HFBoards

Go Back   HFBoards > General Hockey Discussion > National Hockey League Talk
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
National Hockey League Talk Discuss NHL players, teams, games, and the Stanley Cup Playoffs.

sportsclubstats.com please explain

Reply
 
Thread Tools
Old
01-27-2011, 11:51 AM
  #1
Realm
Registered User
 
Realm's Avatar
 
Join Date: Jun 2005
Posts: 5,554
vCash: 500
sportsclubstats.com please explain

How do they get these percentages of making the playoffs? I know some of the formulas but it makes no sense. How are the Kings at 60.4% to make the playoffs and the Sharks are 42.8%? Sharks have 1 more point, they have played the same amount of games, and the Sharks have 2 more home games left than the Kings do.

Realm is offline   Reply With Quote
Old
01-27-2011, 11:54 AM
  #2
HeadLikeAnOrange*
 
Join Date: Jun 2010
Location: Vancouver, BC
Posts: 1,125
vCash: 500
Quote:
Originally Posted by Realm View Post
How do they get these percentages of making the playoffs? I know some of the formulas but it makes no sense. How are the Kings at 60.4% to make the playoffs and the Sharks are 42.8%? Sharks have 1 more point, they have played the same amount of games, and the Sharks have 2 more home games left than the Kings do.
QoC and goal diff are heavily weighted (as they should be)

Kings are much better than their record.

HeadLikeAnOrange* is offline   Reply With Quote
Old
01-27-2011, 11:55 AM
  #3
Orenji
Registered User
 
Orenji's Avatar
 
Join Date: Jan 2010
Location: Winningland
Country: United States
Posts: 5,290
vCash: 500
Goal Differential.

Strength of Schedule.

Orenji is offline   Reply With Quote
Old
01-27-2011, 11:55 AM
  #4
Bill McNeal
Registered User
 
Join Date: Jul 2003
Location: Montreal
Country: Canada
Posts: 12,662
vCash: 50
Change it to 50/50 and the Sharks have a better chance of making it.

The standard results are weighted, so the quality of the opponent, home ice and other factors play a role.

Bill McNeal is offline   Reply With Quote
Old
01-27-2011, 11:57 AM
  #5
Dirty Kari
Registered User
 
Dirty Kari's Avatar
 
Join Date: Aug 2003
Location: Arlington, TX
Country: United States
Posts: 2,043
vCash: 500
They have a better goal differential. One team is technically 25-25 (SJ) while the other is 27-23 (LA). LA has a much better home record. The ratings are also waited based on QOC. These kinds of things could factor in to why one team is rated more likely to make it over the long hall and mean more than (1) point.

Dirty Kari is offline   Reply With Quote
Old
01-27-2011, 12:11 PM
  #6
IdealisticSniper
Registered User
 
IdealisticSniper's Avatar
 
Join Date: Nov 2008
Posts: 10,987
vCash: 500
Quote:
Originally Posted by HeadLikeAnOrange View Post
QoC and goal diff are heavily weighted (as they should be)

Kings are much better than their record.
Thats the main problem with that site. Goal differential is almost as bad of a stat to go by as plus minus when talking about how good a team or player is.

One or two blowouts either way inflate or deflate that goal differential number so badly. The Lightning are a perfect example. They have been routed like 4 times this season so it killed their goal differential. However they win a lot of 1/2 goal games. They are a good team but because of the 4 routes their goal differential is only even. If you take out the four routings they would be like a +26 or +27 in goal differential.

Its a very misleading stat.

IdealisticSniper is offline   Reply With Quote
Old
01-27-2011, 12:19 PM
  #7
Teemu
Moderator
 
Teemu's Avatar
 
Join Date: Dec 2002
Location: Palatine
Country: United States
Posts: 21,950
vCash: 500
Some helpful info:
http://www.sportsclubstats.com/WhatIsThis.html

Teemu is offline   Reply With Quote
Old
01-27-2011, 12:19 PM
  #8
tarheelhockey
Global Moderator
 
tarheelhockey's Avatar
 
Join Date: Feb 2010
Location: The Triangle
Country: United States
Posts: 37,509
vCash: 500
Official NHL records:
San Jose 25-19-6
Los Angeles 27-22-1

"Old school" W-L-T records:
San Jose 22-23-5
Los Angeles 22-23-5

Funny how the OT and SO points can make teams appear far apart when in reality they're tied.

