Patty Kazmaier 2014

[Eeyore]

Re: Patty Kazmaier 2014

Over the course of the season, many of the things you cite will average themselves out over time.

Except that they don't. Coaches consistently use the same players against opponents' top lines and to take defensive zone face-offs. One thing that hockey analytics has shown conclusively is that you must take the quality of opposition faced and where a player starts their shifts into consideration if you want to approximate an individual player's value to their team.

Unlike the other stats, it at least rewards players who forecheck and back check and are more likely to be where they should be when without the puck, and takes into account that at times the other 4 players on the ice with you are not always exactly the same even if your line mates or D partner is consistent. No other stat does that.

None of the commonly available stats, no. Then again, +/- doesn't do a very good job of tracking that, either.

The arguments you make could equally be said for goals and assists.

Those things are true of goals and assists, however, goals and assists are not as dependent upon which teammates you play with as +/- is. For all of their flaws as stats, goals and assists require some sort of affirmative action on the part of the player credited with them. +/- does not. All it tracks is whether you were on the ice when something happened, not whether you had anything to do with it happening.

They are largely a function of your total minutes, which line you play on, and the strength of your line mates. If you happen to be paired on a line with a Kaz nominee, your production is going to benefit dramatically versus someone used as a third line checker. Both players may be equally capable of posting similar scoring stats given the same situation. In the latter case, the + / - can show the true overall value of both players to the team although their roles are different.

That's the theory. In practice it doesn't work out that way.

 

[D2D]

Re: Patty Kazmaier 2014

And here I was thinking the master was back until I saw this hidden message! LOL!

 

[Trillium]

Re: Patty Kazmaier 2014

Except that they don't....

goals and assists are not as dependent upon which teammates you play with as +/- is. .....


That's the theory (referring to overall value of players based on + / -). In practice it doesn't work out that way.

That's all your opinion. I couldn't disagree with you more. And you state no facts to back yourself up.

The highest impact players on a team in any given year for those who see ALL their contributions to the team on a consistent basis do also tend to have among the best +/-. I can't recall a team I've ever followed closely where this was not the case (with the exception of those forwards--and defensemen--who don't pay actually tend to pay to much attention to their defensive responsibilities.

Using your example of players who are put out regularly for defensive zone face-offs (and therefore presumably get scored against more often), I don't know why any coach would choose to put out players in that situation over and over who were not among the teams' strongest defensively, nor unable to get the puck down into the offensive zone. Also, quite a few programs switch the start order of their #1 and #2 lines--probably more often than they change line combinations.

 

[ARM]

I can't recall a team I've ever followed closely where this was not the case (with the exception of those forwards--and defensemen--who don't pay actually tend to pay to much attention to their defensive responsibilities.

I don't understand the piece after the opening parenthesis.

Using your example of players who are put out regularly for defensive zone face-offs (and therefore presumably get scored against more often), I don't know why any coach would choose to put out players in that situation over and over who were not among the teams' strongest defensively, nor unable to get the puck down into the offensive zone.

They are among the team's best defensively, and that is why the coach puts that line out there. However, it is a disadvantage to start out in your defensive zone rather than the offensive zone, and that is the reason coaches complain when they don't like where the officials determine a faceoff will take place. Usually, the good defensive line is successful at getting the puck out of the zone. But over the course of the season and hundreds of faceoffs, it can be a factor in +/-.

I think you are both right and both wrong. The +/- stat isn't perfect, but given we don't have access to anything else, it does provide another piece of information.

 

[TonyTheTiger20]

Re: Patty Kazmaier 2014

And here I was thinking the master was back until I saw this hidden message! LOL!

Just trying to get myself warmed up for tournament time

 

[ARM]

Just trying to get myself warmed up for tournament time

We should create a poll thread on what round and to what opponent BC falls in the HEA Tourney and earns the never-ending hatred of some team that will wind up an ultimately useless 8th in the PWR.

 

[DC78-82]

Re: Patty Kazmaier 2014

We should create a poll thread on what round and to what opponent BC falls in the HEA Tourney and earns the never-ending hatred of some team that will wind up an ultimately useless 8th in the PWR.

There's always tomorrow, for dreams to come true~

 

[TonyTheTiger20]

Re: Patty Kazmaier 2014

We should create a poll thread on what round and to what opponent BC falls in the HEA Tourney and earns the never-ending hatred of some team that will wind up an ultimately useless 8th in the PWR.

I would forgive whichever team beats us in the WHEA tourney if they knocked out your gals at Ridder

**Knocked out as in terms of 'score more goals than,' not knocked out in terms of 'Monique & Jocelyne Lamoureux are hacks and will physically knock you out.'

 

[Eeyore]

Re: Patty Kazmaier 2014

That's all your opinion. I couldn't disagree with you more. And you state no facts to back yourself up.

Actually, I did. It is a fact that percentage of offensive vs. defensive zone face-offs makes a meaningful difference in a player's plus/minus. It is a fact that a plus/minus event doesn't require any affirmative play on the part of each player credited. It is a fact that who you play against has an impact on a player's plus/minus.

What I didn't do was provide a link to someone else stating the same facts or providing an analysis of those facts. Of course, you haven't backed up anything you've said with that sort of thing, either, so acting all huffy about it isn't very impressive.

Actually coming up with a convenient, one stop explanation of this isn't terribly easy because it's something that has been so obvious to hockey analysis for almost a decade that no one writes it up from that perspective much any more. You get lots of bits and pieces of it as it is turned over and examined in different ways but not the whole thing. So Ill try my best.

