Philosophy of hockey sabermetrics: Can hockey accurately be measured?

[AKL]

They're meaningless. Mainly used by teams with highly ranked advanced stats to bring down teams ahead of them in the standings with worse advanced stats.

 

[Pi]

Biggest problem with advanced stats on these boards:

Most people here don't know a thing about Corsi/Fenwick but love to use them anyway.

 

[Muzzinga]

Mean nothing. For people who can't WATCH the games and know what's happening.

advanced stats provide much more meaning than extremely bias and results orientated eyeball tests

but as with every stat in hockey, they are useless on their own, and should be used in context along with other relevant stats

 

[Grant]

They can compliment your viewings from actually watching the games, but to use them on their own isn't going to work well.

 

[ChibiPooky]

So who's going to be the first to post advanced stats about advanced stats?

"RelCorsi is useful 77% of the time when discussing shot differential, but only when Fenwick is not used simultaneously."

 

[habs03]

IMO advance stat are good in the sense that they give you a little more detail, like who a player plays his minutes against, if they get more offensive zone starts vs defensive and stuff like that. But I think you need to take them in context, and need to watch to understand those stats.

Biggest issue I've seen with advance stats has been in the Norris thread, in regards to Subban and his TOI and who he plays against. Advance stat would show that he plays vs weaker opponents, and gets little ice time. But if you watch most of the games, and take everything into context, how he barely played in the first 20 regular season games, missed 6 games, and played about 18-19 minutes on the 3rd pair, you would get a difference picture of who he is playing, specially in a short season.

 

[here come the]

I think a fairly big flaw in the current advanced stats is no differential in the quality of shot given up. Giving up two attempted shots from the wall is worse for your corsi than one wide open in front of the net. Once that adjustment is made I think they will be a better indication

 

[Legionnaire]

So who's going to be the first to post advanced stats about advanced stats?

"RelCorsi is useful 77% of the time when discussing shot differential, but only when Fenwick is not used simultaneously."

I don't think it will be too long!

 

[WarriorOfGandhi]

a deeper understanding of the beautiful game is never a bad thing.

 

[Mr Lahey]

The biggest problem is people make inferences using descriptive statistics. Such a blatant misuse of statistics, something you'd learn in intro stats..

 

[Legionnaire]

a deeper understanding of the beautiful game is never a bad thing.

Statistics are for those who don't understand that beautiful game.

 

[AKL]

Biggest issue I've seen with advance stats has been in the Norris thread, in regards to Subban and his TOI and who he plays against. Advance stat would show that he plays vs weaker opponents, and gets little ice time. But if you watch most of the games, and take everything into context, how he barely played in the first 20 regular season games, missed 6 games, and played about 18-19 minutes on the 3rd pair, you would get a difference picture of who he is playing, specially in a short season.

Advanced stats are used because of this reason. If you looked now, maybe he's playing against tougher competition getting more time on ice, so without advanced stats like you mentioned, we might think this was going on all year. Advanced stats show that he played over half the season against worse competition in more sheltered minutes though.

 

[Canada4Gold]

I've noticd that many Leafs haters like using advanced stats to try and fulfill their dreams.

It's probably because we haven't made the playoffs since these advanced stats were invented so it's really a novelty to them I guess.

Hopefully in a couple years when we make the playoff 3 years in a row they will get bored of posting stats that nobody cares about to prove something that isn't true.

Advanced stats can be useful for sure when complementing them with other factors, but as a stand alone they show nothing at all.

 

[habs03]

Advanced stats are used because of this reason. If you looked now, maybe he's playing against tougher competition getting more time on ice, so without advanced stats like you mentioned, we might think this was going on all year. Advanced stats show that he played over half the season against worse competition in more sheltered minutes though.

Like I said, they are good to use but need to be put in context, I mean basically Subban started off being the teams 5-6th d-men for 15-20 games, and played the rest of the year playing as the teams number 1 d-men. I mean if you just look at his stats, you would think he is a middle pairing d-men, which he isn't.

 

[LuGBuG]

I've noticd that many Leafs haters like using advanced stats to try and fulfill their dreams.

It's probably because we haven't made the playoffs since these advanced stats were invented so it's really a novelty to them I guess.

Hopefully in a couple years when we make the playoff 3 years in a row they will get bored of posting stats that nobody cares about to prove something that isn't true.

Advanced stats can be useful for sure when complementing them with other factors, but as a stand alone they show nothing at all.

The Ducks had their fair share of "unsustainable" comments as well.

 

[The Bored Man]

Like all statistics, they can give you information about a subject but they MUST be used carefully. Biggest problem with their application on the Internet is that they're treated as iron-clad proof of a certain conclusion without looking at context or confounding information. Correlation is not equal to causation. Also, not all of them are equally useful.