tarheelhockey is online now   Reply With Quote
Old
01-27-2011, 01:01 PM
  #9
Fictional Realism
Moustache Power
 
Fictional Realism's Avatar
 
Join Date: Jun 2003
Location: The Crease
Country: Canada
Posts: 15,499
vCash: 894
Quote:
Originally Posted by IdealisticSniper View Post
Thats the main problem with that site. Goal differential is almost as bad of a stat to go by as plus minus when talking about how good a team or player is.
Heh, I always see people say this, but outside of Tampa all the top teams have significant positive goal differentials. I think it's a fairly good stat insomuch as 'if you score more goals then you allow you'll win more games.' Obviously, you'll have outliers like LA, but it's usually a fairly good indicator of who the elite teams happen to be at any given time.

Fictional Realism is offline   Reply With Quote
Old
01-27-2011, 01:07 PM
  #10
Finnish your Czech
BestPosterAvailable
 
Finnish your Czech's Avatar
 
Join Date: Nov 2009
Location: Toronto
Country: Finland
Posts: 46,892
vCash: 50
Quote:
Originally Posted by tarheelhockey View Post
Official NHL records:
San Jose 25-19-6
Los Angeles 27-22-1

"Old school" W-L-T records:
San Jose 22-23-5
Los Angeles 22-23-5

Funny how the OT and SO points can make teams appear far apart when in reality they're tied.
uh, SJ only has one more point than LA

Finnish your Czech is offline   Reply With Quote
Old
01-27-2011, 01:07 PM
  #11
tarheelhockey
Global Moderator
 
tarheelhockey's Avatar
 
Join Date: Feb 2010
Location: The Triangle
Country: United States
Posts: 37,509
vCash: 500
Getting blown out several times is reason enough to move down in people's estimation, don't you think?

tarheelhockey is online now   Reply With Quote
Old
01-27-2011, 01:09 PM
  #12
Kaesar
Registered User
 
Join Date: Oct 2003
Posts: 485
vCash: 500
Quote:
Originally Posted by IdealisticSniper View Post
Thats the main problem with that site. Goal differential is almost as bad of a stat to go by as plus minus when talking about how good a team or player is.

One or two blowouts either way inflate or deflate that goal differential number so badly. The Lightning are a perfect example. They have been routed like 4 times this season so it killed their goal differential. However they win a lot of 1/2 goal games. They are a good team but because of the 4 routes their goal differential is only even. If you take out the four routings they would be like a +26 or +27 in goal differential.

Its a very misleading stat.
You can't be more wrong on this. Historically goal differential is the single most accurate stat in predicting future results. That's not saying you should bet your farm using it, it just means that comparatively, if you had to predict a game between two teams looking at only one of points, wins, or goal differential, GD will give you a more accurate result over the long run.

Kaesar is offline   Reply With Quote
Old
01-27-2011, 01:10 PM
  #13
tarheelhockey
Global Moderator
 
tarheelhockey's Avatar
 
Join Date: Feb 2010
Location: The Triangle
Country: United States
Posts: 37,509
vCash: 500
Quote:
Originally Posted by Finnish your Czech View Post
uh, SJ only has one more point than LA
Maybe not the points themselves, but they way they're represented in the standings. A casual hockey fan would think these two clubs, which have identical records in the traditional sense, were 2-3 games apart.

tarheelhockey is online now   Reply With Quote
Old
01-27-2011, 01:11 PM
  #14
What the Faulk
The Real Swede Shady
 
What the Faulk's Avatar
 
Join Date: May 2005
Location: North Carolina
Country: United States
Posts: 31,399
vCash: 500
Quote:
Originally Posted by tarheelhockey View Post
Getting blown out several times is reason enough to move down in people's estimation, don't you think?
It's the Atlanta Falcon defense. Somehow winning close games against bad teams somehow makes you a good team?

What the Faulk is offline   Reply With Quote
Old
01-27-2011, 01:55 PM
  #15
Realm
Registered User
 
Realm's Avatar
 
Join Date: Jun 2005
Posts: 5,554
vCash: 500
Ok, maybe the better question is do you believe in the way they run the percentages? If the Kings are +18 and the Sharks are +1 in GD and the Sharks still have a better win loss record dont you think that is more important? I would prefer a team that wins the close games with the equal record than the team that has blowout wins with an equal record. Doesnt that prove you are better in pressure situations? Wouldn't that mean in the pressure of a playoff push they would be in better shape? Obviously there is not true formula, but this one doesn't work. Also, like I said the Kings have more road games left, and their road record is nowhere near their home record, blowout home losses shouldnt help them in their future road games.