This is probably the best overview of the question that I could find reasonably quickly. It mentions all of the things that I did with some extra analysis. To the problem of plus/minus reflecting the actions of teammates rather than the individual, video analysis found that around 40% of the attributions of a plus or a minus to an individual player reflected occasions on which they did not have a meaningful impact on the play, either contributing positively to a goal being scored or making some sort of mistake that led to a goal against. So about 40% of the information that plus/minus seems to present for an individual is in fact nothing but random noise. That alone is enough to make it borderline useless even before considering other problems of context.

Another issue is that there is a lot of noise within the goals scored data itself. Scoring a goal in hockey is a sufficiently rare event that there is a large random component to it. There are a lot of goals scored on plays that were not meaningfully better than plays that didn't score. This is one of the reasons that analytics focuses more on shots attempted rather than goals: it has less noise in it and is in fact a better predictor of future goals scored than is the actual figure for goals scored. (This brings us to the difference between descriptive and predictive statistics and I'm going to let it slide because I think that an award like the Kazmeier is better given to someone with good descriptive stats than good predictive ones, relatively speaking. But if you want to have an idea of how good a team is going to be in the playoffs before they happen, predictive stats are more valuable.)

Here are some other links that get at some of the issues.

5 Hockey Analytics Concepts Every Hockey Fan Should Learn
The first part of a description of NHL Adjusted Plus/Minus and how to try to produce good information with it.
An Introductory Guide to Advanced Hockey Analytics

One element that is somewhere in there but I don't remember which piece (and it might be in one I didn't link to) is that plus/minus also suffers from being a net statistic. A player on the ice for five goals for and five goals against has the exact same plus/minus as a player that is on the ice for no goals either way but these are clearly very different situations.

 

[Eeyore]

Re: Patty Kazmaier 2014

I think you are both right and both wrong. The +/- stat isn't perfect, but given we don't have access to anything else, it does provide another piece of information.

This is something I've noticed before and on which we'll just have to disagree. Having more information is not helpful, and can be harmful, if it's bad information. Plus/minus has so many problems that it can often lead you away from correct conclusions.

 

[pokechecker]

Re: Patty Kazmaier 2014

the idea is to put the puck in the opponents net and keep it out of yours, that's what +/- keeps track of

over time it doesn't matter who you go out on the ice with or who is on the ice for the opponents, the more games, the more accurate +/- reflects on a players ability at what really matters: offense AND defense

you can make a bunch of great plays, but if two of your team mates touch the puck after you before the goal is scored you get zero credit in the points standings, that's just one of the reasons +/- is a great statistic to look at

in the end I don't care who wins the PK, it's just a beauty contest with all the politics that goes along with it

 

[Hux]

the idea is to put the puck in the opponents net and keep it out of yours, that's what +/- keeps track of

over time it doesn't matter who you go out on the ice with or who is on the ice for the opponents, the more games, the more accurate +/- reflects on a players ability at what really matters: offense AND defense

you can make a bunch of great plays, but if two of your team mates touch the puck after you before the goal is scored you get zero credit in the points standings, that's just one of the reasons +/- is a great statistic to look at

in the end I don't care who wins the PK, it's just a beauty contest with all the politics that goes along with it

Not really. I've seen plenty of games where particular players cover their responsibilities like a blanket and a teammate doesn't which results in a goal...and a nasty minus in the +/- column.

 

[northhockey]

Re: Patty Kazmaier 2014

'Monique & Jocelyne Lamoureux are hacks and will physically knock you out.'

Just imagine how many more points they would've had in their career had they learned to stay out of the box! Who knows..maybe they could've went down as the top scorers ever in the history of the conference that wins it all every year...

 

[northhockey]

Re: Patty Kazmaier 2014

Wisconsin gives up 1.1/goals per game and the one Badger in the top ten is a forward who is t17th in points?

 

[FiveHoleFrenzy]

Re: Patty Kazmaier 2014

Ammerman's inclusion in the 10 finalists is a bit of a head scratcher.

 

[TonyTheTiger20]

Re: Patty Kazmaier 2014

Just imagine how many more points they would've had in their career had they learned to stay out of the box! Who knows..maybe they could've went down as the top scorers ever in the history of the conference that wins it all every year...

Hey and if I could score 2 PPG and be on a national championship contending team I could probably win the Hobey!

 

[brookyone]

Re: Patty Kazmaier 2014

Wisconsin gives up 1.1/goals per game and the one Badger in the top ten is a forward who is t17th in points?

Ammerman's inclusion in the 10 finalists is a bit of a head scratcher.

Sentiments and puzzlement I share.

 

[NMH]

Re: Patty Kazmaier 2014

Sentiments and puzzlement I share.

Ditto.

I mean, Britt's a good player and an even better person and I feel bad questioning the inclusion, but even I had to wonder how she made the list.

Here's what I've got in terms of trying to figure out justification ....

The votes were cast/ballots filled out while Rigsby was injured, so people didn't want to vote for her in case she didn't return.

They didn't want to not vote for someone from the No. 2 team in the country.

Ammerman is a "feel good" story, returning to the ice after missing an entire season to concussion and thinking she might not skate again and then spending her summer in Africa doing charity work.

They justified it with the character clause, her leading the WCHA in conference points (most points against top talent) and her leading the country (tied) in PPG (there's a "performance in the clutch" clause in the description).


That being said, I'm not entirely sure you couldn't come up with justification for every girl 9-16 in scoring, so ...

 

[PrincetonFan]

Re: Patty Kazmaier 2014

So who will the final three be? Rattray for sure. Who else?

 

[NMH]

Re: Patty Kazmaier 2014

Brandt should be, but I think Minnesota hangover is a very real thing. It would not surprise me if she suffers because of it.