 

[Grant]

I think a fairly big flaw in the current advanced stats is no differential in the quality of shot given up. Giving up two attempted shots from the wall is worse for your corsi than one wide open in front of the net. Once that adjustment is made I think they will be a better indication

Yup, a possibility for this could be looking at a percentage of goals from the area taken. If a shot from the boards only goes in 10% of the time, maybe multiply it by 10% and if a shot from right in front of the net goes in 50% of the time, multiply it by that.

Even still that wouldn't give accurate readings (due to screens, broken sticks, is the goalie in position, etc) but it would still be better than what is currently there.

 

[NaiveLeafsFan]

Statistics are for those who don't understand that beautiful game.

I trust stats far more than I trust my eyes. How many of you have played a professional level? Maybe if you've reached that echelon you can trust your eyes, otherwise you don't know a damn thing.

Advanced stats are useful in pinpointing whats going on with a player currently. Like most stats, they don't explain everything. You look at the Wild last year and the Leafs this year, high PDO. The difference between the two teams is one had enough games to face regression, whereas one has a short enough season it won't matter.

You don't think advanced stats work? Ask Minnesota fans when they're team exploded.

I think a fairly big flaw in the current advanced stats is no differential in the quality of shot given up. Giving up two attempted shots from the wall is worse for your corsi than one wide open in front of the net. Once that adjustment is made I think they will be a better indication

I'm a big proponent of shot quality. I think it would make Corsi and Fenwick alot more useful, although I have heard that shot quality is really negligible and it doesn't make a difference in terms of a possession metric.

 

[Nithoniniel]

Statistics of any kind is great if you understand what they do show. A lot of people just throw out statistics without any context, and context is very important for any stat. But just the same, people who like to disregard advanced stats in favor of eye-test don't seem to realize to what extent we see what we want to see. I've studied cognitive abilities on university level, and it has just made me all the more aware of the limitations of the eye-test.

I'm sure the best analytical hockey minds can just watch the game and pick up all the little details that makes up the basis of what's happening, but the big majority of us needs to support this with good statistics to be able to analyze what is going on accurately.

Unfortunately, very few people seem to be able to do this. There's tons of people who brings up plus/minus in a way to try to judge a player's defensive ability, and then comes along someone who only uses the eye-test and raises someone to the sky without realizing how sheltered the player is, and so on.

At least people on here have a basic understanding of the limitations of the two ways, among normal hockey fans though... not so much.

 

[Ducks DVM]

I trust stats far more than I trust my eyes. How many of you have played a professional level? Maybe if you've reached that echelon you can trust your eyes, otherwise you don't know a damn thing.

Advanced stats are useful in pinpointing whats going on with a player currently. Like most stats, they don't explain everything. You look at the Wild last year and the Leafs this year, high PDO. The difference between the two teams is one had enough games to face regression, whereas one has a short enough season it won't matter.

You don't think advanced stats work? Ask Minnesota fans when they're team exploded.



I'm a big proponent of shot quality. I think it would make Corsi and Fenwick alot more useful, although I have heard that shot quality is really negligible and it doesn't make a difference in terms of a possession metric.

They'll give you the accurate answer that the season imploded when they had massive injury problems, which advanced stats don't predict or take into account.

 

[SettlementRichie10]

Score Adjusted Fenwick is one of the best team stats around in judging possession time. The leading SAF team for four years now has either won the Cup or gone to the Finals. The only exception is LA last year, who were nevertheless the top SAF team in the league from the deadline on.

 

[Doctor No]

Mean nothing. For people who can't WATCH the games and know what's happening.

This sounds like what someone who doesn't grasp advanced statistics would say.

Claiming that advanced stats add no value to "watching the game" is just as silly as saying that watching the game adds no value to advanced stats. Each helps the other.

 

[CrazyCanucks]

It seems like in the mainstram media most media people dismiss it, but non msm people on twitter love to use it

 

[jfb392]

Some pretty funny happenings tonight re: advanced stats.
The Leafs were being outshot 11-3 by the Isles I believe, yet still had a 2-0 lead.
Leafs media guys that don't like advanced stats of words like "unsustainable" (Dave Shoalts, in particular) proceeded to make fun of stat geeks on Twitter because of what had unfolded.

At the end of the 1st period though, the Isles had a 19-5 edge in shots and 3-2 lead.

 

[Verviticus]

Some pretty funny happenings tonight re: advanced stats.
The Leafs were being outshot 11-3 by the Isles I believe, yet still had a 2-0 lead.
Leafs media guys that don't like advanced stats of words like "unsustainable", seemingly because they do not understand them, proceeded to make fun of stat geeks on Twitter because of what had unfolded so far.

At the end of the 1st period though, the Isles had a 19-5 edge in shots and 3-2 lead.

ironically, the msm guys then kinda shutup instead of pointing out that 3 goals in 19 shots is also a lot higher than average

it was really cool because at precisely that moment you realize "wow, these guys understand nothing"