Realm is offline   Reply With Quote
Old
01-27-2011, 02:02 PM
  #16
eklunds source
Registered User
 
Join Date: Jul 2008
Location: Ed Snider's basement
Posts: 8,325
vCash: 500
They're breaking down the likelihood of teams making the playoffs. Don't look at a statistic like goal differential in isolation. There are many other factors at play: LA has played 10 games vs the weak NW (8-1-1), and has 10 remaining. San Jose has played 14 against them (7-6-1) and just 6 remain.. LA likely has an easier schedule in terms of quality of opponents than SJ.

Statistics aren't facts, they're clues. Jesus.

eklunds source is offline   Reply With Quote
Old
01-27-2011, 02:13 PM
  #17
Ron Barr
Doing it to Death
 
Ron Barr's Avatar
 
Join Date: Dec 2009
Location: bdddddddet
Posts: 5,820
vCash: 500
More of a reason for me to believe that the NHL should drop the point system, and just go by wins and losses. Loser points are stupid.

Ron Barr is offline   Reply With Quote
Old
01-27-2011, 02:23 PM
  #18
SLang
Registered User
 
SLang's Avatar
 
Join Date: Nov 2005
Location: Huntington Beach, CA
Posts: 3,685
vCash: 500
Quote:
Originally Posted by Realm View Post
If the Kings are +18 and the Sharks are +1 in GD and the Sharks still have a better win loss record dont you think that is more important?
Yes, but only if it's the last day of the regular season. But that's not the point here. It's just a predictive formula that attempts to forecast performance, by taking into account many factors. I'm sure current record is accounted for, but just not weighted as heavily as GD and strength of upcoming schedule, both of which are hugely important in predicting success/failure.

SLang is offline   Reply With Quote
Old
01-27-2011, 03:22 PM
  #19
Hammer79
Registered User
 
Join Date: Jan 2009
Location: Kelowna
Country: Canada
Posts: 5,217
vCash: 500
Quote:
Originally Posted by Realm View Post
How do they get these percentages of making the playoffs? I know some of the formulas but it makes no sense. How are the Kings at 60.4% to make the playoffs and the Sharks are 42.8%? Sharks have 1 more point, they have played the same amount of games, and the Sharks have 2 more home games left than the Kings do.
If you want the mathematical explanation for that sites formula, then look at page 11 of this pdf link:

http://www.hockeyanalytics.com/Resea...babilities.pdf

E = (Goals for per game + Goals against per game)^P
P = exponent, between 0 and 1, higher the number, the more emphasis that is placed on goal differential in predicting outcome.
Win Probability = ((Goals for)^E) / (((Goals for)^E) + ((Goals against)^E))

That site takes this formula to come up with the team's pythagenpuck weighting based on goal differential and goal weight to come up with strength of match-ups. The site then simulates the rest of season 10 million times based on this weighting. Outcomes where team makes playoffs/Total outcomes of simulation = % chance to make playoffs.

Hammer79 is offline   Reply With Quote
Old
01-27-2011, 03:24 PM
  #20
Kaesar
Registered User
 
Join Date: Oct 2003
Posts: 485
vCash: 500
Quote:
Originally Posted by Hammer79 View Post
If you want the mathematical explanation for that sites formula, then look at page 11 of this pdf link:

http://www.hockeyanalytics.com/Resea...babilities.pdf

E = (Goals for per game + Goals against per game)^P
P = exponent, between 0 and 1, higher the number, the more emphasis that is placed on goal differential in predicting outcome.
Win Probability = ((Goals for)^E) / (((Goals for)^E) + ((Goals against)^E))

That site takes this formula to come up with the team's pythagenpuck weighting based on goal differential and goal weight to come up with strength of match-ups. The site then simulates the rest of season 10 million times based on this weighting. Outcomes where team makes playoffs/Total outcomes of simulation = % chance to make playoffs.
Thanks for posting that pdf

Kaesar 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 02:16 PM.

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

vBulletin Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2015 All Rights Reserved.