand them. They are best when used for predicting long-term trends. They, just like almost every other hockey stat, are terrible for predicting who will win a particular game, but over the long-term of an entire season, or multiple seasons, these new stats seem to be good at predicting success.

Think about all the people who 10-15 years ago hated the idea of "Wins-Above-Replacement" in baseball or "DVOA" in Football, but those have gained pretty universal acceptance.

It will take time, but it's worth following along with the stats.

[saillias]

Originally Posted by mikeecho

Imagine if Backlund has a Morrison of +3, with 5 Smith's and 64 Davidson's.

I'm glad Monahan has picked up his Rhino production this year.

[Rerun]

Came in here expecting to find Sheldon, Leonard, Howard, and Raj.

Left satisfied.

[Rick M.]

I'll feel better about puck possession stats when they are actually measured in time (minutes and seconds) and not by a proxy such as shot attempts.

[Tinordi]

Originally Posted by Rick M.

I'll feel better about puck possession stats when they are actually measured in time (minutes and seconds) and not by a proxy such as shot attempts.

You should only care about that stat if it correlates to something that you care about. If time of possession has no correlation to winning then there's no point in tracking the stat. Which is why so many people are focusing on corsi close, because it correlates to winning, meaning that it is more or less predictive in its value.

[Rick M.]

Originally Posted by Tinordi

You should only care about that stat if it correlates to something that you care about. If time of possession has no correlation to winning then there's no point in tracking the stat. Which is why so many people are focusing on corsi close, because it correlates to winning, meaning that it is more or less predictive in its value.

Fair enough. A stat that correlates SOMEWHAT with winning is better than nothing.

[Tinordi]

Originally Posted by Rick M.

Fair enough. A stat that correlates SOMEWHAT with winning is better than nothing.

Yes that's the thing about statistics, they provide us with insight not certainty. Your statement is less a damning critique of them then it is a definition.

[mrkajz44]

Just curious here on the whole "corsi = possession" argument.

I understand that actual puck possession was measured for about 20 games for the Leafs in close (maybe even tied) situations and compared to corsi. The results showed that corsi was a decent proxy for possession in this case.

My question is, how many other times has this been done? Was it only done a handful of times and now is "proven"? Was it attempted on multiple teams with different playing styles? Was it attempted both early and late in the regular season? In the playoffs?

I'm not trying to bash anyone here, just curious as to the amount of times it was proven out that corsi is a good proxy for possession.

[Street Pharmacist]

Originally Posted by mrkajz44

Just curious here on the whole "corsi = possession" argument.

I understand that actual puck possession was measured for about 20 games for the Leafs in close (maybe even tied) situations and compared to corsi. The results showed that corsi was a decent proxy for possession in this case.

My question is, how many other times has this been done? Was it only done a handful of times and now is "proven"? Was it attempted on multiple teams with different playing styles? Was it attempted both early and late in the regular season? In the playoffs?

I'm not trying to bash anyone here, just curious as to the amount of times it was proven out that corsi is a good proxy for possession.

Some Flyers fans did it too. In the end, why does it matter if it measures "possession"? It's a red herring. It's usefulness and predictive power aren't related to whether or not it measures "possession". Measuring who throws the puck towards the opposition's net more often seems to me to be more important in determining who played better over who had the puck more.