HFBoards

Go Back   HFBoards > NHL Western Conference > Central Division > Winnipeg Jets
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
Notices

Any shots-by-distance-breakdown per individual player?

Reply
 
Thread Tools
Old
07-13-2013, 01:44 AM
  #1
knorthern knight
Registered User
 
Join Date: Mar 2011
Location: GTA
Country: Canada
Posts: 2,876
vCash: 0
Any shots-by-distance-breakdown per individual player?

I'm starting a separate thread to avoid hi-jacking the "Line ups" thread. Question for garret9

Quote:
Originally Posted by garret9 View Post
Same with Jokinen.
His -19 isn't true indication of how he played as it is the very low probability result for the performance he gave and you can know that for certain due to his PDO.
PDO isn't controlled by players and this is a fact that is now widely known.
It's not a repeatable stat its random variance caused by goals being a low occurrance-event. If there were thousands of goals a season when a player is on the ice, it would be completely different.
I think you should read those linked articles.
One advanced-stats question. I think it might be part of the answer as to why we disagree. The crux of our disagreement is... is it "puck luck" or is it "making your own luck"? Advanced stats counts SV% when a player is on the ice. Is there a breakdown for number of shots each way BY DISTANCE for the time a given player is on the ice? The reason I ask is that I think that this may explain why we see things differently.

I think you'll agree that goalies generally have an easier time stopping a dump-in from the blue line than a player in on a breakaway, or all alone in front of the net. However, this is all smeared into 1 "PDO" number. Here's my theory...

Good all-around player forechecks+backchecks and is said to have "puck luck".
  • A good defenseman prevents shots against from in close, and clears opponents from infront of the net.
  • The goalie has an easier time with the long-range shots ("Vesa Toskala specials") and makes more saves.
  • Good defenseman is said to have "puck luck".
  • A good forward (e.g. Kane) gets in close to the net with the puck.
  • The opposing goalie has a harder time with the short-range shots and makes fewer saves.
  • Good forward is said to have "puck luck".
  • A not-so-good defenseman allows attacking forwards to shoot in close to his goal.
  • The goalie has a harder time with the short-range shots and makes fewer saves.
  • [wahhh wahhh wahhh] Bad luck [/wahhh wahhh wahhh]
  • A not-so-good forward doesn't get in close to the net with the puck.
  • The opposing goalie has an easier time with the long-range shots and makes more saves.
  • [wahhh wahhh wahhh] Bad luck [/wahhh wahhh wahhh]

I don't know how well I explained it, but that's my take on the difference between "puck-luck" and "making your own luck".

knorthern knight is online now   Reply With Quote
Old
07-13-2013, 05:57 AM
  #2
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
There is a way shot to do shot location estimations when a player is on the ice; I used it on my Pavelec article to show how the Jets were above avg in distance of shots agains than a lot of team's who had better Sv%.

I can look into it if you want.

I think I didn't explain something good enough. It's not that skill doesn't effect these things, its that everything, even outside of hockey, needs a solid enough sample size to diminish the effects of natural variance.
Some things by their nature need larger than others. On-Ice Sv%, On-Ice Sh%, PDO and +/- are all dependent on goal events which are really rare.
Corsi is superior than +/- for the fact that you on average are getting about 20-30 times the events in a single game. I've read that +/- takes about 3-4 seasons just to remove "white noise".

Anyways, the big thing is, if your hypothesis was true that skill is the dominant factor, you'd tend to see a correlation with higher % for better players, but you don't. It's completely random. Someone like Fehr is top40 one season in both categories, bottom 5 the next season and then top 15 in the next. Superstars have been near the bottom, and AHL replacement players like Machacek have led the league.

I guess what I'm trying to say is...
A 1st line star player's PDO probably should be maybe 5-10 points higher than an AHL callup if we could magically remove luck, due to the factors you mention above; however, natural variance is a far, far, far greater factor in this due to being a low occurrence event. Because of the larger strength of "luck" (ie: factors the player you are looking at can't control) the actual spread from top PDO to bottom PDO player is in the 100.



PS did you look at any of the recommended reading I had for you, because the answer to your questions are actually in them...


Eric T: Shooting percentage regression, Factoring regression into analysis, Fooled by randomness: How to evaluate defensemen
These 3 articles are probably the best for you as it is in response to someone asking very similar questions.

Gabe Desjardins: PDO: If you were going to understand just one NHL statistic, Advanced Metrics: PDO and regression to the mean, Luck vs Shot Quality in Shooting Percentage, Advanced Metrics: How useful is raw +/-?, On-Ice Shooting Percentage as a Player Talent

"Snark": SSW: With so much luck, when are stats useful?

Tyler Dellow: This time is different, Points v. Scoring, PDO Numbers

Vic Ferrari: Real Effects and Team Shooting Percentage at Even Strength

Jared Likens: Even Strength Shooting Percentage, How Much of Shooting Percentage Is Skill?, On Luck, Skill and Sample Size in Shooting Percentage


Last edited by garret9: 07-13-2013 at 07:21 AM.
garret9 is offline   Reply With Quote
Old
07-13-2013, 06:40 AM
  #3
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
I'd also point out the huge problem with your theory.
It would only work if hockey was 1v1.
Kane may be better at getting close, raising his personal shooting %, but that doesn't improve when Buff shoots from the point, etc, etc.


Some examples of randomness:
Olli Jokinen had a high (above average) PDO of 1010 in 2011-12, and a low (bottom 10) PDO of 939 in 2012-13.


George Parros, Steve Bernier, Chris Kelly, Rich Peverley, Jacob Josephson, Sean Couturier, Alex Tanguay, Roman Horak, Matt D'Agonstini, David Clarkson, David Desharnais, Jiri Hudler, Bobby Butler, and Mike Richards.
All players who were in top 20% of PDO for 2011-12 and bottom 20% of PDO in 2012-13.

That's 14 players players were in top ~50-60 one season and bottom ~50-60 the next.
All I did was only look at players who played at least half the season for 2011-12 and listed the top 20% of PDO.
Then I did the same for players listed in bottom 20% in 2012-13 that played at least half the season.
Then looked who was on both.

Take in that some players in each 20% didn't make the GP cutoff for the other year and you can see it's completely random and has players of all skill levels.


Now doing same for defense:
Adam Mcquaid, Andrew Walberts, Sheldon Brookbank, Tyler Myers, Filip Kuba, Philip Larsen, Ryan Ellis, Bryce Salvador, Barret Jackman, Karl Alzner

Thanks to injuries there's a lot less players who made the GP cut in one and not the other so that hurt but still lots of names


Last edited by garret9: 07-13-2013 at 06:59 AM.
garret9 is offline   Reply With Quote
Old
07-13-2013, 06:54 AM
  #4
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
Little
Shots for avg distance 35.8, on-ice sh% 10.79
Shots against avg distance 32.1, on-ice sv% 0.914
2012 PDO 1022
2011 PDO 980

Jokinen
Shots for avg distance 35.4, on-ice sh% 5.10
Shots against avg distance 35.6, on-ice sv% 0.881
2012 PDO 939
2011 PDO 1010

Burmistrov
Shots for avg distance 35.8, on-ice sh% 6.43
Shots against avg distance 35.5, on ice sv% 0.933
2012 PDO 997
2011 PDO 994

Antropov
Shots for average distance 34.7, on-ice sh% 8.56
Shots against average distance 33.9, on-ice sv% 0.946
2012 PDO 1031
2011 PDO 993


Last edited by garret9: 07-13-2013 at 07:13 AM.
garret9 is offline   Reply With Quote
Old
07-13-2013, 07:00 AM
  #5
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
The problem with your theory is there are 9 other shooters on the ice when the player is on. All of varying levels.

Luck is essentially 2 interrelated things in hockey:
1) Things out of the players control (bounces, other players, refs, etc)
2) When something that has high prob of success fails or low prob succeeds


Last edited by garret9: 07-13-2013 at 07:10 AM.
garret9 is offline   Reply With Quote
Old
07-13-2013, 07:10 AM
  #6
GrandChelems
Registered User
 
GrandChelems's Avatar
 
Join Date: Oct 2011
Location: Winnipeg
Country: Canada
Posts: 1,766
vCash: 500
That up at 7 am and off to practice followed by all day road hockey followed by hockey night in Canada followed by road hockey before bed kid in me just died ....

GrandChelems is offline   Reply With Quote
Old
07-13-2013, 07:18 AM
  #7
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
Another way to look at it is defensive teams tend to have low shots against, but vary in Sv% and shot distance.

High offensive teams tend to have high shots for, but vary in Sh% and shot distance.

garret9 is offline   Reply With Quote
Old
07-13-2013, 09:16 AM
  #8
truck
HFB Partner
 
truck's Avatar
 
Join Date: Jun 2012
Posts: 6,641
vCash: 500
Numbers!

truck is offline   Reply With Quote
Old
07-13-2013, 10:46 AM
  #9
allan5oh
#Dive4Five #31Buyout
 
Join Date: Oct 2011
Posts: 6,539
vCash: 500
I LIKE HOCKEY, I LIKE GOALS, I LIKE FIGHTS!

durr durr durr

allan5oh is offline   Reply With Quote
Old
07-13-2013, 11:31 PM
  #10
knorthern knight
Registered User
 
Join Date: Mar 2011
Location: GTA
Country: Canada
Posts: 2,876
vCash: 0
Yes, I did read the articles. Defensemen shooting from the point (blueline) is indeed a crapshoot. I cringe when Byfuglien winds up. If I were the opposition coach, I would instruct my opposite-side winger to...
  • notice whenever Byfuglien winds up to shoot
  • head towards centre on the opposite side
The shot misses and caroms around the net and out the other side past the blueline, more often than not. If our opponents had forwards ready to take advantage of this, we would be burned on so many breakaways, it wouldn't be funny. But I digress.

If advanced stats are to be believed, then give extra time to the players who currently have higher PDO. When their lucky streak ends, and someone else starts "getting lucky", switch to them. As they say in football "Go with the Mo".

knorthern knight is online now   Reply With Quote
Old
07-13-2013, 11:53 PM
  #11
garret9
AKA#VitoCorrelationi
 
garret9's Avatar
 
Join Date: Mar 2012
Location: Vancouver
Posts: 11,198
vCash: 500
Quote:
Originally Posted by knorthern knight View Post
Yes, I did read the articles. Defensemen shooting from the point (blueline) is indeed a crapshoot. I cringe when Byfuglien winds up. If I were the opposition coach, I would instruct my opposite-side winger to...
  • notice whenever Byfuglien winds up to shoot
  • head towards centre on the opposite side
The shot misses and caroms around the net and out the other side past the blueline, more often than not. If our opponents had forwards ready to take advantage of this, we would be burned on so many breakaways, it wouldn't be funny. But I digress.

If advanced stats are to be believed, then give extra time to the players who currently have higher PDO. When their lucky streak ends, and someone else starts "getting lucky", switch to them. As they say in football "Go with the Mo".
I don't know about that.

I say go with probabilities. LLW+Kane create more offensive chances per minute so even if their % drops down they will still make up for it with volume.

Plus some players bring things besides scoring. Possession, reducing number of chances, etc.

garret9 is offline   Reply With Quote
Reply

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 06:13 PM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. ©2014 All Rights Reserved.